QUANTITATIVE RISK ASSESSMENT FOR OFFSHORE WIND FARMS IN THE NORTH SEA
Final Report
Report No. Date
: 23601.621/4 : May 3, 2010
Signature management:
M ARIN P.O. Box 28
6700 AA Wageningen The Netherlands
T +31 317 47 99 11 F +31 317 47 99 99
E
[email protected] I www.marin.nl
Report No. 23601.621/4
1
QUANTITATIVE RISK ASSESSMENT FOR OFFSHORE WIND FARMS IN THE NORTH SEA
MARIN order No.
:
23601.621
Ordered by
:
Ministry of Transport, Public Management Rijkswaterstaat Waterdienst P.O. Box 17 8200 AA LELYSTAD The Netherlands Number 31035696
Reported by
:
C. van der Tak MSc.
Works
and
Water
Report No. 23601.621/4
2
CONTENTS Page
GLOSSARY OF DEFINITIONS AND ABBREVIATIONS ................................................. 4 1
SUMMARY IN DUTCH ........................................................................................... 5 1.1 Introductie ............................................................................................................ 5 1.2 Kader van de SAMSON berekeningen ................................................................ 5 1.3 Werkwijze ............................................................................................................. 5 1.4 Effect van de windparken op de scheepvaart ...................................................... 7 1.5 Risico reducerende maatregelen ......................................................................... 9 1.6 Conclusies en aanbevelingen .............................................................................. 9
2
INTRODUCTION .................................................................................................. 11
3
OBJECTIVE OF THE SAMSON CALCULATIONS .............................................. 11
4
APPROACH ......................................................................................................... 12
5
INPUT FOR THE CALCULATIONS ..................................................................... 13 5.1 Search areas or main variants with traffic databases ........................................ 13 5.2 Traffic databases ............................................................................................... 13 5.3 The wind turbines............................................................................................... 18
6
IMPACT OF WIND FARMS ON SAFETY AND SHIPPING .................................. 20 6.1 Outside the wind farms ...................................................................................... 20 6.2 Collision risk of ramming and drifting against wind turbines .............................. 26 6.3 Total risk due to the wind farms ......................................................................... 30
7
RISK REDUCING MEASURES ............................................................................ 32
8
CONCLUSIONS AND RECOMMENDATIONS .................................................... 33
REFERENCES .............................................................................................................. 35 APPENDIX A 1
QUANTITATIVE RISK ASSESSMENT WITH SAMSON ................ 36
SAMSON.............................................................................................................. 36 1.1 Effect of a wind farm .......................................................................................... 37 1.2 Model input and assumptions ............................................................................ 38 1.2.1 Traffic .................................................................................................... 38 1.2.2 Used models ......................................................................................... 39 1.3 Consequences ................................................................................................... 40 1.3.1 Damage to the wind turbine .................................................................. 40 1.3.2 Personal damage .................................................................................. 42
APPENDIX B ALL VARIANTS WITH THE CORRESPONDING TRAFFIC DATABASE ...................................................................................................... 44
Report No. 23601.621/4
3
TABLE OF FIGURES Figure 5-1 Figure 5-2 Figure 5-3 Figure 5-4 Figure 5-5 Figure 5-6 Figure 5-7 Figure 6-1 Figure 6-2 Figure 6-3 Figure 6-4
Traffic flows for Variant 1 ....................................................................... 15 Traffic flows for Variant 2 ....................................................................... 15 Traffic flows for Variant 3 ....................................................................... 16 Traffic flows for Variant 4 ....................................................................... 16 Traffic flows for Variant 5 ....................................................................... 17 Traffic flows for Variant 6 ....................................................................... 17 Wind turbines placed in area 4a and 4b of Variant 3 ............................. 19 Calculation area for the impact of the wind farms on safety and shipping ................................................................................................. 20 Sub variant 3b growth ............................................................................ 28 Sub variant 5a growth ............................................................................ 28 Ship – wind turbine contacts for Variant 6a growth ................................ 29
TABLE OF TABLES Table 6-1 Overview of the different types of accidents distinguished in SAMSON. .............................................................................................. 21 Table 6-2 Impact of wind farms on shipping: results for the six variants for the year 2008 ............................................................................................... 23 Table 6-3 Impact of wind farms on shipping: results of variant – base for the year 2008 ............................................................................................... 24 Table 6-4 Impact of wind farms as percentage of the total in the main area of influence ................................................................................................ 25 Table 6-5 Probability of ramming and drifting contacts per year summarized over all wind turbines of the sub variant ................................................. 27 Table 6-6 Total extra costs of risk and shipping by the wind farms ............................ 31
Report No. 23601.621/4
4
GLOSSARY OF DEFINITIONS AND ABBREVIATIONS FSA
Formal Safety Assessment
IMO
International Maritime Organization
MKEA
Maatschappelijke kosteneffectiviteitsanalyse
N-ship
A non-route-bound ship. This ship mostly has a mission at sea, such as fishing vessels, supply vessels, working vessels and pleasure crafts.
NCS
Netherlands Continental Shelf
Probability
The probability (or number per year) is generally given with a large number of digits. This does not mean that the accuracy is so large, but the number of digits is used to make comparison possible between different items, also when the absolute values are small.
R-ship
A route-bound ship. It is a merchant ship or ferry sailing along the shortest route from one port to another.
SAMSON
Safety Assessment Model for Shipping and Offshore on the North Sea
TSS
Traffic Separation Scheme
Report No. 23601.621/4
1 1.1
5
SUMMARY IN DUTCH Introductie
De ambitie van de Nederlandse overheid is om een totaal vermogen van 6000 MW aan windenergie te realiseren op de Noordzee. Gezien de tegenstrijdige belangen van de verschillende groepen (stakeholders), die gebruik maken van de Noordzee, is het lastig om de meest geschikte locatie voor de windparken te bepalen. Er zijn verschillende studies uitgevoerd naar bepaalde aspecten van de problematiek. In dit rapport staat de impact van de windparken op de scheepvaart centraal. De kansen op en de kosten van extra ongevallen worden gekwantificeerd evenals de kosten van het omvaren. Deze studie kan als een vervolg gezien worden van de onderzoeken [1] en [2]. In [2] zijn de effecten voor de scheepvaart gekwantificeerd voor de varianten van 2 november 2009. Nadien zijn er discussies geweest met de stakeholders, die hebben geleid tot aanpassingen, verwerkt in [3]. In de maanden maart en april is nog een zesde variant uitgewerkt. Dit rapport is een update van rapport [3] waaraan de resultaten van deze nieuwe variant 6 zijn toegevoegd.
1.2
Kader van de SAMSON berekeningen
Doel van de in deze rapportage opgeleverde berekeningen is het leveren van informatie voor de besluitvorming over de mogelijkheden van windmolenparken op zee, met name in het zoekgebied voor de Hollandse kust. De zoekopdracht is hier ruimte te vinden voor 3000 MW. Voor de realisering hiervan zijn 6 varianten ontwikkeld, met hierbinnen diverse subvarianten. In alle varianten die in deze berekeningen met elkaar worden vergeleken, wordt in totaal 6000 MW vermogen aan windenergie geplaatst, conform de kabinetsdoelstelling. Met de hierbij uitgevoerde berekeningen wordt inzicht gegeven in de effecten die optreden bij verschillende varianten voor locatiekeuzes om deze 6000 MW te realiseren. Daarbij worden per variant de risico’s voor scheepvaart in beeld gebracht , alsook de effecten op vaarafstanden en de daarmee verband houdende extra kosten en uitstoot van milieubelastende stoffen. De uitkomsten van deze studie vormen een bouwsteen voor de Maatschappelijke kosten effectiviteitanalyse. Hierin worden alle effecten op een rij gezet en voor zover mogelijk in geld uitgedrukt ter ondersteuning van de besluitvorming. Ook vormt deze studie een bouwsteen voor een Formal Safety Assessment. Dit is een proces dat doorlopen dient te worden voor goedkeuring in IMO indien een activiteit kan leiden tot veranderingen van in internationaal (IMO) kader aan te wijzen scheepvaart routes. De stap hierbij, waarvoor deze berekeningen relevant zijn, is de kwantificering van risico’s van de ingreep voor de scheepvaart.
1.3
Werkwijze
Dezelfde werkwijze wordt gevolgd als in [2]. Dat wil zeggen dat er één verkeersdatabase wordt aangemaakt voor iedere variant die representatief is voor alle onderliggende subvarianten. De verkeersdatabase uit [2] voor de varianten 1, 2, 3, 4 kan nog steeds gebruikt worden. De verkeersdatabase voor variant 5 is opnieuw bepaald omdat er grote wijzigingen zijn aangebracht. In variant 5 van [2] was het Maas Noord stelsel geheel buiten dienst gesteld. Nu is dit stelsel vervangen door een
Report No. 23601.621/4
6
verkeersbaan met eenrichtingsverkeer dat meer in de NNW richting loopt (verder aangeduid met Maas NNW). Deze lane is bestemd voor het uitgaande verkeer in noordelijke en noordwestelijke richting. Verder is het Maas NW stelsel veranderd in een eenrichtingsverkeer stelsel met alleen inkomend verkeer. Het huidige uitgaande verkeer door Maas NW gaat nu via de Maas NNW lane. Maas NNW en Maas NW vormen nu samen een TSS waarbij de lanes een beetje gedraaid ten opzichte van elkaar staan. Het alternatief waarbij de vaarrichting is omgedraaid, dus bij Maas NNW naar binnen en bij Maas NW naar buiten is ook onderzocht, maar gaf een significant slechter resultaat voor de veiligheid. Daarom zijn alleen verdere berekeningen uitgevoerd voor de hiervoor geschetste situatie met alleen uitgaand verkeer door Maas NNW en alleen inkomend verkeer door Maas NW. De afwikkeling van het scheepvaartverkeer met de opgegeven zoekgebieden is weergegeven in Figure 5-1 tot en met Figure 5-5 van het hoofdrapport. Voor iedere variant zijn de zoekgebieden opgegeven waarbinnen de nieuwe windturbines geplaatst moeten worden. Bij de varianten 2, 3, 4, 5 en 6 wordt een “a” variant onderscheiden waarbij alle windturbines buiten de 12 mijlszone worden geplaatst en een “b” variant waarbij ook de strook tussen 10 en 12 mijl wordt gebruikt voor het plaatsen van windturbines. De totale oppervlakte van de zoekgebieden in iedere variant is groter dan nodig is voor het plaatsen van de windturbines tot een totaal van 6000MW. Deze extra ruimte wordt gebruikt om te voldoen aan andere criteria. De zoekgebieden zijn op twee manieren ingevuld. Eén waarbij wordt voldaan aan de criteria opgelegd met betrekking tot de kleine mantelmeeuw, de basis subvariant genoemd en één waarbij de ontwikkeling zo goedkoop mogelijk plaatsvindt welke de ontwikkel subvariant wordt genoemd. Bij de ontwikkel variant wordt er van uitgegaan dat de bescherming van de kleine mantelmeeuw geen beperking meer oplevert. Het combineren van de a en b variant met de basis en ontwikkel variant levert vier subvarianten op voor de varianten 2, 3, 4, 5 en 6. Variant 1 kent alleen onderscheid in de basis en ontwikkel subvariant. Voor het berekenen van de kans op een aanvaring (ramming) en aandrijving (drifting) met een windturbine moeten er windturbines geplaatst worden in de aangeleverde gebieden voor windparken. Voor het bepalen van de plaatsen van de windturbines is dezelfde methode gevolgd als in [2]. Deze methode bevat de volgende stappen: 1. 5 MW windturbines worden geplaatst in de hele Noordzee in een dichtste bolstapeling met een onderlinge afstand van 960m (8 maal de rotordiameter); 2. Alleen de windturbines worden meegenomen die in de aangegeven gebieden staan; 3. Windturbines die dichter dan 500m van een kabel staan worden verwijderd; 4. Windturbines langs de grenzen van een gebied op een afstand kleiner dan 500m tot een kabel worden niet verwijderd, om er voor te zorgen dat schepen niet binnen het windpark komen, wat anders volgens de huidige reglementen toegestaan zou zijn; 5. Er is aangenomen dat de vereiste afstand van 100m tot een pijpleiding geen echte beperking is en dat hieraan in het uiteindelijke ontwerp altijd kan worden voldaan. Wanneer op deze manier de aangeleverde gebieden voor de 18 subvarianten worden gevuld, blijkt dat het benodigde aantal windturbines niet wordt gehaald. De bedoeling was dat de windturbines in deze gebieden, samen met de al bestaande 225 MW en 500 MW in Wadden Noord voor de basis subvariant en 300 MW voor de ontwikkel subvariant, 6000 MW aan vermogen zouden opleveren. Het blijkt dat steeds 10 tot 20% minder vermogen geplaatst kan worden. Voor het kunnen vergelijken van de resultaten
Report No. 23601.621/4
7
van de berekeningen voor de subvarianten is daarom het resultaat van de berekeningen vermenigvuldigd met een factor gelijk aan het “verwachte aantal te plaatsen windturbines in de aangegeven windparken”, gedeeld door “het geplaatste aantal windturbines”. Hiermee wordt voor alle subvarianten het risico voor 1155 windturbines van 5 MW bepaald. Nadat alle windturbines geplaatst zijn moet de minimale afstand tot het passerende verkeer bepaald worden. Voor een inwendige windturbine kan dat veel meer zijn dan voor een windturbine aan de rand van het windpark.
1.4
Effect van de windparken op de scheepvaart
De belangrijkste effecten voor de scheepvaart zijn: • Verandering van de kans op aanvaringen tussen schepen onderling en kans op andere scheepsongevallen door de verandering van de scheepvaartafwikkeling; • Kosten van omvaren; • Het ontstaan van nieuwe ongevallen, namelijk aanvaringen en aandrijvingen met de nieuwe windturbines. De effecten van de eerste twee punten kunnen worden bepaald met de verkeersdatabases voor de zes varianten. Alle effecten worden bepaald per scheepstype en grootteklasse, waardoor de resultaten zo goed mogelijk in kosten kunnen worden omgezet. De kans op een aanvaring of aandrijving kan worden bepaald met de verkeersdatabase en de voor een subvariant bepaalde locaties van windturbines. Ook deze kansen worden zo goed mogelijk omgezet in kosten, waarbij wordt aangenomen dat de kans dat een windturbine verloren gaat gelijk is aan: • 100% na een aanvaring door een routegebonden schip; • 20% na een aanvaring door een niet-routegebonden schip; • 85% na een aandrijving door een routegebonden schip • 10% na een aandrijving door een niet-routegebonden schip; Hierbij wordt gerekend met M€ 10 voor een windturbine die verloren gaat. Een bedrag van M€ 1,04 wordt gebruikt voor de gemiddelde schadekosten van een schip bij een ongeval buiten het windpark. Vermoedelijk zal de schade aan het schip bij aanvaring of aandrijving van een windturbine kleiner zijn. Het resultaat van alle berekeningen is samengevat in Tabel 1-1. Het middelste deel van de tabel geeft de kosten per jaar van de aanvaringen en aandrijvingen met de windturbines en de omvaarkosten. Bij vergelijking van alle kosten blijkt dat de omvaarkosten meer dan 65% bijdragen in de extra kosten per jaar. Vervolgens is de schade aan de windturbines de grootste kostenpost. Alle varianten zijn gegeven in Appendix B.
Report No. 23601.621/4
Tabel 1-1
8
Extra risico en omvaarkosten voor de scheepvaart door de aanwezigheid van de windparken
basis
x
x
1 ontwikkel
x
2a basis
x
2a ontwikkel
x x
2b ontwikkel
x x
3a ontwikkel
x x
3b ontwikkel
x
4a basis
x
4a ontwikkel
x
4b basis 4b ontwikkel x
5a ontwikkel
x x
6a basis
x
6a ontwikkel
x
6b ontwikkel
x
18,056
0,545
0,328
0,873
0,086
0,408
13,093
18,338
0,578
0,328
0,906
0,086
4,585
0,408
13,093
18,086
0,553
0,328
0,881
0,086
4,557
5,224
0,408
13,093
18,725
0,630
0,328
0,957
0,086
0,500
3,280
3,780
0,604
15,773
20,157
0,461
0,514
0,975
0,154
0,457
3,046
3,503
0,604
15,773
19,880
0,428
0,514
0,941
0,154
0,3199
0,495
3,199
3,694
0,604
15,773
20,071
0,456
0,514
0,970
0,154
0,0050
0,2761
0,426
2,761
3,187
0,604
15,773
19,564
0,398
0,514
0,911
0,154
0,3303
0,0047
0,4043
0,581
4,043
4,624
0,664
16,316
21,605
0,539
0,505
1,044
0,110
0,3995
0,0054
0,5133
0,710
5,133
5,843
0,664
16,316
22,823
0,671
0,505
1,176
0,110
0,0147
0,3064
0,0047
0,3822
0,559
3,822
4,381
0,664
16,316
21,361
0,517
0,505
1,022
0,110
0,0703
0,0130
0,3420
0,0059
0,4311
0,620
4,311
4,932
0,664
16,316
21,912
0,584
0,505
1,089
0,110
0,0597
0,0135
0,3537
0,0047
0,4316
0,613
4,316
4,930
0,240
17,832
23,002
0.570
0.071
0.641
0.056
0,1077
0,0133
0,4074
0,0055
0,5339
0,737
5,339
6,077
0,240
17,832
24,149
0.697
0.071
0.768
0.056
0,0595
0,0152
0,3321
0,0048
0,4116
0,597
4,116
4,713
0,240
17,832
22,785
0.554
0.071
0.625
0.056
0,0561
0,0133
0,3534
0,0059
0,4287
0,621
4,287
4,909
0,240
17,832
22,981
0.585
0.071
0.656
0.056
0,0296
0,0139
0,3292
0,0050
0,3777
0,558
3,777
4,335
0,340
51,159
55,834
0,517
0,039
0,556
0,142
0,0542
0,0144
0,3274
0,0056
0,4016
0,590
4,016
4,606
0,340
51,159
56,105
0,555
0,039
0,594
0,142
0,0285
0,0154
0,3192
0,0051
0,3683
0,554
3,683
4,237
0,340
51,159
55,736
0,513
0,039
0,552
0,142
0,0287
0,0135
0,2893
0,0058
0,3373
0,515
3,373
3,888
0,340
51,159
55,387
0,483
0,039
0,522
0,142
0,590
4,028
4,618
0,326
9,208
0,3833
0,558
3,833
4,391
0,326
9,208
0,0625
0,0136
0,3164
0,0043
0,3967
0,588
3,967
4,555
0,408
0,0793
0,0144
0,3241
0,0046
0,4223
0,614
4,223
4,837
0,0631
0,0145
0,3168
0,0045
0,3989
0,596
3,989
x
0,0865
0,0162
0,3477
0,0053
0,4557
0,667
0,0087 0,0090
0,0144
0,3002
0,0047
0,3280
x
0,0111
0,2798
0,0046
0,3046
0,0084
0,0152
0,2914
0,0049
0,0066
0,0110
0,2535
0,0557
0,0137
0,0957
0,0127
0,0564
x x
x x
x x x
x
0,068
13,093
0,4028
0,0042
x
6b basis
0,068
0,751
0,0041
0,2980
x
5b ontwikkel
0,773
0,226
0,3160
0,0129
x x
0,226
0,525
0,0128
0,0683
x
5b basis
0,547
13,926
0,0699
x
x
5a basis
14,153
Totaal
x x
Totaal
Nschepen
x
3b basis
Buitenwindpark
Rschepen
x
x x
Alleen extra aanvaringen per jaar als voorbeeld
Met windpark
Schip + consequenties
Nschepen
x
Verwachte aantal extra ongevallen per jaar
Totaal extra kosten [M€/jaar]
Rschepen
x
2b basis 3a basis
ontwikkel
vanaf 10 nm
1 basis
Variant
vanaf 12 nm
Verwachte kosten per jaar voor 5775 MW nieuw geïnstalleerd vermogen (1155 windturbines van 5 MW) Kosten voor Kans dat een windturbine verloren gaat na een contact Kosten door contacten met de scheepvaart buiten het wind turbines windpark Subvariant aanvaren aandrijven [M€/jaar] [M€/jaar]
Windpark
Totaal
Extra ongevallen
Omvaren
Report No. 23601.621/4
1.5
9
Risico reducerende maatregelen
De berekeningen zijn uitgevoerd zonder risico reducerende maatregelen. De risico reducerende maatregel met het meeste effect is de inzet van een bergingsvaartuig (emergency towing vessel of ETV genoemd). Tot voor kort vervulde De Waker van de Nederlandse Kustwacht deze functie. Zodra een drifter werd gemeld en assistentie nodig was, werd De Waker er op af gestuurd. Omdat de kans op een storing groter is bij slecht weer condities en de driftsnelheid onder deze omstandigheden ook groter is, lag De Waker bij windkracht vanaf 5 Beaufort buiten gestationeerd, waardoor snel kon worden gereageerd. Door brand aan boord van De Waker is het schip sinds vorig jaar niet meer operationeel. Een ETV in de buurt van windparken kan de kans op een aandrijving met meer dan 50% reduceren.
1.6
Conclusies en aanbevelingen
Conclusies Uit de berekeningen kunnen de volgende conclusies worden getrokken: •
De aanpassingen in variant 5, uitgaande schepen via de Maas NNW lane en alleen inkomend verkeer door het Maas NW stelsel heeft geleid tot: o Reductie van het aantal extra aanvaringen van eens in de 6 jaar tot eens in de 18 jaar; o Een toename van de extra kans op een contact met een offshore platform van eens in de 100 jaar tot eens in 67 jaar, doordat Horizon, P15 en P18 dichterbij of bij een drukkere verkeersstroom komen te liggen; o Een kleinere kans op en contact met een ankerligger, omdat ankergebied 5A beter toegankelijk wordt zonder tegenliggend verkeer. Ook is er ruimte om het ankergebied 5A uit te breiden; o De omvaarkosten zijn afgenomen van M€ 24,8 tot M€ 17,8 per jaar door de korte vaarweg via de gecreëerde Maas NNW lane.
•
De grootste bijdrage (meer dan 65%) van de jaarlijkse kosten wordt geleverd door de omvaarkosten. Deze kosten zijn: Variant 1 M€ 9,2 Variant 2 M€ 13,1 Variant 3 M€ 15,8 Variant 4 M€ 16,3 Variant 5 M€ 17,8 Variant 6 M€ 51,2
•
Daarna is de schade aan windturbines de grootste kostenpost. Deze varieert van M€ 2,8 voor subvariant 3b ontwikkel tot M€ 5,3 voor subvariant 5a ontwikkel;
•
De aandrijfkans en vooral de aanvaarkans hangt sterk af van de manier waarop de zoekgebieden worden ingevuld met wind turbines. Het risico varieert significant tussen de subvarianten van dezelfde variant.
•
De kleinste variatie in risico treedt op bij variant 3 omdat het verkeer in deze variant verder weg van de zoekgebieden wordt geleid.
Report No. 23601.621/4
10
Aanbevelingen Een emergency towing vessel in de buurt van de windparken kan het aandrijfrisico met meer dan 50% reduceren. De huidige berekeningen zijn uitgevoerd zonder inzet van een ETV. De verkeersafwikkeling kan op sommige plekken nog veiliger worden gemaakt. Dit vereist maatwerk. In het kader van de FSA worden hiervoor aanbevelingen gedaan. Aannames In de berekeningen is aangenomen dat een driftend schip niet stopt na een aandrijving tegen een windturbine maar verder drift en daarbij weer andere windturbines kan raken. Dus één drifter kan meerdere windturbines raken die allemaal geteld worden. De extra kans op een contact met een offshore platform is berekend en meegenomen in de tabellen. De schade aan het platform is niet meegenomen omdat geen data beschikbaar is over dit soort schades. Uiteenzetting voor een beter begrip van de kans op een aanvaring De kans op een aanvaring wordt bepaald uit het aantal berekende ontmoetingen tussen schepen. De ontmoeting wordt geclassificeerd op basis van het koersverschil tussen de schepen. Er wordt onderscheid gemaakt tussen een overtaking (koersverschil kleiner dan 60°), crossing (koersverschil tussen 60° en 150°) en een head-on (koersverschil tussen 150° en 180°) ontmoeting. De kans op een aanvaring gegeven een ontmoeting is voor de drie situaties duidelijk verschillend. De kans op een aanvaring gegeven een crossing is veel groter dan bij een overtaking. Soms is deze discrete classificatie van ontmoetingen te grof voor het doen van uitspraken in een klein gebied. Verder is de kans op een aanvaring gebaseerd op enkelvoudige ontmoetingen. Wanneer een derde schip bij de ontmoeting betrokken is wordt de kans op een aanvaring groter. Het effect van het derde schip komt naar voren in het aantal ontmoetingen. Bij een derde schip is het aantal ontmoetingen voor ieder schip twee keer zo groot, daar de ontmoetingen voor ieder schip afzonderlijk geteld worden. Bij een ontmoeting tussen drie schepen worden zes ontmoetingen geteld en bij een ontmoeting tussen twee schepen twee ontmoetingen. Dit geeft al aan dat de kans op een aanvaring bij een ontmoeting tussen drie schepen drie keer zo groot is als de kans op een aanvaring tussen twee schepen, dus bij drie afzonderlijke ontmoetingen tussen twee schepen, die geteld zouden worden wanneer de drie schepen elkaar niet tegelijk ontmoeten maar in drie afzonderlijke ontmoetingen. De kans op een aanvaring gegeven een ontmoeting is niet geografisch gerelateerd en hangt dus niet af van de afstand tot een object, zoals een platform of een windturbine, dat de uitwijkmogelijkheden beperkt. Er gebeuren te weinig aanvaringen om de ongevalskansen voor dit soort situaties goed te kunnen modelleren en te kwantificeren. De kans op een aanvaring gegeven de ontmoeting is wel afhankelijk van het scheepstype en de scheepsgrootte en ook van de weersomstandigheden. Voor het berekenen van het aantal ontmoetingen is gerekend met een snelheidsverdeling voor varen op zee. In sommige gebieden wordt ook met lagere snelheden en met meer snelheidsverschillen gevaren waardoor het aantal ontmoetingen verandert. Voor deze gebieden zou de verkeersafwikkeling in meer detail kunnen worden beschreven. Het aantal ontmoetingen zou dan beter kunnen worden bepaald. Een nieuw probleem wordt dan hoe de aanvaringskans verandert met de lagere snelheid van de betrokken schepen. Maar het verschil tussen twee varianten kan wel beter inzichtelijk worden gemaakt.
Report No. 23601.621/4
2
11
INTRODUCTION
The ambition of the Dutch government is to realize offshore wind farms with a total installed power of 6000 MW. The question “where in the North Sea” is subject of many discussions because of the opposite interests of different stakeholders. Several studies on different topics already have been performed. The topic of this report is the impact of the wind farms on the safety of shipping. Also aspects as costs of extra miles and emissions by detours around the wind farms are quantified. Based on earlier results and discussions with stakeholders, the distribution and composition of the wind farms has changed over time. The impacts on safety and shipping of previous variants are published in [1], [2] and [3]. The versions of the variants described in this report are: • Variants and sub variants 1, 2, 3 and 4 of February 2, 2010; • Variant 5 of February 2, 2010, with only outgoing traffic through the Maas North TSS and only incoming traffic through the Maas Northwest TSS; • Variant 6 of April 2, 2010 This report contains the results of the latest versions of all variants and replaces all earlier results. The report contains the following chapters: Chapter 3 gives the objective of the study. Chapter 4 describes the approach followed. Chapter 5 describes how the input for the calculations is determined. Chapter 6 contains the results of the calculations. Chapter 7 contains further considerations of the result. Chapter 8 contains the conclusions and recommendations.
3
OBJECTIVE OF THE SAMSON CALCULATIONS
The objective of the calculations presented in this report is to provide information for the decision-making process with respect to the possibilities of offshore wind farms, in particular within the search area “Hollandse kust”. The search objective is to find space for 3000 MW in this area. For the realisation of this, 5 variants have been developed, including various sub variants. In each of the variants that are compared within the calculations, a total of 6000 MW power of wind energy is installed, in accordance with the objective of the government. With the results of the calculations, the effects can be compared for the various variants when 6000 MW power of wind energy is realized. For each of the variants, the risks to shipping are quantified, as well as the effects of detours for shipping and the related additional costs and emissions of polluting substances. The results of this study form a building block for the social costs effectiveness analysis. (in Dutch: Maatschappelijke Kosten Effectiviteits Analsyse or MKEA). Herein all effects are listed and monetarised as far as possible, to support the decision-making process. This study also serves as a building block for a Formal Safety Assessment. This is a process that is necessary to go through for approval in the IMO, if an activity may lead to changes in international (IMO) shipping routes. One of the steps, for which these calculations are relevant, is the quantification of the impact on safety for shipping by the changes in the shipping routes.
Report No. 23601.621/4
4
12
APPROACH
The same approach is followed as in November 2009 of which the results are described in [2].The impacts for shipping are quantified with the SAMSON model that has been developed, extended, validated and improved continuously during the last 25 years in studies performed for the Dutch Maritime Safety Authority, European projects and commercial projects. A brief description of the model is given in Appendix A. The following effects can be quantified with the SAMSON the model: • Expected number of accidents per year, divided over type of accident, ship types and sizes involved in the accidents and objects; • Extra miles that have to be travelled as a result of a certain development and the costs involved; • Emitted environmental dangerous goods, e.g. exhaust; • Consequences of the accidents such as the outflow of oil or personal injuries. For quantifying the risk, the model requires the following input: • Description of the geographical area; • A traffic database describing all traffic flows; • Current; • Wind compass; The impact of the offshore wind farms is quantified by comparing the risk of the case with the wind farms with the risk of the base case without the wind farms. An offshore wind farm introduces an additional type of risk, being the risk that a wind turbine is struck by a ship. Two types of collision risk are distinguished, namely: • A ramming contact takes place when a ship is on a collision course with a wind turbine and a navigational error occurs. A navigational error can have various causes, like lack of information, not being able to see the wind farm, not being present on the bridge, getting unwell and not being able to act, making an error etc. A ramming contact will take place with high speed: 90% of the service speed of a vessel. • A drifting contact occurs when a ship in the vicinity of a wind turbine experiences a failure in the propulsion engine or in the steering equipment. Since the ship slowly becomes uncontrollable as it loses speed, the combined effect of wind, waves and current may carry the ship towards the wind turbine. If dropping an anchor does not help or is not practical and the repair time exceeds the available time, the ship may collide against a turbine. This generally happens at a low speed. The probability of a contact with a wind turbine depends on where the traffic flows are with respect to the wind farms. In order to be quantify this ramming and drifting risk the offshore wind farms have to be filled in with wind turbines with a real geographical position.
Report No. 23601.621/4
5 5.1
13
INPUT FOR THE CALCULATIONS Search areas or main variants with traffic databases
Until now, six variants have been developed that require a quantification of the impact. Each of these six variants contains a number of “search” areas that are identified for wind energy. All shipping traffic has to be routed outside these search areas. The summarized surface of the search areas of each variant is larger than required for the 6000 MW. For this reason, different realizations of the 6000 MW are determined based on different aspects. Within variant 2, 3, 4, 5 and 6, four sub variants are distinguished by the choice whether or not to use the 10-12 nm zone outside the coast, and by filling the search areas with wind turbines, based on either ecological considerations or based on a growth scenario. Variant 1 contains only two variants based on the way of filling up the areas. The largest impact on shipping and safety outside the wind farm will occur when all search areas belonging to the variant are cleared from shipping. This is the worst case scenario for that variant, with the largest increase of extra miles for shipping and most likely the largest increase of the probability of accidents. This approach creates the possibility to research different realizations for the 6000 MW without changing the traffic database. In reality, it would have been better to develop a new traffic database for each sub variant, because ships will plan their routes around existing wind farms, based on the regulations, water depths and the shortest distance. However, it is not expected that a more accurate traffic database for each individual sub variant will affect the results of the present calculations significantly. In general, the impact on safety and extra miles will, for a specific sub variant, be less than quantified with this approach, while the contacts with the wind turbines will be slightly underestimated because the traffic flows are modelled outside the search areas instead of only outside the areas with wind turbines. It is expected that this has a minor impact because the largest contribution to the risk for wind turbines is delivered by drifters and this contribution is less sensitive to the distance. Furthermore, ships, using routes that will not change for a sub variant, will deliver the main contribution. This approach is chosen for these global calculations and is accurate enough to balance the pros and cons of the different variants and sub variants.
5.2
Traffic databases
A variant contains a number of search areas combined with the offshore wind farms that are licensed in the second round. All shipping movements that have crossed the North Sea in 2008 are rerouted for the case that the second round wind farms and the search areas are completely filled with wind turbines. The traffic flows for each variant with the search areas and with the wind farms from the second round that can be built within the variant, are presented in Figure 5-1 to Figure 5-5. The traffic databases of November 2009 for Variant 1 to Variant 4 are unchanged. Only the traffic flow of Variant 5 is changed significantly. In Variant 5 of November 2009 the Maas North TSS was removed. In the new Variant of February, an outgoing (from Rotterdam) traffic lane (Maas NNW) replaces the Maas North TSS. Furthermore the Maas NW TSS contains only an ingoing traffic lane.
Report No. 23601.621/4
14
The main changes for the traffic are: • The traffic flow from Rotterdam through the Texel TSS to the North, Hamburg and Baltic is are shortened; • The traffic from Rotterdam to the Hull area is routed through the new Maas NNW lane; • The tankers that have to use the Deep Water Route TSS are now routed through Maas NNW instead of Maas NW. • Ships coming from the North and North West are better able to reach the anchorage area 5A. As alternative, also the case has been analyzed that outgoing ships use the Maas NW lane and ingoing ships use the new Maas NNW, thus a circulation of the traffic in the opposite direction. Because, the results for this alternative were worse, this alternative is not taken into account anymore. The development of Variant 6 has started in February 2010, as a new variant in which a large area off the coast of Zuid-Holland is made free of shipping for wind energy. This variant was not yet implemented in the report of March 18, [3]. The variant has large implications for the TSSs in the approach to Rotterdam. Maas North and Maas Northwest are closed and the remaining TSSs are changed. The effect is that the traffic over the North Sea will sail more concentrated on main routes. The traffic flows for Variant 6 are described in a complete new traffic database. Because of the large changes, this variant requires some more detailed analysis on some spots in case the variant is selected.
Report No. 23601.621/4
Figure 5-1
Traffic flows for Variant 1
Figure 5-2
Traffic flows for Variant 2
15
Report No. 23601.621/4
Figure 5-3
Traffic flows for Variant 3
Figure 5-4
Traffic flows for Variant 4
16
Report No. 23601.621/4
Figure 5-5
Traffic flows for Variant 5
Figure 5-6
Traffic flows for Variant 6
17
Report No. 23601.621/4
5.3
18
The wind turbines
As described before, the summarized surface of the search areas and second round wind farms exceeds the total area required for 6000 MW. This makes a flexible filling up of the areas possible. This flexibility is used to fulfil other requirements. In the “a” sub variant only areas outside the 12 nautical mile line from the coast are filled. In the “b” sub variant also areas between the 10 and 12 nautical mile from the coast are filled. Furthermore, the “base” sub variant meets the requirements with respect to the great black-backed gull and the “growth” sub variant starts with filling up the search areas as cheap as possible, assuming that the protection of the great black-backed gull will give no limitations. In this way four sub variants are designed for each variant. Only variant 1 has only two sub variants because only areas outside the 12 nautical mile are considered. The designation of the areas for the 22 sub variants has been executed before this study and is fixed input. It is assumed that these areas deliver 6000 MW combined with: • 225 MW and 300 MW • 500 MW in the area Wadden North for the base sub variant and 300 MW in Wadden North for the growth sub variant. Appendix B contains the wind farm areas filled with turbines for all variants with the corresponding traffic database. For the calculation of the ramming and drifting risk it is necessary to have real wind turbines instead of an area. For this reason, the areas delivered are filled up with wind turbines. Because it is very time consuming to develop the optimal configuration of wind turbines for each of the areas within a sub variant a more pragmatic approach is followed. The approach, already explained and used in [2], contains the following steps: 1. 5 MW wind turbines are placed in the whole North Sea according to a close packing of spheres with a distance of 960m (8 times the rotor diameter); 2. Only the wind turbines within the designated areas are collected; 3. Wind turbines closer than 500m to a cable are removed; 4. Wind turbines along the borders of the area within 500m of a cable are not removed, to avoid that ships will enter, or better are allowed to enter, the wind farm; 5. It is assumed that the required 100m distance to pipeline will not be a real restriction and this requirement can be fulfilled in the final design. The result of this process for area 4a and 4b of Variant 3 is shown in Figure 5-7. The red striped lines are cables. Some cables are not in use anymore. Which ones, is not visible in this figure, but for this reason the inside wind turbines closer than 500m to some cables are not removed. The figure shows that the wind turbines along the borders are not removed. The pipelines (grey striped lines) are ignored. Generally, the process results in a number of wind turbines that is 10 to 20% less than assumed by the surface offered, due to the cables and use of smaller areas. Because the risk is strongly related to the number of wind turbines, a multiplier is used to upgrade the collision risk. This multiplier is different for each sub variant and is the number of assumed wind turbines within the sub variant divided by the number of wind turbines that could be placed.
Report No. 23601.621/4
19
After all wind turbines were determined within a sub variant, the minimum distance from each individual wind turbine to the passing shipping was determined, because this value is part of the input for each wind turbine.
Figure 5-7
Wind turbines placed in area 4a and 4b of Variant 3
Report No. 23601.621/4
6
20
IMPACT OF WIND FARMS ON SAFETY AND SHIPPING
This chapter contains the results of the calculations for the impact of the wind farms on safety and shipping. Chapter 6.1 gives the result for safety and shipping by the changed routes and larger concentration of ships over some routes due to the presence of the wind farms. Chapter 6.2 contains the effect in expected ramming and drifting contacts against the wind turbines.
6.1
Outside the wind farms
The impacts on shipping and safety outside the wind farm area are related to the traffic database. This means that these results are the same for all sub variants of a variant, because they all use the same traffic database. The influence of the new traffic database of variant 5 stretches over a larger area than was used in [2]. Therefore, the calculation area for quantifying the effects is enlarged to the calculation area presented in Figure 6-1. Of course, most effects will occur on the Netherlands Continental Shelf (NCS).
Figure 6-1
Calculation area for the impact of the wind farms on safety and shipping
Report No. 23601.621/4
21
The impacts are calculated with SAMSON. Within the models 36 ship types and 8 ship sizes classes are distinguished within the route-bound ships. All results, thus probabilities of accidents and extra miles, are calculated for each ship type and size class. This makes it possible to quantify the consequences in monetary terms, for which references [4] to [9] are used. The probability of each type of accident is related to the calculated exposures from the traffic database. An exposure is a kind of dangerous situation that can result in an accident. For example, the encounter between two ships is the exposure for a collision. An encounter occurs when one ship penetrates the domain around another ship. Within the calculations three types of encounters are distinguished based on the difference in course, namely the overtaking (<60°), crossing (60°-150°) and head-on (150°-180°) encounter. This distinction is made because the probability that the encounter ends up in an accident differs for these three types of encounters. The different type of accidents defined in SAMSON with their cause and exposure measure can be found in Table 6-1. The ramming opportunity and the danger miles are derived from the traffic flow in which the ship type, size and distance are used. Table 6-1
Overview of the different types of accidents distinguished in SAMSON.
Type
Subtype
Ship/Ship collision Ship/Anchored ship Ramming ship/anchored ship in port collision approach area Ramming ship/anchored ship in anchor area Ramming ship/anchored ship outside port approach area or anchor area Drifting ship/anchored ship
Cause
Exposure measure
Human error
Encounter
Human error Human error Human error
Ship miles
Engine failure
Danger miles
Ramming anchored ship during anchoring Drifting anchored ship/ anchored ship Engine failure after breaking anchor chain Contacts
Anchored ship ramming opportunity Anchored ship ramming opportunity
Danger miles Danger miles
Objects with fixed position
Engine failure
Danger miles
Objects with fixed position
Navigational error
Ramming opportunity
Contacts with object that can take place everywhere.
Ship miles
Foundering (Sinking)
Ship miles
Fire and Explosion
Ship miles
Stranding Incident Hull/Machinery
Engine failure
Danger miles
Navigational error
Stranding opportunity Ship Miles
In [2] the calculations have been performed and compared for the traffic databases of 2004 and 2008. It could be concluded that changes in the intensities and ship types and sizes had not caused a movement in the relation of the variants to each other. Therefore the calculations are now only performed for the year 2008.
Report No. 23601.621/4
22
Table 6-2 contains the results of the calculation summarized over all ship types and ship size classes. It shows that on average, nearly 600 route-bound ships are at sea in the calculation area. Yearly about 24 ships are involved in collisions, which corresponds with 12 collisions because nearly always two ships are involved in one collision at sea. The base case gives that 66,7 million nautical mile are sailed, thus on average about 110 000 nautical mile per ship per year. The total shipping costs amount M€ 4861 per year, which includes fixed costs and fuel costs. The first column with results shows the values for the present situation. The next six columns the values for the traffic databases of the six variants. The extra accidents that are expected per year and the extra costs by the wind farms are presented in Table 6-3, in which the results of the “base for 2008” are subtracted from the results for each variant. Thus Table 6-3 shows the impact of the wind farms on safety and shipping outside the wind farms. The costs are estimated based on the collected data and models of SAMSON. The collision costs contain the costs of deaths, pollution, salvage, repairing, delay, loss in income, collected within several projects and publications from P&O clubs. All costs are indexed with 3% per year. The costs of contacts with offshore platforms are not included completely. Only the damages to the ship are included. The costs of damages to offshore platforms could not be quantified because no data is available for this type of consequences. The extra emissions are quantified. The corresponding societal costs are taken into account in the MKEA. For a better understanding of the changes in safety presented in Table 6-3, these values are given as percentage of the level without any wind farm. Because the calculation area is taken very large to ensure that all changes are included, a rough estimate is made of the area, where the most changes will occur. The main area of impact is the Dutch area off the coast. About 20% of all movements will take place here, indicated by the 20% in the column for collisions. The influence area for collisions with ships in the anchorage areas is set to 100%, because only the anchorage areas for the Western Scheldt and the Dutch ports are included in the calculations. Applying the estimated percentages for the influence areas, Table 6-4 contains the changes in percentages in safety and economics. For example this means for Variant 6 that: • a collision with a ship at anchor decreases by just over 13%; • the collision risk for platforms decreases by 3 to 4%; • the risk of collision with another sailing ship increases by 5.9%; • the foundering, hull failures and fire/explosion and also the costs change roughly by the number of shipping miles;
Report No. 23601.621/4
Table 6-2
23
Impact of wind farms on shipping: results for the six variants for the year 2008 Unit
Average number of ships in the area OBO's Ships Chemical tankers Ships ships Oil tankers ships Gas tankers ships Bulkers ships Unitised ships General Dry Cargo ships Passengers + conv, ferries ships High Speed Ferries ships Other ships Total Route-bound ships Total Non-route-bound Safety Ships involved in collisions ships Stranding after navigat, failure ships Stranding after technical failure ships Ramming against platform ships Drifting against platform Ramming against ship at anchor ships ships Drifting against ship at anchor ships Foundering ships Hull Failure ships Fire/ Explosion ships Total Economy Shipping costs, fixed + fuel M€ / year Ship miles Mnm/year Emissions KW used GWh CO2 kton / year CO kton / year kton / year SO2 kton / year NOx Oil Shipping accidents probability/yr probability/yr Chem+olie tankers in accidents probability/yr Oil tanker in accidents probability/yr Oil spills probability/yr Oil spill more than 10000 m3 probability/yr Oil spill more than 30000 m3 probability/yr Oil spill more than 100000 m3 Oil spilt m3 / year Chemical spills after collision probability/yr Very Large Ecological Risk probability/yr Large Ecological Risk probability/yr Medium Ecological Risk probability/yr Low Ecological Risk probability/yr Negligible Ecological Risk Costs collisions and foundering Repairing M€ /year M€ /year Salavage Cleaning and environment,costs M€ /year M€ /year Costs of delay M€ /year Loss in income M€ /year Willingness to pay for deaths M€ /year Ship+cargo when sinking Total M€ /year
Base for 2008
Variant 1
Variant 2
Variant 3
Variant 4
Variant 5
Variant 6
0,982 80,975 41,586 23,458 44,546 128,402 232,009 14,168 0,184 28,656 594,966 281,650
0,986 81,455 41,762 23,476 44,571 128,423 232,102 14,179 0,184 28,658 595,796 281,650
0,987 81,511 41,790 23,480 44,623 128,510 232,428 14,191 0,184 28,707 596,411 281,650
0,983 81,453 41,753 23,494 44,646 128,649 232,714 14,202 0,184 28,765 596,843 281,650
0,987 81,536 41,826 23,484 44,633 128,587 232,588 14,194 0,184 28,842 596,861 281,650
0,985 81,469 41,729 23,504 44,644 128,813 232,895 14,210 0,184 28,757 597,190 281,650
0,986 82,207 41,811 23,608 44,779 130,023 234,733 14,256 0,184 29,091 601,678 281,650
24,028 9,168 1,505 0,146 0,075 2,554 0,025 3,244 3,970 7,164 51,878
24,163 9,166 1,504 0,147 0,075 2,631 0,025 3,244 3,979 7,171 52,104
24,200 9,168 1,504 0,146 0,075 2,685 0,025 3,246 3,982 7,174 52,205
24,335 9,160 1,502 0,148 0,075 2,732 0,026 3,251 3,981 7,181 52,391
24,248 9,160 1,501 0,149 0,075 2,804 0,026 3,250 3,985 7,183 52,382
24,139 9,160 1,502 0,157 0,079 2,475 0,024 3,253 3,980 7,180 51,948
24,311 9,145 1,510 0,143 0,072 2,216 0,023 3,271 4,004 7,223 51,917
4861 66,704
4870 66,792
4874 66,859
4877 66,910
4878 66,907
4879 66,949
4912 67,449
29517 14693 56,540 150,004 377,050
29565 14716 56,626 150,243 377,645
29585 14727 56,668 150,351 377,914
29604 14737 56,706 150,448 378,156
29607 14737 56,707 150,453 378,177
29621 14746 56,740 150,539 378,379
29851 14858 57,170 151,683 381,248
33,4712 8,1521 2,9411 0,3902 0,0727 0,0382 0,0154 5353
33,6010 8,2401 2,9741 0,3941 0,0736 0,0385 0,0155 5396
33,6475 8,2481 2,9771 0,3946 0,0736 0,0385 0,0155 5398
33,7795 8,2512 2,9732 0,3949 0,0735 0,0385 0,0155 5397
33,7025 8,2762 2,9912 0,3958 0,0739 0,0387 0,0155 5418
33,5817 8,2001 2,9561 0,3922 0,0731 0,0383 0,0154 5373
33,7790 8,2212 2,9502 0,3920 0,0728 0,0381 0,0152 5339
0,0214 0,0043 0,0120 0,0325 0,0457
0,0214 0,0043 0,0124 0,0327 0,0466
0,0214 0,0044 0,0124 0,0327 0,0467
0,0212 0,0043 0,0125 0,0326 0,0473
0,0213 0,0043 0,0125 0,0327 0,0472
0,0210 0,0043 0,0124 0,0322 0,0468
0,0204 0,0042 0,0128 0,0318 0,0481
3,130 1,565 15,963 0,585 3,844 8,946 25,526 59,559
3,157 1,579 16,176 0,587 3,873 8,987 25,526 59,886
3,166 1,583 16,192 0,589 3,884 9,007 25,545 59,967
3,188 1,594 16,209 0,597 3,917 9,020 25,638 60,163
3,182 1,591 16,315 0,590 3,899 9,017 25,629 60,223
3,143 1,572 16,020 0,589 3,865 8,957 25,652 59,800
3,147 1,573 15,840 0,593 3,881 9,018 25,847 59,899
Report No. 23601.621/4
Table 6-3
24
Impact of wind farms on shipping: results of variant – base for the year 2008 Unit
Average number of ships in the area OBO's Chemical tankers Oil tankers Gas tankers Bulkers Unitised General Dry Cargo Passengers + conv, ferries High Speed Ferries Other Total Route-bound Total Non-route-bound Safety Ships involved in collisions Stranding after navigational failure Stranding after technical failure Ramming against platform Drifting against platform Ramming against ship at anchor Drifting against ship at anchor Foundering Hull Failure Fire/ Explosion Total Economy Shipping costs, fixed + fuel Ship miles Emissions KW used CO2 CO SO2 NOx Oil Shipping accidents Chem+olie tankers in accidents Oil tanker in accidents Oil spills Oil spill more than 10000 m3 Oil spill more than 30000 m3 Oil spill more than 100000 m3 Oil spilt Chemical spills after collision Very Large Ecological Risk Large Ecological Risk Medium Ecological Risk Low Ecological Risk Negligible Ecological Risk Costs collisions and foundering Repairing Salavage Cleaning and environmental costs Costs of delay Loss in income Willingness to pay for deaths Ship+cargo when sinking Total
Variant 1 Variant 2 Variant 3 Variant 4 – base – base – base – base
Variant 5 - base
Variant 6 - base
ships ships ships ships ships ships ships ships ships ships ships ships
0,004 0,480 0,176 0,018 0,025 0,021 0,093 0,011 0,000 0,002 0,830 0,000
0,005 0,536 0,204 0,022 0,077 0,108 0,419 0,023 0,000 0,051 1,445 0,000
0,001 0,478 0,167 0,036 0,100 0,247 0,705 0,034 0,000 0,109 1,877 0,000
0,005 0,561 0,240 0,026 0,087 0,185 0,579 0,026 0,000 0,186 1,895 0,000
0,003 0,494 0,143 0,046 0,098 0,411 0,886 0,042 0,000 0,101 2,224 0,000
0,004 1,232 0,225 0,150 0,233 1,621 2,724 0,088 0,000 0,435 6,712 0,000
ships ships ships ships ships ships ships ships ships ships
0,135 -0,002 -0,001 0,001 0,000 0,077 0,000 0,000 0,009 0,007 0,226
0,172 0,001 -0,001 0,000 0,000 0,131 0,001 0,002 0,012 0,010 0,328
0,307 -0,008 -0,003 0,002 0,001 0,178 0,001 0,007 0,011 0,017 0,514
0,220 -0,008 -0,004 0,004 0,001 0,250 0,001 0,006 0,015 0,019 0,505
0,111 -0,008 -0,003 0,011 0,004 -0,079 -0,001 0,009 0,010 0,016 0,071
0,283 -0,023 0,005 -0,003 -0,002 -0,338 -0,002 0,027 0,034 0,059 0,039
M€ / year Mnm/year
9,208 0,088
13,093 0,155
15,773 0,206
16,316 0,203
17,832 0,245
51,159 0,745
GWh kton / year kton / year kton / year kton / year
48,263 22,979 0,086 0,238 0,595
68,720 33,790 0,128 0,346 0,863
87,737 43,556 0,166 0,444 1,106
90,264 44,008 0,167 0,449 1,126
104,484 52,507 0,200 0,534 1,328
334,285 164,871 0,630 1,678 4,197
probability/yr probability/yr probability/yr probability/yr probability/yr probability/yr probability/yr m3 / year
0,1298 0,0880 0,0330 0,0039 0,0008 0,0004 0,0001 43,215
0,1763 0,0960 0,0360 0,0044 0,0009 0,0004 0,0001 45,392
0,3083 0,0991 0,0321 0,0046 0,0008 0,0004 0,0001 43,923
0,2313 0,1241 0,0501 0,0055 0,0011 0,0005 0,0002 65,751
0,1105 0,0480 0,0150 0,0020 0,0004 0,0002 0,0000 20,133
0,3078 0,0691 0,0091 0,0018 0,0001 -0,0001 -0,0002 -13,636
probability/yr probability/yr probability/yr probability/yr probability/yr
0,0000 0,0000 0,0004 0,0002 0,0009
0,0000 0,0000 0,0004 0,0002 0,0010
-0,0002 0,0000 0,0005 0,0001 0,0017
-0,0001 0,0000 0,0005 0,0002 0,0015
-0,0004 -0,0001 0,0004 -0,0002 0,0011
-0,0010 -0,0001 0,0008 -0,0007 0,0024
0,027 0,014 0,213 0,003 0,029 0,041 0,000 0,326
0,036 0,018 0,229 0,004 0,041 0,061 0,019 0,408
0,058 0,029 0,246 0,012 0,073 0,074 0,112 0,604
0,052 0,026 0,352 0,006 0,055 0,071 0,103 0,664
0,013 0,007 0,057 0,005 0,022 0,011 0,126 0,240
0,016 0,008 -0,123 0,009 0,037 0,072 0,320 0,340
M€ /year M€ /year M€ /year M€ /year M€ /year M€ /year M€ /year M€ /year
Report No. 23601.621/4
Table 6-4
25
Impact of wind farms as percentage of the total in the main area of influence
Safety
Ships involved in collisions Stranding after navigational failure Stranding after technical failure
Additional by wind farms Rough estimate of influence Var 1 var 2 area 20%
Variant - base Var 3
Var 4
Var 5
Var 6
2,81% 3,58% 6,39% 4,58% 2,31%
5,89%
50%
Ships
-0,05% 0,01% -0,17% -0,17% -0,17%
-0,50%
50%
Ships
-0,12% -0,16% -0,40% -0,53% -0,37%
0,65%
Ramming against platform
70%
Ships
1,19% -0,09% 2,35% 3,52% 10,60%
-3,10%
Drifting against platform
70%
Ships
0,62% 0,57% 1,76% 1,64% 8,02%
-4,47%
Ramming ship at anchor
100%
Ships
3,00% 5,14% 6,96% 9,80% -3,08% -13,24%
Drifting against ship at anchor
100%
Ships
0,77% 2,65% 5,03% 5,93% -3,54%
-7,58%
Foundering
20%
Ships
0,00% 0,31% 1,08% 0,92% 1,39%
4,16%
Hull Failure
20%
Ships
1,13% 1,51% 1,39% 1,89% 1,26%
4,28%
Fire/ Explosion
20%
Ships
0,49% 0,70% 1,19% 1,33% 1,12%
4,12%
Economy route bound ships Shipping costs Ship miles
20% 20%
0,94% 1,34% 1,61% 1,67% 1,82% 0,66% 1,16% 1,54% 1,52% 1,84%
5,25% 5,58%
Report No. 23601.621/4
6.2
26
Collision risk of ramming and drifting against wind turbines
All wind turbines are defined as separate objects. The probability of ramming and drifting is determined for each wind turbine. It is assumed that a drifting ship continues drifting after a wind turbine is hit. Thus one drifting ship can hit more than one wind turbine. The ship stops drifting when: • anchoring is successful (no contact at all); • the failure is repaired, using the distribution function for repairing; • 24 hours are passed (the maximum drifting time); • an emergency towing vessel (ETV) has arrived with sufficient capacity for the prevailing weather conditions to recover the ship; In the present calculations no ETV is included, which means that the probability of drifting contacts can be decreased by including one or more ETVs. Table 6-2 contains the probability of a ramming and drifting contact per year for each of the 22 sub variants. The table shows that drifting by route-bound ships (R-ships) delivers the largest contribution. The expected number of contacts varies from 0,4097, once in 2,4 years for the sub variant 3b growth to 0,7088, once in 1,4 years for sub variant 5b growth. In the right part of the table the probabilities are compared with each other for each type of contact separately. The 3b growth variant contains in this part four times 1,00, which means that this sub variant delivers the lowest probability of contacts with wind turbines. This is clearly the effect of the largest distance between the wind turbines and the traffic flows of passing traffic. It is especially visible when comparing the ramming contact of route-bound ships. The probability of a ramming contact for the sub variant 5a growth is 16,24 times larger than for 3b growth. The reason for this, becomes clear when comparing the layout of 3a growth in Figure 6-2 with the layout of 5a growth Figure 6-3. The factor for a drifting contact by a route-bound ship is with 1,61 much lower, because a drifting contact is less sensitive on the distance.
Report No. 23601.621/4
27
Table 6-5 Probability of ramming and drifting contacts per year summarized over all wind turbines of the sub variant
base
x
x
1 growth
x
2a base
x
2a growth
x
x x
2b growth
x x
3a growth
x x x x x x x x x
5b growth
x x x
x x
x
6a growth
x x
5b base 6a base
x x
4b growth 5a growth
x x
4b base 5a base
x x
3b growth 4a growth
x x
3b base 4a base
x x
2b base 3a base
growth
from 10 nm
1 base
Variant
from 12 nm
Expected probability per year of a ramming and drifting contact against a wind turbine for 5775 MW new installed power (1155 wind turbines of 5 MW) Probability per year Sub variants compared with each other Sub variant ramming Drifting ramming drifting
x x x x
6b base
x
6b growth
x
x x
R-ships
N-ships
R-ships
N-ships
total
R-ships
N-ships
R-ships
N-ships
total
0,0701
0,0689
0,3847
0,0434
0,5671
10,57
1,25
1,29
1,02
1,38
0,0684
0,0673
0,3583
0,0428
0,5368
10,32
1,22
1,20
1,00
1,31
0,0627
0,0728
0,3852
0,0446
0,5654
9,46
1,32
1,29
1,04
1,38
0,0795
0,0748
0,3890
0,0468
0,5901
11,98
1,36
1,30
1,09
1,44
0,0634
0,0774
0,3856
0,0470
0,5734
9,55
1,40
1,29
1,10
1,40
0,0866
0,0839
0,4168
0,0543
0,6417
13,06
1,52
1,40
1,27
1,57
0,0087
0,0719
0,3532
0,0471
0,4808
1,31
1,30
1,18
1,10
1,17
0,0090
0,0555
0,3292
0,0459
0,4397
1,36
1,01
1,10
1,07
1,07
0,0084
0,0762
0,3428
0,0486
0,4760
1,27
1,38
1,15
1,14
1,16
0,0066
0,0551
0,2982
0,0498
0,4097
1,00
1,00
1,00
1,17
1,00
0,0557
0,0683
0,3886
0,0466
0,5591
8,40
1,24
1,30
1,09
1,36
0,0957
0,0633
0,4700
0,0540
0,6830
14,43
1,15
1,58
1,26
1,67
0,0564
0,0734
0,3604
0,0473
0,5375
8,51
1,33
1,21
1,11
1,31
0,0703
0,0651
0,4023
0,0587
0,5964
10,59
1,18
1,35
1,37
1,46
0,0597
0,0674
0,4162
0,0466
0,5899
9,01
1,22
1,40
1,09
1,44
0,1077
0,0666
0,4793
0,0552
0,7088
16,24
1,21
1,61
1,29
1,73
0,0595
0,0762
0,3907
0,0479
0,5743
8,96
1,38
1,31
1,12
1,40
0,0561
0,0664
0,4158
0,0591
0,5974
8,46
1,21
1,39
1,38
1,46
0,0296
0,0697
0,3873
0,0500
0,5366
4,46
1,26
1,30
1,17
1,31
0,0542
0,0721
0,3852
0,0556
0,5671
8,17
1,31
1,29
1,30
1,38
0,0285
0,0771
0,3756
0,0514
0,5326
4,30
1,40
1,26
1,20
1,30
0,0287
0,0677
0,3403
0,0581
0,4948
4,32
1,23
1,14
1,36
1,21
Report No. 23601.621/4
Figure 6-2
Sub variant 3b growth
Figure 6-3
Sub variant 5a growth
28
Report No. 23601.621/4
29
The total probability of a contact between a ship and a wind turbine is of the same order of magnitude for all wind farms. However, there are substantial differences in collision risk between wind farms belonging to the same variant. This is illustrated for Variant 6a growth in Figure 6-4. The wind farm areas are numbered. The number of turbines and the collision risks are presented as percentages of the total given in the row “Total absolute”. In total 864 wind turbines could be placed. For example wind farm 44 contains 5,3% of all wind turbines but delivers 43,8% of the total ramming risk for routebound ships. The percentage drifting risk is more in line with the percentage wind turbines. Wind farm location 44 is the reason why the ramming risk for Variant 6a growth is substantially higher than the ramming risk of the other variants of 6. The last row of the table in the figure contains the risk after applying a correction factor to bring the number of wind turbines to 1095 in these areas and adding the contact risk for 60 wind turbines (300MW) in Wadden North.
44 99 32 41 33
Ship ‐ wind turbine contacts for Variant 6a growth
Area
37
35 42
34
36 39 43
Borssele 17 Figure 6-4
ramming drifting Wind Total tur‐ R‐ships N‐ships R‐ships N‐schips share bines
17 21.6% 8.6% 20.7% 21.7% 22.0% 20.4% 32 1.0% 0.0% 6.2% 0.8% 1.7% 1.5% 33 7.3% 42.4% 6.5% 13.9% 5.2% 14.9% 5.9% 0.0% 4.4% 7.4% 5.2% 6.1% 34 35 2.8% 0.0% 0.6% 1.9% 2.8% 1.7% 2.5% 0.0% 0.2% 1.7% 2.4% 1.4% 36 37 2.1% 0.0% 1.8% 1.3% 2.9% 1.4% 39 3.1% 0.1% 6.2% 3.5% 3.8% 3.6% 41 17.0% 3.6% 19.2% 16.7% 19.7% 16.0% 42 20.8% 0.0% 13.2% 12.9% 21.8% 12.6% 43 2.9% 0.2% 6.2% 3.5% 3.6% 3.5% 44 5.3% 43.8% 7.8% 8.5% 4.7% 11.5% 99 7.5% 1.3% 7.1% 6.2% 4.2% 5.6% Total % 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Total absolute 864 Share 100.0% After factor + 1155 Wadden
Ship – wind turbine contacts for Variant 6a growth
0.0426 9.7%
0.0547 12.5%
0.2978 68.0%
0.0542
0.0721 0.3852
0.0429 0.4380 9.8% 100.0% 0.0556
0.5671
Report No. 23601.621/4
6.3
30
Total risk due to the wind farms
Chapter 6.1 contains the extra costs of accidents and detours outside the wind farms due to the changes in the traffic flows. Chapter 6.2 contains the probabilities of contacts with the wind turbines but not the costs involved with these contacts. In this chapter all costs, thus risk and additional sailing costs are quantified. For this purpose some assumptions have been made with respect to the damage of the wind turbines. These assumptions are extracted from the studies carried out for individual wind farms. The most important assumption is the probability that the wind turbine is lost by the contact. The following percentages are used: • 100% after a ramming contact by a route-bound ship; • 20% after a ramming contact by a non-route-bound ship; • 85% after a drifting contact by a route-bound ship; • 10% after a drifting contact by a non-route-bound ship; An amount of M€ 10 is taken for a wind turbine that is lost. An amount of M€ 1,04 is used for the average damage costs for the ship that has a contact with a wind turbine It is expected that the damage to the ship will be lower in general. Table 6-6 is composed with these assumptions. The left part contains the probabilities that a wind turbine is lost, thus the probabilities of Table 6-5 multiplied with the with the percentages above. The center part of the table contains the yearly costs of accidents with the wind farm and of accidents and extra miles outside the wind farm. By comparing the columns, it becomes clear that the largest contribution (more than 65%) of the yearly costs are delivered by costs of the extra miles. Thereafter the damage to the wind turbines gives the largest contribution.
Report No. 23601.621/4
Table 6-6
31
Total extra costs of risk and shipping by the wind farms Expected costs per year for 5775 MW new installed power (1155 wind turbines of 5 MW) Probability of a wind turbine is lost after a contact
base
x
x
1 growth
x
2a base
x
2a growth
x
2b base 3a base
x
3a growth
x x x
4a base
x
4a growth
x
4b growth
x
5a base
x
5a growth
x
5b growth
x
6a base
x
6a growth
x
6b growth
x
0,226
0,773
0,068
0,525
0,226
0,751
0,068
13,093
18,056
0,545
0,328
0,873
0,086
0,408
13,093
18,338
0,578
0,328
0,906
0,086
4,585
0,408
13,093
18,086
0,553
0,328
0,881
0,086
5,224
0,408
13,093
18,725
0,630
0,328
0,957
0,086
0,4028
0,590
4,028
4,618
0,326
9,208
0,0042
0,3833
0,558
3,833
4,391
0,326
9,208
0,0625
0,0136
0,3164
0,0043
0,3967
0,588
3,967
4,555
0,408
0,0793
0,0144
0,3241
0,0046
0,4223
0,614
4,223
4,837
0,0631
0,0145
0,3168
0,0045
0,3989
0,596
3,989
x
0,0865
0,0162
0,3477
0,0053
0,4557
0,667
4,557
0,0087 0,0090
0,0144
0,3002
0,0047
0,3280
0,500
3,280
3,780
0,604
15,773
20,157
0,461
0,514
0,975
0,154
x
0,0111
0,2798
0,0046
0,3046
0,457
3,046
3,503
0,604
15,773
19,880
0,428
0,514
0,941
0,154
0,0084
0,0152
0,2914
0,0049
0,3199
0,495
3,199
3,694
0,604
15,773
20,071
0,456
0,514
0,970
0,154
0,0066
0,0110
0,2535
0,0050
0,2761
0,426
2,761
3,187
0,604
15,773
19,564
0,398
0,514
0,911
0,154
0,0557
0,0137
0,3303
0,0047
0,4043
0,581
4,043
4,624
0,664
16,316
21,605
0,539
0,505
1,044
0,110
0,0957
0,0127
0,3995
0,0054
0,5133
0,710
5,133
5,843
0,664
16,316
22,823
0,671
0,505
1,176
0,110
0,0564
0,0147
0,3064
0,0047
0,3822
0,559
3,822
4,381
0,664
16,316
21,361
0,517
0,505
1,022
0,110
0,0703
0,0130
0,3420
0,0059
0,4311
0,620
4,311
4,932
0,664
16,316
21,912
0,584
0,505
1,089
0,110
0,0597
0,0135
0,3537
0,0047
0,4316
0,613
4,316
4,930
0,240
17,832
23,002
0.570
0.071
0.641
0.056
0,1077
0,0133
0,4074
0,0055
0,5339
0,737
5,339
6,077
0,240
17,832
24,149
0.697
0.071
0.768
0.056
0,0595
0,0152
0,3321
0,0048
0,4116
0,597
4,116
4,713
0,240
17,832
22,785
0.554
0.071
0.625
0.056
0,0561
0,0133
0,3534
0,0059
0,4287
0,621
4,287
4,909
0,240
17,832
22,981
0.585
0.071
0.656
0.056
0,0296
0,0139
0,3292
0,0050
0,3777
0,558
3,777
4,335
0,340
51,159
55,834
0,517
0,039
0,556
0,142
0,0542
0,0144
0,3274
0,0056
0,4016
0,590
4,016
4,606
0,340
51,159
56,105
0,555
0,039
0,594
0,142
0,0285
0,0154
0,3192
0,0051
0,3683
0,554
3,683
4,237
0,340
51,159
55,736
0,513
0,039
0,552
0,142
0,0287
0,0135
0,2893
0,0058
0,3373
0,515
3,373
3,888
0,340
51,159
55,387
0,483
0,039
0,522
0,142
x
x x
x x
x x x
x
0,547
13,926
0,0041
0,2980
x
6b base
14,153
0,3160
0,0129
x x
Total
0,0128
x
5b base
Outside wind farm
0,0683
x x
With wind farm
Extra accidents
0,0699
x
4b base
Wind farm
Only extra collisions per year as example
Nships
x
3b growth
Ship + consequences
Total extra costs [M€/year]
Rships
x
3b base
Total
Expected number of extra incidents per year
Nships
x
x
drifting
Costs for shipping outside the wind farms [M€/year]
Rships
x x
2b growth
growth
from 10 nm
ramming
1 base
Variant
from 12 nm
Sub variant
Costs by contacts with the wind turbines [M€/year]
x x
Total
Detour
Report No. 23601.621/4
7
32
RISK REDUCING MEASURES
The calculations are performed without any risk reducing measures. The risk reducing measure that would have the most impact is the involvement of an emergency towing vessel (ETV). Until shortly, “De Waker” had this function. As soon son as a report of a drifter was received by the Netherlands Coastguard and assistance was required, De Waker was alarmed. The probability to get a failure is higher during severe weather conditions and also the drifting speed will be larger. For this reason, De Waker was positioned at sea during wind conditions from 5 Beaufort, resulting in low response times. Due to a fire on board of “De Waker”, she is out of service. An ETV in the vicinity of a concentration of wind farms can reduce the probability of a drifting contact considerably, with more than 50%.
Report No. 23601.621/4
8
33
CONCLUSIONS AND RECOMMENDATIONS
Conclusions The conclusions with respect to the impact of the wind farms are: •
The changes in variant 5, outgoing ships via a new Maas NNW traffic lane and only ingoing ships through the Maas NW TSS has resulted in: o Reduction of the number of extra collisions form once in 6 years to once in 18 years; o An increase of the extra probability of a contact with an offshore platform from once in 100 year to once in 67 year, because Horizon, P15 and P18 are located to more intense used traffic flows and close to new traffic flows; o A reduction in the probability of a contact with a ship at anchor, because anchorage area 5A is better accessible without opposite traffic. Also the anchorage area 5A can be extended; o The detour costs are decreased from M€ 24,8 to M€ 17,8 per year, by the creation of the new Maas NNW lane.
•
The largest contribution (more than 65%) of the total yearly costs are delivered by costs of the extra miles. These costs are: Variant 1 M€ 9,2 Variant 2 M€ 13,1 Variant 3 M€ 15,8 Variant 4 M€ 16,3 Variant 5 M€ 17,8 Variant 6 M€ 51,2
•
Thereafter the damage to the wind turbines is the largest item. This varies from 2,8 M€ for the sub variant 3b growth to 5,3 M€ for sub variant 5a growth.
•
The drifting and in particular the ramming risk is depends strongly on the way of filling the search areas with wind turbines. The risk varies significantly among the sub variants within the same variant.
•
The lowest variation in risk occurs within variant 3 because the traffic flows in this variant are routed further away from the search areas than in the other variants.
Recommendations An Emergency Towing Vessel (ETV) to recover drifting ships can deliver a reduction in the probability of drifting of more than 50%. The present calculations have been carried out without an ETV. For safety, the traffic flows can be improved on some locations. This required a more detailed research. This aspect is dealt with in the FSA. Assumptions Within the calculations it is assumed that the drifting ship is not stopped by a contact with a wind turbine, but continues drifting. Thus one drifter can count for more than one drifting contact with wind turbines.
Report No. 23601.621/4
34
The extra probability of a contact with an offshore platform is calculated and included in the tables. However, the costs of the damage to the platform could not be included, because no data with respect to such a type of accident is available. Explanation for a better understanding of the probability of a collision The probability of a collision is determined from the calculated number of encounters between ships. The encounter is classified based on the course difference between the ships. A distinction is made between an overtaking (course difference is less than 60°), crossing (course difference between 60° and 150°) and a head on (course difference between 150 ° and 180 °) encounter. The probability of a collision given an encounter, is clearly different for the three situations. The probability of a collision, given a crossing encounter is larger than in case of an overtaking encounter. Sometimes this discrete classification of encounters is too rough for a sound result in a small area. The probability a collision is based on singular meetings. When a third ship is involved at the same time the probability of a collision will increase, because the number of encounters increases. For the triple ship encounter, in total six encounters between two ships are counted, thus three times as much as in a singular encounter. This means that the probability of a collision for a triple encounter is equal to the probability of a collision in three encounters between ships, thus as if the ships would have met each other in separate encounters. The probability of a collision given an encounter is not geographically related and therefore does not depend on the distance to an object such as a platform, or a wind turbine. Too few collisions have occurred under these circumstances to be able to include this in the modeling. The probability of a collision given an encounter depends on the type of ship, the ship's size and the weather conditions. The calculation of the number of encounters is based on the distribution of the speed of ships at sea. In some areas, when boarding or disembarking the pilot, some ships will sail at lower speeds. The difference in speed can result in more encounters. The results of the calculations can be improved when these areas are modeled in more detail. The number of meetings can then be determined more precisely. A new issue will be how the collision probability changes with the speed of the ships concerned. But in any case the differences between variants can be quantified more precisely.
Report No. 23601.621/4
35
REFERENCES [1]
C.F. Christensen (DNV), D. Brandon (DNV), C. van der Tak (MARIN), R. Zwart (Dirkzwager), A. Gerretsen (STC), C. Westra (ECN) Identification of Suitable Sea Areas for Wind Farms with Respect to Shipping and Safety DNV, Report No. 646092-REP-01 , December 2008
[2]
C. van der Tak Verkeerseffecten Windparken MARIN, Report 23601.620/4, 18 december 2009
[3]
C. van der Tak Quantitative risk assessment for offshore wind farms in the North Sea MARIN, Report 23601.621/2, March 16, 2010
[4]
J. Barentse Nadere toelichting: Gevolgen van aanvaringen van de windturbine-installatie Jaconbs Comprimo Nederland, juli 2000
[5]
Charterprijzen en vlootontwikkeling in de Amsterdamse haven Dynamar, 2006
[6]
M.H. Nijdam, L.M. van der Lugt, P.W. de Langen, Zeesluis IJmuiden, Economische gevolgen van stremmingen, Erasmus Universiteit, november 2006
[7]
Day costs for Cost-Benefit Analysis for MarNIS based on Drewry charter prices. Ecorys, acitivity within MarNIS project
[8]
Analysis of major claims Ten-year trends in Maritime risk UK P&I Club, 1998
[9]
M.P. Lehman, E. Sørgård Consequence Model For Ship Accidents Det Norske Veritas AS, Strategic research, Høvik, Norway, 2000
Report No. 23601.621/4
APPENDIX A
36
QUANTITATIVE RISK ASSESSMENT WITH SAMSON
This appendix describes the generic approach for the quantitative risk assessment with SAMSON.
1
SAMSON
The SAMSON1-model has been developed to predict the effects of spatial developments in the North Sea (or other sea areas). These developments can vary from development within the shipping industry itself (e.g. changes in ship sizes) to measures for the shipping (e.g. Traffic Separation Schemes). The following effects can be determined by the model: • Expected number of accidents per year, divided over type of accident, ship types and sizes involved in the accidents and objects; • Extra miles that have to be travelled as a result of a certain development and the costs involved; • Emitted environmental dangerous goods, e.g. exhaust; • Consequences of the accidents such as the outflow of oil or personal injuries. During the last 25 years the SAMSON model has been developed, extended, validated and improved continuously in studies performed for the Dutch Maritime Safety Authority and European projects. The system diagram of the SAMSON-model is shown in Figure 1-1. Almost all blocks presented in the diagram are available. The large block “Maritime traffic system” contains four sub-blocks. These sub-blocks describe the complete traffic pattern: the number of ship movements, the characteristics of the ships (length etc.) and the layout of the sea area. The different casualty models are used to determine the accident frequencies based on the complete traffic image. The block “Impacts” contains the sub-blocks that are used to determine the different consequences of the different accidents.
1
Safety Assessment Model for Shipping and Offshore on the North Sea
Report No. 23601.621/4
37
Maritime traffic system Traffic Demand - Cargo - Passengers - Fishing - Recreational vessels
Existing traffic management system
Traffic: - Traffic intensity - Traffic mix
Characteristics of sea areas
Traffic management measures: - Routing (TSS) - Waterway marking - Piloting - Vessel Traffic Services
Ships: - Technology on board - Quality of ships - Quality of crew
Traffic accidents: - Collisions - Contacts - Strandings - Founderings
Tactics (New traffic management measures)
Other accidents: - Fire and explosions - Spontaneous hull accidents - Cargo accidents
Pipe accidents: - Foundering on pipe - Cargo on pipe - Anchor on pipe - Anchor hooks pipe - Stranding on pipe
Impacts Financial costs: - Investment costs - Operating costs
Environmental consequences: - Oil spills - Amount of oil on coast - Chemical spills - Dead and affected organisms
Economic consequences: - Loss of income - Repair costs - Cleaning costs - Delay costs caused by accidents - Extra sea miles caused by the use of a tactic Human safety: - Individual risk - Societal risk
Search and Rescue Contingency planning
Figure 1-1
1.1
System diagram SAMSON-model
Effect of a wind farm
The construction of a wind farm will have some consequences for the shipping near the location of the wind farm. In Article 60 of UNCLOS is formulated that the coastal State may, where necessary, establish reasonable safety zones around artificial islands, installations and structures with a maximum of 500m. By applying this safety zone of 500m around a wind turbine, the wind farm is an area to be avoided. For so far as known all States apply this safety zone of 500m for route-bound traffic, which means that it is prohibited for merchant ships to sail through a wind farm. Therefore, it is possible that ships may have to change their sailing routes in the future and pass the wind farm at a minimum of 500 m (equal to the safety zone around an offshore platform). This means that the wind farm will cause a nuisance to the passing ships. The policy with respect to small ships varies per country. Some countries allow smaller ships to sail through or fish within the area of the wind farm. The rerouting of ships can also have some effects outside the location of the wind farm. Because ships are forced to sail a different route the traffic density will increase on the other traffic routes outside the wind farm. Because of the increase in the traffic intensity
Report No. 23601.621/4
38
on some routes, the number of encounters between ships and therefore the number of collisions will also increase. The construction of the wind farm will also introduce a new type of accident. Due to different causes, a ship can collide with a wind turbine. In the SAMSON model two types of contacts are distinguished, a ramming and a drifting contact. •
•
A ramming contact takes place when a ship is on a collision course with a wind turbine and a navigational error occurs. A navigational error can have various causes, like lack of information, not being able to see the wind farm, not being present on the bridge, getting unwell and not being able to act, making an error etc. A ramming contact will take place with high speed: 90% of the service speed of a vessel. A drifting contact occurs when a ship in the vicinity of a wind turbine experiences a failure in the propulsion engine or in the steering equipment. Since the ship slowly becomes uncontrollable as it loses speed, the combined effect of wind, waves and current may carry the ship towards the wind turbine. If dropping an anchor does not help or is not practical and the repair time exceeds the available time, the ship may collide against a turbine. This generally happens at a low speed.
All these accidents apply to all shipping near the location of the wind farm, and do not necessarily only apply to the ships that are usually sail across the location of the wind farm. In order to be able to calculate the effects of the wind farm, it is necessary to make a “new” traffic database for the SAMSON-model. All passing ships have to keep a distance of at least 500m from the border of the wind farm. The number of changes in the traffic pattern by the wind farm depends on the location and the size of the wind farm. Subsequently, the different models within SAMSON are used together with the new traffic database to carry out a complete (nautical) risk assessment. This risk assessment implies the “change” of risk for the complete shipping near the wind farm and the “new” risk being the probability of colliding against a wind turbine.
1.2
Model input and assumptions
The following model inputs, assumptions and parameters are used in the calculation. 1.2.1 Traffic In the calculation, a traffic database is used. The traffic database contains links, traffic intensities on the different links and the link characteristics. A link is defined as a straight connection between two so-called waypoints. The traffic intensity of a link describes the number of ships sailing on the link per year divided over different ship types and ship sizes. The link characteristic of the connection contains its width and the lateral distribution of the ships across the join. The maritime traffic is divided into two main groups: the route-bound ships and the non-route-bound (or random) ships. The route-bound traffic consists of merchant vessels and ferries sailing along the shortest route from one port to another. The non-
Report No. 23601.621/4
39
route-bound traffic contains vessels that mainly have a mission at sea, such as fishing vessels, supply vessels, working vessels and pleasure crafts. Within the SAMSON-model both groups are modelled in a different way. The route-bound traffic is modelled on the shipping routes in the region of the wind farm. Because of the location of the different ports and the traffic separation schemes in the area, most of the route-bound ships sail on a large network of links, comparable with the road network. Every ship is allowed to sail “anywhere” as long as one complies with the rules and regulations. Thus ships could also sail outside the defined network. However, the number of ships sailing outside the network is very small, because the network contains the shortest route between ports. A new traffic database is generated for the location of the wind farm. The location of the wind farm is assumed to be a forbidden area, so route-bound traffic will not sail through the wind farm. This new database is used to determine the probabilities for contact with a wind turbine. The non-route-bound traffic cannot be modelled the same way as the route-bound traffic, because the information about the journeys is not included in the Lloyds database. Moreover, the behaviour of these non-route-bound ships at sea is very different. A non-route-bound ship does not sail from port A to port B along a clear route, but from port A to one or more destinations at sea and then usually back to the port of departure A. The behaviour of these ships at sea is mostly unpredictable. Fishing vessels also usually sail from one fishing ground to another during one journey. Therefore, the traffic image of the non-route-bound traffic is modelled by densities of ships in a so-called grid cell of 8 x 8 km. The size of this cell is historically connected to the grid cell size used in the North Sea. Other sizes of the grid cells can be applied but it has only sense when the densities are known on a finer grid. The local policy for making use of the region of the wind farm is modelled. 1.2.2 Used models The total SAMSON-model consists of many different sub-models for the different accidents. Not all sub-models have to be used to determine the effects of the wind farm. The following models are applied to determine the expected number of rammings and driftings with a wind turbine per year: • Contact with a fixed object (wind turbine) - as a result of a navigational error (ramming) - as a result of an engine failure (drifting) The drift model contains a table with the probability of failure to anchor for all wind force classes. The failure probability increases strongly with the wind force. Further, the self repairing time function is modelled to determine the maximum time of the drifting process without assistance from outside. The effect of one or more ETVs (Emergency Towing Vessels), thus vessels that can get control over the drifting ships by taking her in tow, can be determined by the input of one or more ETVs. For each ETV the geographical position and a number of characteristics, such as speed and towing capacity are required. No ETVs are modelled in this study. In case the wind farm has a large impact on the safety of shipping outside the location of the wind farm, the risk levels with and without the wind farm can be compared.
Report No. 23601.621/4
40
To determine the “general” risk level the following models are used: • Ship-Ship collision model; • Contact with a platform - as a result of a navigational error (ramming); - as a result of an engine failure (drifting); • Contact with a pier - as a result of a navigational error (ramming); - as a result of an engine failure (drifting); • Stranding - as a result of a navigational error (ramming); - as a result of an engine failure (drifting);
1.3
Consequences
A collision (ramming or drifting) of a ship with a wind turbine can lead to (serious) damage to the wind turbine, to the ship, to the environment as a result of an outflow of oil from the ship, or personal damages (injuries, fatalities). The risk assessment results in drifting and ramming collision frequencies for the full range of all thirty-six ship types and eight ship size classes. This subdivision makes it possible to perform additional calculations. Based on the mean displacement of each colliding vessel, added mass and the expected speed at collision (90% of the service speed for ramming and the drifting speed for drifting), the kinetic energy of the ship at the moment of the collision is calculated. This energy level is used to calculate the damage to the ship based partially on experience and partially on complex calculations. The calculations are based on the assumption that all energy is being absorbed. 1.3.1 Damage to the wind turbine Due to the limited energy absorption of the collided object (wind turbine), not all the kinetic energy of the colliding ship will be absorbed. The collapse behaviour of the wind turbine has been studied ([4] of main report). The conclusion of the study is that for almost all ship types the wind turbine collapses (static) and that only a fraction of the energy of the ship will be absorbed. For further analysis of the damage to the wind turbine the following two collapse types are being distinguished: •
Pile failure; the wind turbine collapses by bending at the point of impact caused by plastic deformation (first picture in Figure 1-2). The collapsed part of the wind turbine will stay attached to the rest of the turbine. Finally the turbine falls towards the ship or away from the ship. When the turbine collapses towards the ship the rotor and nacelle can fall on the deck of the ship.
•
Soil failure; the wind turbine collapses by bending at the foundation of the turbine at the bottom of the sea caused by plastic deformation (second and third picture from Figure 1-2).The turbine can, as a result of this deformation, fracture at the bottom of the sea or it can also be pushed over (including the bottom of the sea).
Report No. 23601.621/4
Figure 1-2
41
Figures of the different collapse types.
The collapse type that will occur after a collision can only be determined by dynamic calculations. Some experts have estimated the frequency of occurrence of the different collapse types based on their research. When it was not possible (yet) to estimate the frequency, one has chosen a conservative result. For example, the pile with the turbine can fall towards the ship or away from the ship depending on the construction and the environmental factors. For the calculations, it is assumed that the turbine will fall on the ship in all cases of a pile failure. In Table 1-1 an overview is given of the different collapse types as a result of a ramming or drifting collision per ship size class. In addition, the expected damage to the ships is given in the table. In case of a frontal or frontal/lateral (grazing) collision (ramming) of the turbine there will be (serious) damage to the bow of the ship, but no (serious) damage to the side of the ship, where the cargo tanks are located. The construction of the ship in front of the collision bulkhead is very rigid, which causes the damage to be limited to the front of the ship. Thus, it will not cause cargo (oil) or fuel oil to flow out of the ship. In case of grazing the rigid construction with the bow, the ship will absorb the kinetic energy without causing much damage. There could be some damage to the ship because the pile and the nacelle fall on the ship. No environmental damage is expected in case of a ramming collision, because the pile is constructed in such a way that no parts stick out which could penetrate the hull of the ship and cause the outflow of oil and/or chemicals. Environmental damage can be expected in case of a drifting collision, where the hull of the ship can be penetrated by some thickenings of the pile. This can cause the outflow of oil and/or chemicals. Personal damages (injuries/fatalities) are only expected when the pile and/or a part of the turbine collapses on the ship. When a wind turbine collapses it can be expected that some oil from the turbine will flow out. The pollution that results would be, at most, 250 litres of mineral oil, which is with respect to the viscosity and evaporation comparable to the cargo oil in the SAMSONmodel, and 100 litre diesel oil (comparable to bunker oil in the SAMSON-model).
Report No. 23601.621/4
Table 1-1
Collapse type
Pile failure
Soil failure
42
Collapse types with the estimated percentages and the estimation of the resulting damage to the turbine and the ship.
ship size [GT] Part <500 500-1000
0% 0%
1000-1600
5%
1600-10000
10%
10000-30000
10%
30000-60000
10%
60000-100000
10%
>100000
10%
<500 500-1000 1000-1600 1600-10000 10000-30000 30000-60000 60000-100000 >100000
100% 100% 95% 90% 90% 90% 90% 90%
Ramming Frontal Grazing (10%) (90%) Damage Damage Part TurTurShip Ship bine bine No None 0% No None Yes None 0% No None Nos Deck 0% Yes None 2 Pos Nos Nos Deck 5% Deck Pos Pos Nos Nos Deck 10% Deck Pos Pos Nos Nos Deck 10% Deck Pos Pos Nos Nos Deck 10% Deck Pos Pos Nos Nos Deck 10% Deck Pos Pos No None 100% No None Yes None 100% No None Yes None 100% Yes None Yes None 95% Yes None Yes None 90% Yes None Yes None 90% Yes None Yes None 90% Yes None Yes None 91% Yes None
Drifting Lateral mid-ships Lateral eccentric (100%) (0%) Damage Damage Part Part TurTurShip bine bine
100% 100% 100% 100% 100% 100% 100% 100%
No No No Yes Yes Yes Yes Yes
None None Hull Hull Hull Hull Hull Hull
100% No None 100% No None 100% No None 100% No None 100% Yes None 100% Yes None 100% Yes None 100% Yes None
1.3.2 Personal damage Starting from the estimated frequencies for ramming and drifting against a wind turbine, the following steps are made per ship type and ship size: • •
•
•
2
The number of estimated collisions (ramming/drifting) is multiplied with the matching probability of a certain collapse type. Multiplication with the probability that the pile/turbine will fall upon the ship. Since it is uncertain whether the pile will fall towards the ship of away from the ship, the probability is set to 1. Thus, the worst-case scenario is addressed. Multiplication with the part of the deck that is damaged. There are two worst-case assumptions for this multiplication factor: - The pile falls completely on the deck. In case of grazing it is very well possible that the pile does not fall on the deck completely. - The surface of the pile and the rotor are being applied completely. In this case it is assumed that the rotor is falling while it is still turning. Multiplication with the probability that a person is located on the deck when the turbine falls onto it. The probability that someone is located on the deck is assumed to be 10% In reality this probability is (much) lower, because only on fishing vessels will the crew be mostly on deck, but for this type of vessel the turbine will not bend.
NosPos = Nacelle On Ship and Pile On Ship after plastic deformation
Report No. 23601.621/4
•
43
This 10% also includes the probability that someone inside the ship is hit by a part of the falling turbine. Multiplication with the average number of people on board.
Injuries or fatalities caused by the impact itself are not included. In addition, the personal damages in case of a very small vessel that is totally destroyed by the impact are not taken into account. For this category of ships, the probability models are very unreliable. Besides, these smaller ships will usually just graze the pile and not hit it frontally.
Report No. 23601.621/4
APPENDIX B
ALL VARIANTS WITH TRAFFIC DATABASE
THE
44
CORRESPONDING
Report No. 23601.621/4
45
Report No. 23601.621/4
46
Report No. 23601.621/4
47
Report No. 23601.621/4
48
Report No. 23601.621/4
49
Report No. 23601.621/4
50
Report No. 23601.621/4
51
Report No. 23601.621/4
52
Report No. 23601.621/4
53
Report No. 23601.621/4
54
Report No. 23601.621/4
55
Report No. 23601.621/4
56
Report No. 23601.621/4
57
Report No. 23601.621/4
58
Report No. 23601.621/4
59
Report No. 23601.621/4
60
Report No. 23601.621/4
61
Report No. 23601.621/4
62
Report No. 23601.621/4
63
Report No. 23601.621/4
64
Report No. 23601.621/4
65
Report No. 23601.621/4
66
