Compendium

Walkthrough: Design an Offshore Wind Farm (1 GW Fixed-Bottom)

21 min read

Walkthrough: Design an Offshore Wind Farm (1 GW Fixed-Bottom)

This walkthrough takes a 1 GW fixed-bottom offshore wind farm from site characterization through turbine and foundation selection, electrical export, marine installation, O&M strategy, and full-life economics. The reference case: 67 × 15-MW direct-drive turbines, 30-50 m water depth, 30-80 km offshore, 25-year design life with 5-year extension option.

Anchored against actual programs: Hornsea 1/2/3/4/5 (Ørsted, North Sea UK), Dogger Bank A/B/C (SSE-Equinor-Vårgrønn, 3.6 GW), Vineyard Wind 1 (Avangrid-CIP, US MA 800 MW first US commercial-scale, first power Jan 2024), Revolution Wind (Ørsted-Eversource, RI 704 MW), South Fork Wind (Ørsted-Eversource, NY 132 MW first US fully commercial completed Mar 2024), Coastal Virginia Offshore Wind (Dominion 2.6 GW), Empire Wind 1+2 (Equinor 2 GW, BP exited 2023), Borssele (NL), He Dreiht (Germany, EnBW 960 MW Vestas V236-15), Yunlin (Taiwan 640 MW), Hesheng (Japan).

1. Site characterization

Wind resource

A bankable wind resource assessment (WRA) is the cornerstone of project finance. Required minimum mean wind speed at hub height (140-160 m for 15-MW class): 8.5-10.5 m/s. Below 8 m/s mean, P50 economics typically fail at current $60-100/MWh CfD/PPA bands.

Tools and inputs:

  1. Long-term reanalysis — ERA5 (ECMWF, hourly 0.25°), MERRA-2 (NASA, hourly 0.5°), CFSR (NCEP), CFSv2. Provides ~30+ year long-term climate baseline.
  2. Mesoscale modeling — WRF (Weather Research & Forecasting) at 1-3 km resolution, run by DNV, UL, K2 Management, Wood Mackenzie / formerly Garrad Hassan, Vortex, ProPlanEn.
  3. Microscale CFD — Meteodyn WT, WAsP CFD, OpenFOAM-based; resolves wakes and complex flow.
  4. On-site measurement — floating LiDAR buoys: Fugro SEAWATCH Wind LiDAR Buoy, Akrocean Wind Sentinel, EOLOS FLS200, AXYS WindSentinel, with Vaisala/ZX Lidars/Leosphere WindCube V2 onboard. 12-24+ month campaigns, 2-4 buoys for a 1 GW site.
  5. Hub-height met masts — increasingly displaced by LiDAR but still installed at flagship sites (e.g., Hornsea, Borssele) at ~150 m for IEC 61400-12-1 power-performance verification.

P50/P75/P90 energy yield estimates with uncertainty propagation (long-term ref, vertical extrapolation, wake models, blockage, availability) feed financial models. Independent engineer (IE) review by DNV, Wood Mackenzie, K2 Management, Ramboll prior to financial close.

Wake losses

Internal wake losses on a 1 GW dense array typically 8-15%; blockage (the “global blockage” upstream slowdown across the array) 1-4% additional. Far-wake interactions with neighboring farms now material in the North Sea (Hornsea wake reaches Dogger Bank).

Seabed and geotechnical

Survey campaign by Fugro, Gardline, MMT, Geoquip Marine, Ocean Infinity:

  • Bathymetry — multibeam echosounder (Kongsberg EM 2040 / EM 304, R2Sonic 2024, Reson SeaBat T50-P) at 0.5-1 m resolution
  • Geophysical — sub-bottom profiler (Edgetech 3400, Innomar parametric SES-2000), side-scan sonar (Edgetech 4205, Klein 5900) to image seabed features, boulders, sand waves, paleo-channels, gas pockets
  • Geotechnical — CPTu (cone penetration test with pore pressure, Geomil CFP-50 / Fugro SeaCPT) to 30-50 m sub-seabed; boreholes by jack-up or geotechnical vessel (Fugro Voyager, Geoquip Saentis) sampling to 80-120 m
  • UXO survey — magnetometer arrays (Geometrics G-882, JW Fishers Pulse 12) and AUV-mounted; North Sea hot zone post-WW2; survey + clearance via controlled detonation or relocation, usually $20-100M line item

Met-ocean

  • Wave hindcast (DHI MIKE 21 SW, WaveWatch III) + on-site wave buoys (Datawell Waverider DWR-MkIII)
  • Current profiles (ADCP — Teledyne Workhorse Sentinel, Nortek Signature500)
  • Storm-surge statistics for foundation extreme-load design
  • Marine growth thickness (typical 100 mm North Sea, structural drag impact)
  • Ice (Baltic, Bohai Bay China — adds ice load case per IEC 61400-3)

Environmental + permitting

  • Avian baseline — boat-based and aerial visual + radar (DHI Wind, HiDef Aerial Surveying)
  • Marine mammal baseline — passive acoustic monitoring (PAM, Wildlife Acoustics SongMeter SM4, Chelonia C-POD) + visual MMOs
  • Benthic ecology — grab samples, drop video, ROV
  • Fisheries impact + commercial routing
  • AIS shipping density analysis

Permitting timelines:

  • US BOEM — Bureau of Ocean Energy Management lease auction → site assessment plan → COP (Construction & Operations Plan) → ROD (Record of Decision) → 7-10 years lease-to-COD
  • UK Crown Estate — leasing round (Round 4 closed 2022, ScotWind 2022 25 GW awarded) + Section 36 consent + DCO (Development Consent Order)
  • DE — BSH (Bundesamt für Seeschifffahrt und Hydrographie) tendering + permitting
  • NL — RVO competitive tender + WBR permit
  • Denmark — Energistyrelsen open-door + tender
  • France — competitive auction CRE

2. Turbine selection

15-MW class options (2024-2026 commercial)

OEMModelRotor (m)Power rating (MW)DrivetrainNotable deployments
Siemens GamesaSG 14-222 DD22214 (15 boost)Direct-drive PMSGHe Dreiht, Sofia, Moray West, Vineyard Wind 1 (13 MW)
Siemens GamesaSG 14-236 DD23614Direct-drive PMSGEmpire Wind 1, Coastal Virginia, Revolution Wind
VestasV236-15.0 MW23615Medium-speed hybrid PMG + 2-stage gearboxHe Dreiht, Baltic Eagle (recently switched some sites)
GE VernovaHaliade-X 14 MW22014 (12-14 variants)Direct-drive PMGDogger Bank A/B/C, Vineyard Wind 1 (13 MW variant)
GE VernovaHaliade-X 15.5 MW22015.5 (boost from 14)Direct-driveDogger Bank C (uprated)
MingyangMySE 18.X-20MW26018 (-20 variant)Hybrid drivetrainChinese domestic; offered to overseas markets 2024+
MingyangMySE 16.X-26026016.6Hybrid PMSGFujian, Shandong; targeting EU
CSSC HaizhuangH260-18MW26018Direct-drivePingtan Strait
GoldwindGWH252-16MW25216Hybrid PMSGFujian
Dongfang Electric18MW26018Hybrid PMSGFujian 2023

For this reference design: Siemens Gamesa SG 14-236 DD at 14.9 MW boost (67 units × 14.9 MW = 998 MW). Rationale: mature production from SGRE Cuxhaven (DE) and Le Havre (FR), direct-drive eliminates gearbox failure mode (the historical #1 unplanned outage source on offshore turbines), 236-m rotor produces high specific yield in low-medium wind regimes (8-9 m/s mean), Ørsted/RWE/Vattenfall purchase history confirms financial-grade reliability data.

Direct-drive vs geared drivetrain

  • Direct-drive PMSG (Siemens Gamesa, GE Haliade-X, Mingyang MySE DD): generator at hub speed (~5-8 rpm), large-diameter ring stator/rotor; eliminates gearbox; ~30% lower component count; permanent-magnet rare-earth (NdFeB + Dy) requirement ~600-1,000 kg per turbine at 15 MW.
  • Medium-speed hybrid (Vestas V236 with 2-stage gearbox + PMG): partial speed-up (gearbox 1:30-1:50), smaller generator, lower nacelle mass than DD but retains some gearbox risk.
  • Geared DFIG (legacy onshore architecture, doubly-fed induction generator): largely abandoned for new offshore platforms post-2018.

Permanent magnet supply: NdFeB magnets ~95% sourced from China (Baotou Steel + Inner Mongolia, BGRIMM, Ningbo Yunsheng); diversification efforts (Vacuumschmelze DE, Less Common Metals UK, MP Materials US, Lynas AU). This is a meaningful Western supply-chain vulnerability for the DD architecture.

Blade

15-MW class blades are 115-118 m long (LM Wind Power 107P 107 m for Haliade-X; LM 115.5P for newer 15 MW boosts; SGRE B115 for SG 14-236). Construction:

  • Carbon-fiber spar caps (Toray T700/T800, Hexcel HiMax, Mitsubishi Rayon) — pultruded UD planks bonded into the laminate, ~30-40% of cap weight, ~85% of stiffness
  • Glass-fiber UD plies (Vetrotex/3B, Owens Corning, Jushi) for shell laminate
  • Epoxy resin (Hexion EPIKOTE, Olin Epoxy, Westlake Polymer, Swancor 2511) infused via VARTM (vacuum-assisted resin transfer molding)
  • Balsa or PVC/PET foam core in shell sandwich
  • Lightning-protection system (LPS) — receptors at multiple points + down-conductor to root
  • Edge bonding via PU/MMA paste (Henkel Loctite, Sika SikaPower)

Recyclability: ZEBRA project (LM Wind / Arkema Elium recyclable thermoplastic resin), Siemens Gamesa RecyclableBlade (commercialized 2022, deployed Kaskasi DE 2023). Standard thermoset blades end up landfilled or co-processed in cement kilns (Veolia, Geocycle Lafarge).

Tower

Tubular steel, S355 + S420 + S460ML higher-strength grades. 4-9 m diameter tapered sections, total 100-130 m hub-height-above-water (~140-160 m hub above LAT incl transition piece). 4-6 sections, bolted L-flange (M64-M80 high-strength studs). Tower mass 800-1,200 tonnes. Fabricators: Haizea Wind (ES Bilbao), CS Wind (KR, Vietnam, Türkiye, US Pueblo CO post-IRA), Welcon (DK), GRI Renewable (ES), SGRE Cuxhaven (offshore facility), Bladt Industries (DK), Steelwind Nordenham (DE).

3. Foundations

Monopile (>80% of fixed-bottom offshore installs)

  • Diameter 8-12 m (XL/XXL monopile)
  • Length 80-110 m (typical penetration 30-40 m, plus exposed transition + water column)
  • Wall thickness 90-150 mm
  • Mass 1,500-2,500 tonnes
  • Steel S355ML / S420ML / S460ML
  • Fabricators: Sif Group (NL Maasvlakte 2), Steelwind Nordenham (DE), EEW SPC (DE Rostock), Haizea (ES — newest Bilbao yard 2024), Bladt Industries (DK), SeAH Wind (KR/UK Teesside Phase 1 2024, largest single yard globally)

Pile-driving: impact hammer Menck MHU 3500S (3,500 kJ blow energy) / MHU 4400S / MHU 5300S (5,300 kJ, capable of 12 m monopile). Hammer counts 4-6 globally; bottleneck for large campaigns.

Marine-mammal-protection during piling:

  • Big Bubble Curtain (BBC, Hydrotechnik Lübeck) — perforated air ring around pile; 10-15 dB sound reduction
  • Hydro Sound Damper (HSD, OffNoise Solutions) — net with foam bodies; 8-15 dB additional
  • IHC NMS Noise Mitigation Screen — double-wall steel
  • AdBm Helmholtz resonators
  • Soft-start procedure (low-energy ramp) before full blow energy
  • Marine mammal observers (MMOs) + PAM monitoring with shutdown protocol

Transition piece (TP): traditionally bolted-grout connection to monopile (the historic 2009-2013 grouted-connection slippage epidemic forced wholesale retrofit campaigns — Sif/SLP cone-grouted geometry, bolted-flange now industry default since ~2015). TP carries boat-landing, J-tubes for cable, internal platform, davit crane.

Jacket

Four-legged steel-lattice (~50-80 m tall, 600-1,500 tonnes), pin-piled at corners. Better for:

  • Deeper water (40-60 m) where monopile diameter exceeds installation hammer capability
  • Sandy/silty soils with low lateral resistance
  • Sites where transport / installation of XL monopiles infeasible
  • Offshore substations (almost always jacket)

Fabricators: Lamprell (UAE/UK), Bladt Industries (DK), HSG Sungdong (KR), Smulders (BE/NL), Heerema Fabrication, Dragados Offshore (ES), Saipem.

Suction bucket

Suction-installed inverted bucket — embedded by underpressure rather than driving. Deployed at:

  • DanTysk + Sandbank (DE)
  • Borkum Riffgrund 1 (Ørsted demo)
  • Aberdeen Bay (12 × Vattenfall European Offshore Wind Deployment Centre)
  • Hornsea Project Two (test articles)
  • Hywind Tampen (floating, suction anchors)

Advantages: silent installation (no piling — major permit driver in fish-spawning zones), reversible, suitable for soft cohesive soils. Disadvantages: slower installation, more design uncertainty, limited supplier base (Universal Foundation, SPT Offshore).

Floating (out of scope, but for context)

For water depths >60-80 m (Pacific US, much of Atlantic Europe, Japan, Korea, South China Sea):

  • Semi-submersible — Principle Power WindFloat (Atlantic, Kincardine, WindFloat Pacific PG&E), Equinor Hywind Tampen (88 MW concrete spar, 2023 world’s largest floating farm)
  • Spar — Hywind Scotland (5 × 6 MW, 2017 world’s first commercial floating)
  • Tension-leg platform (TLP) — concept stage (X1 Wind, Bluewater)
  • TetraSpar — Stiesdal-Shell-RWE demo Norway 2021
  • Damping pool — Ideol Floatgen (FR demonstrator)
  • Mooring: catenary chain + clump weight, or taut-leg synthetic (polyester) — Bridon-Bekaert, Lankhorst, DSM Dyneema

4. Electrical export

Inter-array cable

  • Voltage: 66 kV (became standard ~2017, displacing 33 kV; 66 kV roughly doubles MVA per circuit and tolerates 4 turbines per string at 15 MW vs 2 at 33 kV)
  • Construction: 3-core XLPE-insulated, lead-sheathed, steel-wire-armored submarine cable; copper or aluminum conductor 95-630 mm²
  • Length: 1.6-3.5 km hop between turbines (longer in spread-out array); typically 6-10 turbines per string
  • Suppliers: Prysmian (IT), Nexans (FR), NKT (DK), Sumitomo Electric (JP), Hellenic Cables (GR), ZTT (CN), LS Cable (KR), Furukawa (JP), TFKable (PL — acquired by GUSE/HASCO 2023)
  • Layout: ring-string or radial-string topology; ring offers single-fault resilience at ~10-15% extra cable cost; most modern farms radial

Offshore substation (OSS) — HVAC

For a 1 GW farm at <80 km offshore, HVAC is economical. Configuration:

  • 66/220 kV step-up transformers (2-3 units of 400-600 MVA, ABB Hitachi Energy, Siemens Energy)
  • 220-275 kV GIS or AIS switchgear
  • Static var compensator (SVC) or STATCOM for reactive power
  • Shunt reactors for cable capacitance compensation
  • Auxiliary diesel + battery + harmonic filters

OSS platform: jacket foundation (12-16 piles, 4,000-8,000 tonne topside). Single OSS for 1 GW or two split (more common for resilience). EPC: Petrofac, Saipem, Aibel, Heerema, Bladt, McDermott. Topside mass 6,000-15,000 tonnes for 1 GW HVAC platform.

Offshore substation — HVDC (for distances >80 km or 2+ GW)

For longer-distance / larger-scale, HVDC export displaces HVAC due to cable charging-current losses:

  • Converter platform: voltage-source converter (VSC) topology, ABB Hitachi Energy HVDC Light, Siemens Energy HVDC Plus, GE Grid Solutions
  • Voltage levels: ±320 kV (BorWin1, DolWin1), ±525 kV (DolWin5, IJmuiden Ver, BorWin5, SCR), ±640 kV (rising)
  • Notable HVDC offshore projects:
    • DolWin1 800 MW ±320 kV (DE, COD 2015)
    • BorWin3 900 MW ±320 kV (DE)
    • DolWin5 900 MW ±320 kV (DE, COD 2024)
    • Dogger Bank A/B/C 3× 1.2 GW ±320 kV (UK, in commissioning 2024-2026)
    • Hornsea 3 2.85 GW ±320 kV (UK, in construction)
    • IJmuiden Ver Alpha + Beta 2 GW each ±525 kV (NL, COD 2029)
    • Empire Wind 1 (HVAC 230 kV) vs Empire Wind 2 (suspended)

Onshore converter: similar topology mirrored; substation connects to AC transmission (typically 400 kV pan-European, 345 kV US).

Export cable

  • HVAC: 220-275 kV 3-core XLPE submarine, 800-1,500 mm² Cu, single circuit per cable; for 1 GW typically 2 circuits parallel
  • HVDC: ±320-525 kV single-core MI (mass-impregnated) or XLPE-DC, +pole and -pole separately, 1,500-2,500 mm² Cu
  • Burial: 1-3 m below seabed via jet trenching (post-lay), plough (Sif Group, Boskalis, Subsea 7), or mechanical trencher (Soil Machine Dynamics)
  • Landfall: HDD (horizontal directional drilling) under beach to onshore TJB transition joint bay; +Joint Box transition to onshore underground cable
  • Length: 30-80 km typical for the reference site; ~150-200 km for far-out Borssele or Dogger Bank

5. Installation

Vessels

Heavy-lift jack-up installation vessels (WTIVs) for 15-MW turbines:

VesselOwnerYearCrane (t)Leg length (m)Deck (m²)Notes
Cadeler O-class (Wind Osprey, Orca)Cadeler (DK, IPO 2020)2012/20241,2001095,500Upgraded for 14-15 MW
Cadeler X-class (Wind Peak, Pace)Cadeler2025-20262,6001305,600NG-20000X, for 20 MW
Cadeler A-class (acquired Eneti 2023)Cadeler2025-20262,6001265,5007 × 15 MW per load
Van Oord AeolusVan Oord (NL)20141,6001054,000Upgraded 2021
Boskalis Bokalift 2Boskalis (NL)20204,000n/a (floater)8,000Crane vessel, not jack-up
GustoMSC NG-14000XL designsvarious2018+1,500110-1304,500Used by Seajacks Scylla
Heerema SleipnirHeerema (NL)20192 × 10,000n/a12,000Largest crane vessel globally; jacket installation
Saipem 7000 / 8000Saipem1988/202514,000 / 22,000n/a19,000Heavy-lift floating
Subsea 7 / Seaway 7 Stanislav Yudin (acquired)Seaway 7 (NO)1985/20232,50080 (rebuilt)5,000Foundation + turbine
Maersk Supply Service Maersk VoyagerMaersk20242,200 (planned)1305,000Newbuild WTIV
Dominion Energy CharybdisDominion (US)2024 (delayed to 2025)2,2001305,000+First US Jones-Act compliant WTIV
Eneti NG-16000XCadeler (acq)20242,6001266,000
Innovation, Pacific OrcaDEME2012/20131,500/1,200105/973,400Pile-driving spreads

Charter rates: $200-500k/day for top-tier WTIVs (2024 market peak); cost can dominate balance-of-plant economics. Fleet shortage of large-crane WTIVs is the binding constraint on 15+ MW-class deployment globally through 2027.

Cable-lay vessels

  • NKT Victoria — NKT (DK), 2017, 9,000-tonne cable carousel
  • Prysmian Leonardo da Vinci — Prysmian (IT), 2021, 17,000-tonne capacity
  • Boskalis Ndurance — Boskalis (NL), 2014
  • Sumitomo Electric Subaru — JP, 2018
  • Nexans Aurora — Nexans (FR), 2021, 10,000-tonne carousel
  • Asso.subsea Asso 1
  • Helix Q4000 (US, Charybdis-paired)

Installation logistics

For 1 GW at 67 turbines × ~10 days per turbine (single WTIV, weather-dependent):

  • Mobilization + commissioning vessel: 30 days
  • Foundation installation campaign (monopile + TP): 6-9 months (3-5 days per monopile + TP)
  • Array cable lay + burial: 4-6 months overlapping
  • OSS topside + jacket: 2 months
  • Export cable lay + landfall: 4-6 months
  • Turbine installation (tower + nacelle + 3 blades): 5-8 months at 3-5 turbines per week per WTIV
  • Pre-commissioning + commissioning: 4-6 months
  • Total installation campaign: 24-36 months from start to COD

Marshalling ports: Esbjerg (DK), Cuxhaven (DE), Eemshaven (NL), Vlissingen (NL), Ostend (BE), Rotterdam Maasvlakte 2 (NL), Le Havre (FR), Hull/Grimsby (UK), Belfast (NI), Mostyn (UK), New London CT (US), New Bedford MA (US), Salem MA (US), Sparrows Point MD (US). Quayside crane capacity, water depth, and lay-down area drive port selection. US ports under significant ramp investment 2022-2025 (NJ Wind Port, NY South Brooklyn Marine Terminal, MA Salem) but most still constrained.

6. Operations and maintenance

O&M strategy

Annual planned maintenance per turbine ~120-200 person-hours; unplanned ~50-150 person-hours. Total O&M staff for 1 GW: 80-150 dedicated technicians + ops/management.

Vessels

  • Crew Transfer Vessels (CTV) — daily round-trip <90 min each way; 12-24 m catamarans accommodating 12-24 technicians; Damen Fast Crew Suppliers, Tidal Transit, Hst Marine, Northern Offshore Services, ESVAGT
  • Service Operation Vessels (SOV) — accommodation 40-90 technicians on multi-week rotations; motion-compensated walk-to-work gangway (Uptime Bridge, Ampelmann, SMST Velocity); DP2 dynamic positioning; Esvagt (DK; supplied Ørsted), Edda Wind (NO; Vard newbuilds 2022-2025), Bourbon Subsea Services, North Star (UK), VARD partner programs, REM Offshore
  • Commissioning Service Operation Vessels (CSOV) — pre-commissioning role
  • Helicopters — Bristow, NHV, Babcock; hoist transfer to turbine deck; used for far-out farms >70 km or weather-marginal CTV days

Predictive + condition monitoring

  • SCADA (Bachmann, Mita-Teknik, KK Wind Solutions) — every 10 min: 10,000+ data tags per turbine
  • Vibration monitoring on drivetrain (Brüel & Kjær Vibro, SKF IMx, GE Bently Nevada, Romax InSight) — 0.5-15 kHz accelerometers on bearings, generator stator
  • Oil-condition monitoring (Hydac, Eaton, Bosch Rexroth) — for geared turbines
  • Strain monitoring (fiber-optic FBG sensors — fiberSensing, Micron Optics) on critical structural elements
  • Blade lightning damage detection (Polytech, Global Lightning Protection Services)
  • Acoustic emission for blade crack detection (Brüel & Kjær, Vestas iVOR)

OEM service agreement (5-25 year full-scope) typically locked at SOP; lapse to merchant O&M (or owner-operated) is the dominant O&M cost reduction lever post-warranty.

Availability and capacity factor

  • Time-based availability target: 96-98%
  • Production-based availability: 95-97%
  • Capacity factor net: 45-55% (high-quality North Sea / US East Coast site)
  • For comparison: onshore CF 28-42%, utility solar CF 22-30%

O&M cost

$40-65/MWh-year typical (~25-30% of LCOE for modern projects). Decomposes:

  • Crew + labor: 30-35%
  • Vessels (CTV+SOV): 25-30%
  • Major component (gearbox, generator, blade, bearing) replacements: 20-25%
  • Spares, consumables: 10-15%
  • Onshore facility, port, admin: 5-10%

7. Output and energy economics

Annual energy production

  • Gross AEP: 4.5-5.5 GWh per MW installed in good North Sea sites
  • For 1 GW: 4.5-5.5 TWh/yr gross
  • Wake + electrical losses: 12-18%
  • Availability losses: 3-5%
  • Net AEP: 3.5-4.5 TWh/yr
  • Equivalent: power supply for 3.0-3.8 million EU homes (average 1,200 kWh/yr per house consumed, plus heat-pump and EV penetration scaling demand)

Levelized cost of energy (LCOE)

2024 benchmarks (BNEF, IRENA, Wood Mackenzie):

  • North Sea mature markets: 50-70 in 2019 due to inflation + interest rates
  • US East Coast: $90-130/MWh post-2022 (PPAs being renegotiated upward)
  • UK CfD Allocation Round 5 (2023): clearance £44/MWh in 2012 prices (≈£65/MWh today), zero awards for offshore (price too low)
  • UK CfD AR6 (2024): clearance £58/MWh (raised 66% from AR5), 3.3 GW awarded
  • UK CfD AR7 (2025): underway with structural changes for floating
  • Dutch tenders 2023-2024: zero-subsidy (IJmuiden Ver Alpha/Beta) becoming difficult; some EUR 0/MWh + non-price criteria

Capacity cost

  • 2024 CapEx: 3.5-4.5M/MW)
  • For 1 GW: $3.5-4.5 billion total capital
  • Floating: $5-8 M/MW (premium 1.5-2× fixed-bottom)

Cost decomposition (1 GW fixed-bottom)

  • Turbines (supply + transport + installation): $1,500M (~40%)
  • Foundations (supply + installation + UXO): $700M (~18%)
  • Offshore + onshore substations + transmission: $700M (~18%)
  • Inter-array + export cable: $400M (~10%)
  • Project development (survey, permit, engineering, IE): $200M (~5%)
  • Marshalling port + logistics: $150M (~4%)
  • Owner contingency: $200M (~5%)

Revenue streams

  • Wholesale market sales (Day-Ahead, Intraday, Balancing)
  • PPA (corporate buyers — Microsoft, Amazon, Google, Meta, big offtakers in Europe)
  • CfD strike-price top-up (UK contractual)
  • Tax credits (US IRA ITC 30%+ for projects beginning construction by 2025, PTC alternative; Inflation Reduction Act expansion)
  • Green certificates (REC, GO)

25-year cash flow shape

Years 1-3: Negative — devex + capex spend Year 4-5: Negative-flat — first power + ramp commissioning Years 6-15: Strong positive — full availability + lower OpEx Years 16-25: Moderate positive — major component refurbs (gearbox swap at year 12-15 for geared; first generator refurb DD at year 15-18; blade refurb / repower) Year 25+: Decommissioning provision (~1.5-3B for 1 GW deco bond. Repower option for partial component reuse.

8. Major developers and ownership

Top global offshore wind developers by operational capacity (2024):

DeveloperHQOperational capacity (GW)Major projects
ØrstedDK~9.0Hornsea 1+2 (UK, 2.4 GW total), Walney, London Array (50% sold), Anholt (DK), Borkum Riffgrund (DE), Block Island (US first), South Fork, Revolution Wind, Sunrise Wind
RWEDE~3.3Triton Knoll, Galloper, Sofia (UK 1.4 GW HVDC), Kaskasi (DE), Thor (DK 1.0 GW under-construction)
Iberdrola AvangridES/US~2.0Vineyard Wind 1 (US first commercial-scale 800 MW), East Anglia 1+3 (UK), Saint-Brieuc (FR), Wikinger (DE)
EquinorNO~1.7Hywind Tampen (88 MW floating), Sheringham Shoal, Dudgeon (UK 50% sale 2024), Empire Wind 1 (BP exited 2023), Dogger Bank A/B/C (40%)
EDF RenewablesFR~1.5Provence Grand Large, Saint-Nazaire (480 MW, first French commercial 2022), Fécamp, Courseulles
VattenfallSE~3.0Hollandse Kust Zuid 1.5 GW (NL — first zero-subsidy), Kriegers Flak (DK), DanTysk, Sandbank, Norfolk Vanguard +Boreas (UK; canceled 2023 due to inflation)
SSE RenewablesUK~3.7Dogger Bank A+B+C (40%), Seagreen 1.1 GW, Beatrice
Iberdrola Scottish Power RenewablesUK~1.5East Anglia 1+3, West of Duddon Sands
Shell New EnergiesNL~1.0Mayflower Wind (Atlantic Shores 50%), CrossWind (HKN, NL 50%)
TotalEnergiesFR~1.0Seagreen 51%, OW Ocean Winds JV (with EDPR)
Ocean Winds (EDPR + Engie JV)PT/FR~1.5Moray West, Noirmoutier (FR floating), East Anglia THREE 50%
BPUK(exiting)Empire Wind 1+2 (NY, 50% with Equinor; BP wrote down $1B 2023 and exited Dec 2024); Asian + UK leases retained but pace cut
IberdrolaES~1.5Saint-Brieuc, East Anglia 1+3, Wikinger
China Three Gorges RenewablesCN~3.0+Yangjiang, Putian, Fujian
China Longyuan PowerCN~2.0+Rudong, Putian
Mingyang IPP armCN~1.0+Vertical-integrated turbine + IPP
China General NuclearCN~1.0+Yangjiang, Shanwei
IberdrolaES(above)
Mitsubishi CorporationJP~0.5Akita Noshiro 140 MW (2023), Akita Yurihonjo, Kashima
SumitomoJP~0.3Akita JV, UK/EU stakes

9. Notable projects

Operating or near-operational (2024-2025)

  • Hornsea 1 (UK) — Ørsted, 1.2 GW, 174 × 7 MW Siemens Gamesa SG 7.0-154, COD 2020, world’s largest at COD
  • Hornsea 2 (UK) — Ørsted, 1.32 GW, 165 × 8 MW SGRE SG 8.0-167, COD 2022, retook world’s largest title
  • Hornsea 3 (UK) — Ørsted, 2.85 GW, ~190 × 14 MW, FID 2023, COD 2027 target — currently world’s largest under construction
  • Dogger Bank A+B+C (UK) — SSE-Equinor-Vårgrønn, 3.6 GW, 277 × 13 MW GE Haliade-X, first power 2024, COD 2026
  • Vineyard Wind 1 (US MA) — Avangrid-CIP, 800 MW, 62 × 13 MW GE Haliade-X, first power Jan 2024 (first US commercial-scale offshore wind), full COD 2024
  • Revolution Wind (US RI/CT) — Ørsted-Eversource, 704 MW, 65 × 11 MW SGRE SG 11.0-200, COD 2025
  • South Fork Wind (US NY) — Ørsted-Eversource, 132 MW, 12 × 11 MW SGRE, first US fully completed commercial Mar 2024
  • Coastal Virginia Offshore Wind (US VA) — Dominion Energy, 2.6 GW, 176 × 14.7 MW SGRE SG 14.0-236, first power 2026, COD 2026
  • Empire Wind 1 (US NY) — Equinor (now sole post-BP exit), 810 MW, COD 2027 (Equinor Pause Apr 2025 post Trump EO)
  • Sunrise Wind (US NY) — Ørsted (post-Eversource exit), 924 MW, COD 2026
  • He Dreiht (DE) — EnBW, 960 MW, 64 × 15 MW Vestas V236-15.0 MW (Vestas first 15-MW commercial), COD 2025
  • Saint-Brieuc (FR) — Iberdrola, 496 MW, 62 × 8 MW SGRE, COD 2024
  • Hollandse Kust Zuid (NL) — Vattenfall + Allianz, 1.5 GW, 139 × 11 MW SGRE, first zero-subsidy offshore farm, full COD 2023
  • Yunlin (Taiwan) — wpd-Yushan, 640 MW, 80 × 8 MW SGRE
  • Akita Noshiro / Yurihonjo (Japan) — Marubeni-led JV, 140 MW + 819 MW Round 1 2021 auction

Cancelled or paused (the 2023-2025 reset)

  • Ocean Wind 1+2 (US NJ) — Ørsted, 2.2 GW total, cancelled Nov 2023, $4B writedown (the largest single offshore wind cancellation in industry history); FID-stage abandoned due to inflation + supply chain
  • Park City Wind (US CT) — Avangrid, 804 MW, PPA terminated 2023
  • Commonwealth Wind (US MA) — Avangrid, 1.2 GW, PPA terminated 2022 (price renegotiation)
  • Skipjack 1+2 (US MD) — Ørsted, paused
  • Norfolk Vanguard + Boreas (UK) — Vattenfall, AR4 CfD strike price too low, written off Jul 2023, ~£250M loss
  • Beacon Wind (US NY) — Equinor-BP, suspended after BP exit 2024
  • Empire Wind 2 (US NY) — Equinor, PPA terminated Jan 2024
  • US offshore wind broadly — Jan 2025 Trump executive order paused new leasing and review of existing permits

Floating projects to watch

  • Hywind Tampen (NO) — 88 MW, 2023, world’s largest floating; powering Snorre + Gullfaks oil platforms
  • Hywind Scotland — 30 MW, 2017, first commercial floating
  • Kincardine (UK) — 50 MW, 2021, Principle Power WindFloat
  • WindFloat Atlantic (PT) — 25 MW, 2020
  • Provence Grand Large (FR) — 25 MW, 2024
  • ScotWind 2022 awards — 14 GW of 25 GW awarded for floating
  • US California Pacific lease auction Dec 2022 ($757M, 4.6 GW floating capacity awarded)
  • Korea, Japan floating tenders launching 2024-2026

10. Risk register

Market + financial

  • Interest rate sensitivity — offshore wind is capital-heavy, debt-financed; 2022-2024 rate rise added ~20-30% to LCOE
  • Inflation + supply-chain cost escalation — turbine + steel + cable prices up 25-40% from 2020-2022
  • CfD / PPA price re-opener risk — multiple US projects renegotiated post-Vineyard Wind; UK AR6 raised strike price after AR5 zero-award
  • PPA tenor + counterparty — corporate PPA market deepening but still <5 year typical tenors
  • US federal policy — Jan 2025 Trump EO pause; lease auctions frozen; existing permits under review

Technical

  • WTIV / cable-lay vessel shortage — binding constraint through 2027; charter rates volatile
  • 15+ MW turbine warranty experience — limited fleet hours; reliability data still building; SGRE 5.X platform suffered major reliability issues 2022-2024 (~€2.2B charges); Vestas first 15 MW units installed 2024
  • Foundation manufacturing capacity — XL monopile yards (Sif, SeAH, Steelwind, EEW, Haizea) ramping but at most ~30-40 GW/yr global combined capacity through 2026
  • Permanent magnet supply — NdFeB DD reliance on Chinese supply; tariff / geopolitical risk
  • Marine growth + cathodic protection — corrosion management 25-year life
  • Cable failures — historically a major insurance claim category (50-70% of claims by value); CIGRE TB 689 + IEC 63026 + improved testing protocols + dynamic cables for floating

Environmental + permitting

  • Marine mammal litigation — North Atlantic right whale concerns in US (NOAA permitting), recent litigation in MA / VA
  • Fisheries opposition — US lobster/scallop industry in MA/RI/ME organized opposition
  • Bird strike — annual mortality estimates 1-10 birds per turbine; cumulative effects assessment increasingly required (Norfolk Vanguard rejected on cumulative-impact grounds, overturned, eventually cancelled for economic reasons)
  • UXO survey + clearance — North Sea WWII ordnance; $20-100M+ project-level line item
  • Visual + radar impact — onshore radar interference, military aviation (Cape Wind cancellation 2017 — Massachusetts opposition + Pentagon concerns)

11. Adjacent

Graph View

Backlinks