Offshore Wind Investment Intelligence 2026–2030: Where Capital Is Moving and Where Risks Are Rising
By the Green Fuel Journal Research & Intelligence Team
Last reviewed: October 2026 · Next review trigger: close of the KKR–RWE transaction for 50% of Norfolk Vanguard, expected during 2026
Published 29 September 2026, under the ISSN 2979-3777
GreenFuelJournal.com · Global scope · Evidence date: 1 October 2026 · Reader: energy investors
THIS IS REPORT IS IN 2 PARTS
Legal Disclaimer
This report is informational research published by Green Fuel Journal. It is not investment, legal, tax or financial advice, and it does not recommend buying, selling or holding any security or asset. Figures come from named third-party sources, were read from the published pages as of 1 October 2026, and may change or be corrected. Forward-looking statements are analysis, not guarantees. Readers should verify figures against the cited sources and seek independent professional advice before acting.
Read the full terms at greenfueljournal.com/disclaimers.
Executive Summary
Offshore Wind investment is stalling at the decision stage: TGS | 4C counts only 2.1 GW at final investment decision (FID) so far in 2026, against more than 12 GW due to enter operation this year. Capital is moving where a two-way contract for difference (CfD) removes merchant price risk, led by the UK's seventh CfD round (AR7), which awarded 8.4 GW at an average £90.91/MWh (2024 prices) in January 2026. Cost of capital then sets the returns: TGS | 4C puts the effect of each percentage point of financing cost at about 9–11% on the cost of electricity. TGS | 4C expects 5.6 GW of FID by year-end, which would make 2026 the second-lowest year for investment decisions since 2020, and BloombergNEF (BNEF) reports offshore wind investment down 72% year on year in the first half of 2026. The shortage is not projects: the Global Wind Energy Council (GWEC) counts about 25 GW outside China that hold consents and await FID. Each headline figure rests on a single publisher, and TGS | 4C has not published a country breakdown.
Contract design explains where the money goes. Poland awarded 3.4 GW at about €113–117/MWh, and Denmark drew bids again in May 2026 after switching to a CfD. Germany and the Netherlands, where unsubsidised rounds drew no bids in August 2025 and October 2025, are moving the same way. Capital is impaired where that protection is absent: the US has paid $3.9bn in lease buyouts to six developers since March 2026 with no new FID identified, India cancelled its 4 GW and 500 MW tenders in August 2025, and the Philippines suspended its 3.3 GW auction in July 2026. Even in CfD markets, no AR7 project has reached a confirmed FID as of 1 October 2026.
Long-term borrowing sits near 15-year highs in the UK, US and euro area, and Ørsted booked a DKK 1.2bn US impairment in Q2 2026 that it tied to higher long-dated US interest rates. Published cost benchmarks differ by basis: $78/MWh (International Renewable Energy Agency, IRENA, 2025 commissioned projects), $100/MWh (BNEF, 2025 global benchmark) and $105–167/MWh (Lazard, US, unsubsidised), against $102/MWh for new gas on BNEF's benchmark. For an energy investor the screen runs in three steps: revenue-risk design first, sponsor sell-down and funding second, cost of capital third. The evidence has limits: no second source confirms capital expenditure (CAPEX), and the make-up of the 2.1 GW is not public.
1. Executive Intelligence Synthesis
DIRECT ANSWER
Five signals matter for offshore wind investors in 2026. Conversion, not pipeline, is the constraint: 2.1 GW reached FID against more than 12 GW entering operation. Two-way CfDs decide which markets get funded. Cost of capital moves returns by about 9–11% per percentage point. Sponsors reach FID by selling equity to partners. Policy risk has stranded capital in the US, Japan, India and the Philippines.
1.1 The Five Executive Signals
Signal | Detail |
2.1 GW | FID in 2026 to date vs 12 GW+ commissioning (TGS | 4C, 21 Sep 2026) |
£90.91/MWh | UK AR7 average, 8.4 GW, two-way CfD (14 Jan 2026) |
9–11% | Rise in levelised cost of electricity (LCOE) per percentage point of financing cost (TGS | 4C) |
50% | Norfolk Vanguard stake sold to KKR; deal still to close (RWE, 10 Aug 2026) |
$3.9bn | US lease buyouts to six developers since Mar 2026 (Roll Call, 30 Sep 2026) |
Signal 1. Conversion, not pipeline, is the constraint
FINDING: The 2026 decision count is a fraction of the build count: 2.1 GW at FID so far, set against more than 12 GW entering operation this year (TGS | 4C, single-source, 21 September 2026).
SO WHAT: Today's commissioning numbers reflect decisions taken years ago, so they say little about the capital an investor would be joining for 2027–2030.
NOW WHAT: Before treating a project's pipeline stage as evidence of investability, check whether it has a signed offtake contract and a funded sponsor.
Signal 2. Contract design decides who gets funded
FINDING: The UK's seventh CfD round (AR7) awarded 8.4 GW at an average £90.91/MWh (2024 prices) on 14 January 2026, and Poland awarded 3.4 GW at about €113–117/MWh in December 2025, both under two-way contracts for difference.
SO WHAT: Germany (August 2025), the Netherlands (October 2025) and Denmark (December 2024) each ran unsubsidised or negative-bid rounds that drew no bids, so contract design, not site quality, separated rounds that attracted bidders from rounds that did not.
NOW WHAT: Before reading an auction result as a demand signal, check whether the revenue contract is two-way, indexed and long enough to lend against.
Signal 3. Cost of capital moves returns faster than capex
FINDING: On TGS | 4C's estimate each point of financing cost adds roughly 9–11% to electricity cost (single-source), and long-term borrowing sits near 15-year highs in the UK, US and euro area.
SO WHAT: A project priced before rates rose can lose its margin through financing alone, which is the mechanism Ørsted cited for its DKK 1.2bn Q2 2026 US impairment.
NOW WHAT: Ask for the financing assumption behind any quoted return, and re-run it one percentage point higher before comparing projects.
Signal 4. Sponsors fund FID by selling down, not from cash flow
FINDING: RWE's 6.9 GW AR7 win relies on KKR taking 50% of Norfolk Vanguard, a deal RWE's 10 August 2026 report says is "expected to close over the course of this year", while Apollo-managed funds bought 50% of Ørsted's Hornsea 3 for DKK 39bn.
SO WHAT: Large projects reach FID when a sponsor sells equity to a financial partner, so the partner's closing, not the auction win, is the practical gate.
NOW WHAT: Check whether a project has a signed equity partner, and whether that deal has closed, before treating an award as investable.
Signal 5. Policy risk has stranded capital outside Europe
FINDING: The US has reimbursed about $3.9bn to six developers under 12 lease buyout agreements since March 2026, India cancelled its 4 GW and 500 MW tenders in August 2025, the Mitsubishi-led consortium left Japan's Round 1 in August 2025, and the Philippines suspended its 3.3 GW auction on 4 July 2026.
SO WHAT: In these markets the loss came from the rules or the procurement process, not from the technology, so a good site does not protect the investment.
NOW WHAT: Test any non-European entry against that market's procurement record, such as cancelled or reset tenders, before assessing the project itself.
1.2 How to Read This Report's Evidence Labels
FINDING: Each figure in this report carries an evidence label: TWO-SOURCE only where two independent publishers report the same metric on a comparable basis, and SINGLE-SOURCE where one publisher, or publishers using different bases, dates or geographies, stand behind it.
SO WHAT: A trade-press repeat of one publisher's number is still one number, so repetition adds no confirmation.
NOW WHAT: Treat every SINGLE-SOURCE figure as provisional and open the publisher's own page before using it in a model.
Two more labels appear. "Inference" marks the analysts' reasoning from sourced facts, not a published claim. "Author arithmetic" marks a number we derived from sourced figures, which no publisher states. The evidence limits are set out in Section 12.1.
2. Macro Context & Strategic Drivers
DIRECT ANSWER
Offshore wind capacity keeps growing because projects decided years ago are now being built, while new decisions fall. GWEC counts 92.5 GW installed at end-2025, and 41.2 GW is under construction (TGS | 4C), which supports commissioning through about 2029. Only 2.1 GW reached FID in 2026 to date. High cost of capital and weak offtake design explain the gap.
2.1 Installed Base and Why the Numbers Differ
FINDING: GWEC counts 92.5 GW of offshore wind installed at end-2025, with China at 48.4 GW, Europe at 38 GW and North America at 0.2 GW, after 9.3 GW of additions in 2025.
SO WHAT: IRENA, Orrick and TGS | 4C publish totals of 91.4 GW, 89.2 GW and 92.7 GW, so any market-share or growth calculation depends on which baseline it uses.
NOW WHAT: Name the source and date beside every capacity figure, and never blend two publishers into one total.
The gaps may reflect dataset definitions or cut-off dates, but no source confirms the cause. China shows the widest spread: GWEC gives 48.4 GW and 6.6 GW added in 2025, Orrick gives about 46.4 GW and about 6 GW, and state-linked Global Times reports more than 47 GW grid-connected at end-February 2026. Europe differs too: WindEurope puts it at 40.9 GW in the first half of 2026, against GWEC's 38 GW at end-2025, which is a later date. The IRENA figure of 91.4 GW comes from a headline only and not from IRENA's own page.
Evidence depth also varies by region. The sources hold full data for the US, Canada, the EU, the UK, China, Taiwan, Japan, South Korea and India; partial data for the Philippines, Vietnam, Australia, Brazil and South Africa; and no relevant data for the Gulf.
Source | Figure | Basis and date | Note |
GWEC | 92.5 GW (China 48.4 GW; Europe 38 GW; North America 0.2 GW) | End-2025 | Global Offshore Wind Report 2026 |
IRENA | 91.4 GW | 2025 | Headline of a trade write-up only; verify against IRENA |
Orrick | 89.2 GW (China about 46.4 GW) | End-2025 | About 6 GW added by China in 2025 |
TGS | 4C | 92.7 GW operating | "Currently", September 2026 | Later date; not like-for-like with end-2025 |
Global Times | China above 47 GW grid-connected | End-February 2026 | State-linked media |
WindEurope | Europe 40.9 GW; EU-27 23.2 GW | H1 2026 | Against GWEC Europe 38 GW at end-2025 |
T1. Installed offshore wind capacity baselines in conflict. Sources: GWEC (9 June 2026); Orrick (February 2026); TGS | 4C via Windtech International (22 September 2026); Global Times (6 April 2026); WindEurope (Autumn 2026 data); IRENA via GreentechLead. All SINGLE-SOURCE on their own bases.
CHART PLACEHOLDER V4 — INSTALLED OFFSHORE WIND BY REGION, END-2025
Type: bar (comparison). Placement: Section 2.1, after T1. Data (GWEC, end-2025): China 48.4 GW; Europe 38 GW; North America 0.2 GW. Three named regions only; no derived remainder.
Caption: Installed offshore wind by region, end-2025. Source: GWEC, Global Offshore Wind Report 2026, 9 June 2026.
2.2 The Pipeline-to-FID Evidence Ladder
FINDING: Published figures for each stage run from 11.4 GW auctioned worldwide in 2025 (GWEC) and about 25 GW consented outside China awaiting FID (GWEC), through 22.2 GW of consents expected in 2026 (TGS | 4C), to 41.2 GW under construction, 92.7 GW operating and only 2.1 GW at FID in 2026 so far.
SO WHAT: The figures come from two publishers with different years and methods, so they show where projects sit, not what share of them converts.
NOW WHAT: Do not compute a ratio between stages; ask which stage a given project occupies and what contract and sponsor evidence supports it.
Read from the top, GWEC reports that only 11.4 GW was auctioned in 2025, one-fifth of the 2024 record. The projects that already hold consents are numerous: GWEC counts about 25 GW outside China awaiting FID, and TGS | 4C forecasts 22.2 GW of consents in 2026, the second-highest on record. Further down the ladder, TGS | 4C counts 7.1 GW "at FID", which is a stock of projects at that stage and not yet in construction, a different measure from the 2.1 GW flow of 2026 decisions. The 41.2 GW under construction and the 92.7 GW operating are TGS | 4C figures at September 2026.
CHART PLACEHOLDER V1 — PIPELINE-TO-FID EVIDENCE LADDER
Type: horizontal bar (comparison), colour-coded by publisher. Placement: Section 2.2; also the one-page brief.
Data: 2.1 GW FID 2026 to date; 5.6 GW FID expected by year-end 2026; 7.1 GW at FID (stock); 22.2 GW consents expected 2026 (all TGS | 4C). 41.2 GW under construction; 92.7 GW operating (TGS | 4C). About 25 GW consented outside China awaiting FID; 11.4 GW auctioned in 2025 (GWEC).
Caption: Different publishers and methods; not a conversion rate. Source: TGS | 4C, 21–22 September 2026; GWEC, 9 June 2026.
METHODOLOGY NOTE
The ladder lists published figures for each stage from auction to operation, in the order a project moves through them. It is not a computed index or conversion rate. The figures come from two publishers (TGS | 4C and GWEC) and different years, definitions and geographies, and no source reconciles them; the report therefore shows each figure with its publisher and date and draws no percentage from the ladder. The one figure used for the 2026 FID count (2.1 GW) is single-source.
2.3 Policy Design as the Dividing Line
FINDING: Nine North Sea governments signed the Hamburg Declaration and Investment Pact on 26 January 2026, making two-sided CfDs the standard auction mechanism and setting 15 GW a year for 2031–2040.
SO WHAT: Germany, the Netherlands and Denmark each tried an unsubsidised or negative-bid design, each saw it fail, and each has moved to two-way support, which leaves the US, Japan and India as the large markets without an equivalent in force.
NOW WHAT: Sort any market by whether a two-way revenue mechanism is in force today, not announced, before comparing resource quality.
Ørsted's Group President and CEO, Rasmus Errboe, described the German shift on the company's Q2 2026 call on 13 August 2026:
"It is a return to CfDs to two-way CfDs that was, in our view, sort of about time to be a bit blunt in Germany."
The Hamburg Investment Pact also includes an industry pledge of 30% cost reduction by 2040 against 2025, per WindEurope.
Market | Mechanism | Latest outcome | Date |
UK | Two-way CfD, 20-year, CPI-indexed | AR7: 8.4 GW awarded at £90.91/MWh average (2024 prices) | 14 Jan 2026 |
Poland | Two-way CfD, up to 25 years | 3.4 GW+ awarded at PLN 476.88–492.32/MWh (about €113–117) | 17 Dec 2025 |
Denmark | Two-sided CfD after a failed no-subsidy tender | Bids received; results expected January 2027 | 20 May 2026 |
Germany | Unsubsidised round failed; two-stage redesign (adoption pending) | No bids in the round; 2026 auctions pushed to 2027 | Aug 2025; 28 Jan 2026 |
Netherlands | Temporary subsidy until a CfD in mid-2027 | No bids for Nederwiek I-A; 2 GW tender, deadline December 2026 | Oct 2025; 6 Apr 2026 |
US, Japan, India | No equivalent two-way mechanism in force | Lease buyouts (US); Round 1 withdrawal (Japan); tenders cancelled (India) | 2025–2026 |
T2. Revenue mechanism and latest auction outcome by market. Sources: UK Department for Energy Security and Net Zero (DESNZ) and UK Parliament written statement (14 January 2026); offshoreWIND.biz (19 December 2025, 22 May 2026, 6 April 2026); Windtech International (28 January 2026); WindpowerNL (31 October 2025); Freshfields (Germany, adoption date not published).
2.4 Why 2030 Forecasts Disagree
FINDING: GWEC projects 24% compound annual growth for 2026–2030 and 420 GW by end-2035, while TGS | 4C cut its 2030 outlook by 28% in December 2025 (non-China 2030 from 192 GW to 121 GW) and the International Energy Agency (IEA) cut its five-year forecast by more than 25% on 8 October 2025.
SO WHAT: The forecasters differ mainly on how many consented projects reach FID and on US and Asian policy (inference), so the gap between forecasts is a gap in FID assumptions.
NOW WHAT: Before using any 2030 figure, find the FID assumption behind it and compare it with the 2.1 GW seen so far in 2026.
GWEC also says additions roughly double in 2026 from 9.3 GW, while TGS | 4C expects more than 12 GW to enter operation; no source states the doubled figure in GW. TGS | 4C's longer outlook is 474 GW by 2040. Wood Mackenzie's November 2025 outlook expects offshore additions to more than double in 2026 and warns of a slowdown in build-out from 2028, without a GW forecast. These forecasts are SINGLE-SOURCE each and rest on press relays of reports not available in full.
3. India-Specific Analysis
DIRECT ANSWER
India's offshore wind market is not yet investable: no project has been awarded. In August 2025 the Solar Energy Corporation of India (SECI) cancelled a 4 GW Tamil Nadu seabed-lease auction and a 500 MW Gujarat project for lack of bids. A Rs 7,453 crore scheme, including Rs 6,853 crore of viability gap funding for 1 GW, is approved, but as of 3 July 2026 the 1 GW tender was only expected, not issued. On that record, the evidence does not support investment.
3.1 Policy Architecture and the VGF Scheme
FINDING: India's Cabinet approved a Rs 7,453 crore scheme: Rs 6,853 crore of viability gap funding (VGF) for 1 GW, split 500 MW each off Gujarat and Tamil Nadu, and Rs 600 crore for two ports (DD News, undated).
SO WHAT: Public money exists for a first 1 GW, but the sources do not show how much of the estimated tariff gap it covers.
NOW WHAT: Obtain the MNRE VGF guidelines to see whether any quoted tariff is before or after VGF.
India has had a National Offshore Wind Energy Policy since 2015 and a stated aim of 30 GW by 2030 (US International Trade Administration, updated 1 October 2024). The Power Grid Corporation of India (PGCIL) is to build transmission, and India and the UK launched an Offshore Wind Task Force in February 2026. Every Indian figure in this section is SINGLE-SOURCE and rests on trade press and government media. We read no primary document from the Ministry of New and Renewable Energy (MNRE), the Press Information Bureau (PIB), SECI, the Central Electricity Regulatory Commission (CERC) or the National Institute of Wind Energy (NIWE). Readers following the demand side of India's power system can see our analysis of India's AI power strategy and electricity infrastructure.
Item | Figure | Note |
Total scheme | Rs 7,453 crore | Cabinet-approved; approval reported June 2024, DD News page undated |
VGF for projects | Rs 6,853 crore for 1 GW | 500 MW off Gujarat, 500 MW off Tamil Nadu |
Ports | Rs 600 crore | Two ports |
Transmission | PGCIL to build | Only confirmed grid actor |
Policy aim | 30 GW by 2030 | US International Trade Administration, October 2024 |
T12. India's offshore wind VGF scheme at a glance. Sources: DD News (undated); US International Trade Administration (updated 1 October 2024). SINGLE-SOURCE.
3.2 The Tender Record, 2025–2026
FINDING: SECI cancelled a 4 GW seabed-lease auction off Tamil Nadu (issued February 2024) and a 500 MW Gujarat project (issued September 2024) in August 2025 for lack of bids (Down To Earth, March 2026).
SO WHAT: Both competitive offshore processes we found failed to attract bidders, so no price discovery has happened.
NOW WHAT: Treat the first issued 1 GW VGF tender, and the number of bids it draws, as the next test, and check for a tender notice before assuming it exists.
On 11 March 2026 an unnamed MNRE official told Down To Earth that the 1 GW tender was ready but that the ministry might not draw enough interest. This is reporting, not an attributed quote. As of 3 July 2026, Indian Infrastructure described fresh tenders as expected in the second half of 2026. No source after 3 July 2026 shows the tender issued, awarded or cancelled, so we describe it as approved and not yet tendered. Tender cycles in India run 100–120 days, against 6–7 months in Europe (Down To Earth), which gives developers less time to line up financing and supply.
3.3 Economics: What the Tariff and Cost Signals Say
FINDING: Reported tariff estimates conflict: Rs 7–9/kWh (Down To Earth, March 2026) against about Rs 9.60/kWh for Tamil Nadu and Rs 10.50/kWh for Gujarat (Indian Infrastructure, 3 July 2026), against Rs 2.8–2.9/kWh for onshore wind.
SO WHAT: The Indian Infrastructure offshore tariff estimates run at more than three times the onshore figure (author arithmetic: Rs 9.60 against Rs 2.9), and neither tariff source says whether VGF is included.
NOW WHAT: Ask which tariff basis a state buyer would accept before modelling any Indian project, and treat both as unverified published estimates.
Tamil Nadu has committed to a Rs 4/kWh premium, according to Outlook Business (5 November 2025), which paraphrases the managing director of the state's Green Energy Corporation; we did not find the underlying document. Each figure here is SINGLE-SOURCE, and the two tariff sets cannot be reconciled from published sources.
3.4 Ports, Grid and Supply Chain
FINDING: The V.O. Chidambaranar Port (Tuticorin) outer-harbour project is reported at Rs 15,000 crore, with a renewable-energy hub designation (The Week, 6 June 2026), and PGCIL is the only confirmed grid actor.
SO WHAT: A first project needs a port, a grid connection and installation vessels, and the public record confirms a port programme and one transmission builder but nothing on the rest.
NOW WHAT: Ask any sponsor which vessels, substations and turbine supply it has secured, because the public record shows none.
Vessels, offshore substations and domestic manufacturing are all unknown in the sources. No grid-evacuation plan is published beyond PGCIL being named.
3.5 Investment-Readiness Pathway and the Binding Constraint
FINDING: Of five gates a project must pass, namely policy support, competitive procurement, offtake, financing and FID, the public record confirms policy support (a Rs 7,453 crore scheme) and finds procurement failed twice in August 2025.
SO WHAT: Policy support has arrived before the evidence that anyone will bid, so India stands at the second gate, not the fifth.
NOW WHAT: Track the first issued VGF tender notice, its tariff basis and its bid count, which together decide whether the market moves to the offtake gate.
Gate | What the public record confirms | What it does not confirm |
1. Policy support | 2015 policy; 30 GW by 2030 aim; Rs 7,453 crore scheme | VGF guidelines and tariff treatment |
2. Competitive procurement | 4 GW and 500 MW tenders cancelled, August 2025; 1 GW tender "expected" at 3 July 2026 | A tender notice, bid count or award |
3. Offtake or power purchase agreement (PPA) | Tamil Nadu Rs 4/kWh premium (reporter's paraphrase) | Any signed PPA or buyer commitment |
4. Financing | Nothing | Capital cost per MW, financing cost, lenders |
5. FID | No project at FID | A realistic commissioning date |
T13. India investment-readiness checklist. Status record to 3 July 2026. Sources: Down To Earth (March 2026); Indian Infrastructure (3 July 2026); Outlook Business (5 November 2025); DD News (undated).
METHODOLOGY NOTE
The checklist lists five gates a project must pass (policy support, competitive procurement, offtake, financing, FID) and records for each what the public record confirms and what it does not. It is a status record built from the published record to 3 July 2026; it contains no score. India's evidence rests on trade-press and government-media reporting, which is stated beside the table.
Analyst reading (inference). The constraint most likely to bind first is tariff economics. The estimated tariff is far above onshore wind, and the ministry itself reportedly doubted that the ready tender would draw interest. Procurement timing, ports and grid come next. Our report on green ammonia exports from India and China covers an Indian export case with a fuller public record.
4. Operational / Technical Deep-Dive: Project Economics and the Auction-to-FID Chain
DIRECT ANSWER
Published offshore wind LCOE ranges from $78/MWh (IRENA, 2025 commissioned projects) to $100/MWh (BNEF, 2025 global benchmark) and $105–167/MWh (Lazard, US, unsubsidised), because the bases differ. The UK DESNZ medium case is £88.5/MWh with CAPEX of £3,976/kW. TGS | 4C estimates each percentage point of financing cost adds about 9–11% to the cost of electricity. Offshore wind costs about two to two-and-a-half times onshore wind on BNEF's and IRENA's figures (author arithmetic). All three benchmarks are single-source.
4.1 Unit Economics: LCOE, CAPEX and OPEX
FINDING: BNEF's 2025 benchmark puts offshore wind at $100/MWh (up 12%), IRENA's 2025 commissioned-project average at $78/MWh, and Lazard's US unsubsidised range at $105–167/MWh, against $102/MWh for new gas on BNEF's benchmark.
SO WHAT: The spread reflects China's weight in IRENA's average against Europe- and US-centred benchmarks (inference), so the three figures are not contradictory data.
NOW WHAT: Quote the publisher and basis with every LCOE figure and never average them.
IRENA's figure is a global weighted average of projects commissioned in 2025, with China about 75% of new wind. BNEF's is a 2025 global benchmark, and BNEF notes that UK offshore projects cost 69% more than five years ago. Lazard's range is US-only, and the drivers it cites are higher capital costs, sustained interest rates, tariff pass-through and supply-chain repricing. Offshore wind has the highest or equal-highest cost of the technologies shown in each dataset: $100 against onshore wind at $40, solar at $39 and gas at $102 on BNEF's benchmark, and $78 against onshore at $33 and solar at $44 on IRENA's. That puts offshore at roughly two to two-and-a-half times onshore wind on BNEF's and IRENA's figures, and at parity with new gas on BNEF's (author arithmetic). For readers weighing the gas comparison, our analysis of LNG as a transition fuel to 2030 covers the gas side of the trade.
Technology | BNEF 2025 global benchmark (18 Feb 2026) | IRENA, 2025 commissioned projects (1 Jul 2026) | Lazard LCOE+ 2026, US, unsubsidised (16 Jul 2026) |
Offshore wind | $100/MWh, up 12% | $78/MWh | $105–167/MWh |
Onshore wind | $40/MWh, up 2% | $33/MWh | $37–99/MWh |
Solar PV | $39/MWh (fixed-axis), up 6% | $44/MWh, unchanged on 2024 | $40–98/MWh (utility-scale) |
Gas combined cycle | $102/MWh, up 16% | Not available | $51–129/MWh |
Label | SINGLE-SOURCE | SINGLE-SOURCE | SINGLE-SOURCE |
T3. Levelised cost by publisher, basis and date. Sources: BNEF (18 February 2026); IRENA, Renewable Power Generation Costs in 2025, as listed by PNNL Tethys (the IRENA PDF could not be opened); Lazard LCOE+ 2026 as reported by Utility Dive (16 July 2026).
UK unit costs come from one official source, the DESNZ/Arup cost study dated 24 July 2025, which is more than 12 months old. For fixed-bottom projects reaching commercial operation in 2028–2032, it gives CAPEX of £3,321 / £3,976 / £4,439 per kW (low / medium / high), operating expenditure (OPEX) of £73.7k / £138.5k / £207.2k per MW a year, and a net load factor of 46.0% / 50.5% / 56.1% over a 35-year life. Outside the UK, OIES, citing IRENA, puts China's average offshore installed cost at $1,520/kW against $3,389/kW in Europe (single-source).
UK fixed-bottom offshore wind (2023 real prices) | Low | Medium | High |
LCOE, commercial operation 2028–2032 | £55.4/MWh | £88.5/MWh | £124.0/MWh |
CAPEX | £3,321/kW | £3,976/kW | £4,439/kW |
OPEX | £73.7k per MW a year | £138.5k per MW a year | £207.2k per MW a year |
Net load factor | 46.0% | 50.5% | 56.1% |
Asset life and hurdle rate | 35 years; 6.2% pre-tax real hurdle rate, derived from BNEF H2 2023 data | ||
T4. UK DESNZ/Arup unit costs. Source: DESNZ/Arup, Renewable Energy Generation Cost and Technical Assumptions: Offshore Wind (24 July 2025). SINGLE-SOURCE; older than 12 months. DESNZ notes a bias risk in a sample of 11,760 MW.
CHART PLACEHOLDER V2 — OFFSHORE WIND LCOE AGAINST ONSHORE WIND, SOLAR AND GAS
Type: grouped bar with floating ranges (cost). Placement: Section 4.1, after T3. Data: BNEF $100 / $40 / $39 / $102; IRENA $78 / $33 / $44 (offshore / onshore / solar / gas, gas not in IRENA); Lazard US $105–167 / $37–99 / $40–98 / $51–129 shown as floating ranges.
Caption: Three publishers shown separately; each basis differs; SINGLE-SOURCE. Source: BNEF (18 Feb 2026); IRENA (1 Jul 2026); Lazard via Utility Dive (16 Jul 2026).
4.2 The Auction-to-FID Risk Chain
FINDING: UK AR7 cleared 19.29–20.81% below the administrative strike prices (£113/MWh fixed-bottom and £271/MWh floating, 2024 prices), yet no FID is confirmed for any AR7 project as of 1 October 2026.
SO WHAT: Clearing below a ceiling shows developers will bid for a CfD at that level, not that financing is secured.
NOW WHAT: Follow every award to its financing close, and treat the gap between award date and FID as the real measure of risk.
The Auction-to-FID Risk Chain runs in six links: bid price, inflation, financing, supply chain, revised economics, FID. Each link has at least one named, dated example. Japan's Round 1 is the clearest: the Mitsubishi-led consortium won with bids of JPY 11.99–16.49/kWh against a JPY 29 ceiling, costs were reported to have more than doubled, and the group withdrew in August 2025. Germany's August 2025 round and the Netherlands' October 2025 round drew no bids at all. Ørsted's Hornsea 4 (2.4 GW) was cancelled in its current form in May 2025.
INFOGRAPHIC PLACEHOLDER F1 — AUCTION-TO-FID RISK CHAIN
Type: flow infographic (SVG). Placement: Section 4.2. Flow: Bid price → Inflation → Financing → Supply chain → Revised economics → FID. One dated example under each link, as in T5.
Caption: A qualitative chain; no probabilities attached. Source: IEEFA; offshoreWIND.biz; DESNZ; Ørsted; see T5.
Link | Evidence | Date | Source |
Bid price | Japan Round 1 bids JPY 11.99–16.49/kWh against a JPY 29 ceiling; AR7 cleared 19.29–20.81% below administrative prices | 2021; Jan 2026 | IEEFA; DESNZ |
Inflation | Japan Round 1 costs reported to have more than doubled; steel up 53% year on year | 2025; Sep 2026 | IEEFA; TGS | 4C |
Financing | Borrowing costs near 15-year highs; Ørsted DKK 1.2bn US impairment tied to long-dated US rates | Sep 2026; Aug 2026 | TGS | 4C; Ørsted |
Supply chain | Vestas offshore order intake 0 MW in Q2 2026; DESNZ reports a severe high-voltage direct current (HVDC) shortfall in 2025–26 | Aug 2026; 2026 | Vestas; DESNZ |
Revised economics | Hornsea 4 cancelled in current form; Mitsubishi-led exit with JPY 52.2bn impairment | May 2025; Aug 2025 | Windtech International; IEEFA |
FID | 2.1 GW in 2026 to date; no confirmed AR7 FID; Germany and the Netherlands drew no bids | Sep 2026 | TGS | 4C; Clean Energy Wire; WindpowerNL |
T5. Auction-to-FID evidence by link in the chain. Where a link rests on analyst inference, the text says so.
METHODOLOGY NOTE
The chain is a qualitative causal framework, not a model. The links (bid price, inflation, financing, supply chain, revised economics, FID) were chosen because each has at least one named, dated example in public sources. Where the evidence for a link is an inference rather than a published finding, the text says so. No probabilities or return figures are attached to the chain.
4.3 Cost-of-Capital Sensitivity
FINDING: TGS | 4C estimates that each percentage point of financing cost lifts the cost of electricity by about 9–11%, with borrowing near 15-year highs (single-source), and Ørsted booked a DKK 1.2bn US impairment in Q2 2026 on higher long-dated US rates.
SO WHAT: Only the one-point effect is sourced, so any larger rate shock is an unpublished extrapolation and can mislead.
NOW WHAT: Stress-test one percentage point at a time, and ask for a second source before extending the 9–11% to two or more points.
An Ørsted white paper (2025), developer-authored, puts the rise in weighted average cost of capital (WACC) at roughly 3–4 points over 2020–2024, with CfD backing lowering WACC by about 2 points and merchant risk adding about 1.5 points (single-source). Aegir Insights estimates that a premium of EUR 79–91/MWh is needed for a 10% equity internal rate of return (IRR) on Japan's Round 1 rebid sites. A straight-line extension of the 9–11% to the white paper's 3–4 point rise would give a very large LCOE effect, but that is author arithmetic with no published support, so we do not state it as a number. For a related look at how financing costs move through power-sector deals, see our new energy M&A playbook for 2026–2027.
Worked example — Does a £91.20/MWh strike price survive a one-point rise in financing costs?
Entity: Northbank Offshore, a fictional developer holding a 1 GW fixed-bottom project under an AR7-style CfD. Sourced figures and illustrative figures are kept apart below.
Sourced inputs | Value | Source |
Strike price | £91.20/MWh (2024 prices) | DESNZ, AR7 results, January 2026 |
Medium-case LCOE | £88.5/MWh (2023 real prices) | DESNZ/Arup, 24 July 2025 |
Net load factor (medium) | 50.5% | DESNZ/Arup, 24 July 2025 |
CAPEX (medium) | £3,976/kW | DESNZ/Arup, 24 July 2025 |
LCOE sensitivity | 9–11% per percentage point of financing cost | TGS | 4C, 21–22 September 2026 (single-source) |
Illustrative inputs (assumptions, not sourced): project size of 1,000 MW; 8,760 hours in a year; no curtailment; the strike price paid on all output; a rise of exactly one percentage point in financing cost.
Calculation (author arithmetic):
1. Annual output = 1,000 MW × 8,760 hours × 0.505 = 4,423,800 MWh (about 4.42 TWh).2. Capital at stake = 1,000,000 kW × £3,976 = £3,976m (£3.976bn).3. Starting margin = £91.20 − £88.50 = £2.70/MWh, or about £11.9m a year (£2.70 × 4,423,800 MWh).4. LCOE after a one-point rise = £88.50 × 1.09 to £88.50 × 1.11 = £96.47 to £98.24/MWh.5. New margin = £91.20 − £96.47 = −£5.27/MWh to £91.20 − £98.24 = −£7.04/MWh.6. Annual shortfall = £5.27 × 4,423,800 to £7.04 × 4,423,800 = £23.3m to £31.1m a year.
Reading. On these figures, the cushion of £2.70/MWh disappears with a one-point rise in financing cost, and the project would need an extra £5.27–£7.04/MWh, or lower costs, to break even. The result tests direction, not outcome: the strike price is in 2024 prices and the LCOE in 2023 real prices, TGS | 4C's sensitivity is applied to a DESNZ figure from a different publisher, and no tax, indexation or construction timing is modelled.
Apply it to your own project: annual shortfall = (base LCOE × (1 + s) − strike price) × annual MWh, where s is 0.09 to 0.11 for each percentage point of financing-cost rise. Replace the LCOE, strike price and annual MWh with your own.
4.4 Supply Chain and Grid, Translated into Schedule and Cost
FINDING: Steel prices are up 53% year on year (TGS | 4C), Vestas booked 0 MW of offshore orders in Q2 2026 against an offshore backlog of EUR 12.0bn, and Siemens Gamesa's orders fell 78.5% on a comparable basis to EUR 1,050m.
SO WHAT: Turbine makers are returning to profit while offshore orders stall, so supplier recovery has not yet shown up as new project commitments.
NOW WHAT: Ask whether a sponsor has reserved turbines, vessels and HVDC capacity before FID, because supply slots are now secured ahead of capital.
Vestas confirmed a firm order for 1,380 MW (92 V236-15.0 MW turbines) for Norfolk Vanguard West on 20 February 2026, ahead of FID. DESNZ's supply-chain study rates five components as high risk (turbines, floating foundations, HVDC and alternating-current HVAC stations, cable vessels and ports), reports "Severe HVDC shortfall in 2025–26", and finds only 3–5 steel mills worldwide able to supply heavy plate for monopiles. On grid, the National Energy System Operator (NESO) plans Gate 2 transmission offers through mid-January 2027, though the number of offers issued is not published. We found no vessel day rates, cable lead times or offshore-specific curtailment data. Our report on renewable energy grid bottlenecks in 2026 covers the wider grid constraint.
4.5 Floating Wind: Investable Before 2030?
FINDING: UK AR7 floating wind cleared at £216.49/MWh for only 0.1925 GW, about 2.4 times the £91.20/MWh fixed-bottom price (author arithmetic).
SO WHAT: Floating wind is a policy-supported niche today, and we found no verified independent global floating LCOE to test the price against.
NOW WHAT: Count a floating project as investable only if it holds a CfD-type contract; GWEC counts just 1.5 GW of floating capacity with active offtake.
GWEC reports 1.5 GW of floating capacity with active offtake and 5.9 GW committed through auctions, while TGS | 4C counts 5.5 GW of floating awarded in 2025. TGS | 4C expects about 11% of the 19,400 foundations installed in 2026–2040 to be floating.
5. Named Company Case Studies
DIRECT ANSWER
Four companies show where offshore wind capital sits. Ørsted raised about DKK 46bn from divestments and booked a DKK 1.2bn US impairment in Q2 2026. RWE's 6.9 GW AR7 win depends on a KKR sell-down. Vattenfall took the one named 2026 FID, the 1.6 GW Nordlicht project. The Mitsubishi-led consortium withdrew from Japan's Round 1 after winning at low bids. Each case is read from company disclosures and named reports, with figures kept in the reporting currency.
5.1 Ørsted: Balance-Sheet Repair, Rate-Driven Impairments, Asia-Pacific Retreat
FINDING: Ørsted raised about DKK 46bn from divestments (against a DKK 35bn target) and about DKK 60bn gross in a rights issue, cutting net interest-bearing debt to DKK 18.978bn at FY2025 from DKK 58.027bn at FY2024.
SO WHAT: Its position rests on asset rotation, not organic cash flow (inference), so US litigation or a further rate rise would press it again.
NOW WHAT: Watch the next divestment announcement and the next US impairment test, because both signal how much headroom remains.
The FY2025 results of 6 February 2026 showed earnings before interest, tax, depreciation and amortisation (EBITDA) of DKK 25.1bn, net profit of DKK 3.2bn and impairments of DKK 3.633bn. In H1 2026, EBITDA was DKK 15.0bn and impairments DKK 2.537bn, of which DKK 1.168bn fell in Q2. Net debt rose from DKK 18.978bn at FY2025 to DKK 21.96bn at H1 2026; the improvement is against DKK 67.137bn a year earlier. Apollo-managed funds bought 50% of Hornsea 3 (about 2.9 GW) for DKK 39bn of total consideration with DKK 20bn upfront, signed on 3 November 2025, and Hornsea 3 was about 30% complete. We use DKK figures throughout because dollar values for the rights issue conflict between sources.
In the US, Revolution Wind's final turbine, number 65, was installed on 18 September 2026, and Sunrise Wind was about 50% complete in Q2 2026. Outside the US, UK and EU, Ørsted installed all turbines at Greater Changhua 2b and 4 in Taiwan (920 MW) by 20 January 2026, with Cathay Life holding 50% of Greater Changhua 4 and acquiring 55% of Greater Changhua 2.
Per Mejnert Kristensen, CEO of Region APAC at Ørsted, said on 20 January 2026:
"With all wind turbines now installed, we've achieved an important milestone in Taiwan's offshore wind expansion."
Ørsted also bid Dadu 1 (840 MW) in Taiwan's Round 3.3. In South Korea, it suspended its bid for the 1.4 GW Incheon project in August 2026 because the project no longer met internal investment criteria (Ocean Energy Resources, 28 August 2026).
CHART PLACEHOLDER V3 — ØRSTED NET INTEREST-BEARING DEBT
Type: line (trend). Placement: Section 5.1. Data: DKK 58.027bn (FY2024); DKK 67.137bn (H1 2025); DKK 18.978bn (FY2025); DKK 21.96bn (H1 2026).
Caption: Debt rose between FY2025 and H1 2026. Source: Ørsted FY2025 announcement (6 Feb 2026) and H1 2026 announcement (13 Aug 2026).
5.2 RWE with KKR and Masdar: An AR7 Winner Not Yet at FID
FINDING: RWE won 6.9 GW of the 8.4 GW awarded in AR7 across five projects, and its 10 August 2026 report still says the KKR transaction is expected to close during 2026, with no FID confirmed on any of the five by 1 October 2026.
SO WHAT: RWE's FID is gated on a partner sell-down and a returns hurdle, and KKR's price is undisclosed, so the capital-recycling model behind 3 GW-scale UK projects cannot be priced from public sources.
NOW WHAT: Treat the KKR closing announcement as the trigger to re-check Norfolk Vanguard's status.
The five projects are Norfolk Vanguard East and West (3.1 GW combined, £91.20/MWh in 2024 prices,
KKR taking 50% in each, commissioning 2029 for West and 2030 for East), Dogger Bank South East and West (3.0 GW, RWE 51% and Masdar 49%, completion 2031–2032 per RWE's release) and Awel y Môr (about 0.8 GW, 2031; RWE 60%, Stadtwerke München 30%, Siemens 10%). Dogger Bank South received planning permission in May 2026. RWE's H1 2026 report records gross capex of €7,180m (€5,364m in H1 2025), a FY2026 net capex forecast raised to €9–11bn from €6–8bn, and leverage above 2025's 2.2x. It also records a settlement with the US Department of the Interior with a
"negative impact on earnings in the low to mid double-digit millions".
Markus Krebber, CEO of RWE AG, said on 14 January 2026:
"By combining KKR's investment know-how in large-scale, complex infrastructure projects with RWE's extensive offshore wind expertise, we are well positioned to jointly realise these major projects."
Vestas confirmed a firm 1,380 MW order for Norfolk Vanguard West on 20 February 2026, so supply is reserved before the money is committed.
5.3 Vattenfall: Nordlicht I and II, the One Named 2026 FID
FINDING: Vattenfall took FID on 13 January 2026 for Nordlicht I and II in German waters: 1.6 GW (980 MW and 630 MW), 112 Vestas V236-15.0 MW turbines, with both farms operating in 2028.
SO WHAT: If all of it sits inside TGS | 4C's 2.1 GW, only about 0.5 GW of other projects reached FID in nine months (author arithmetic), but TGS | 4C has not published its project list.
NOW WHAT: When comparing FID datasets, ask each publisher what "FID" and "financed" mean, because WindEurope's 0.6 GW financed in H1 2026 sits uneasily beside a 1.6 GW decision in January.
Monopile installation for Nordlicht I starts in Q3 2026. The investment amount and offtake terms were not disclosed in the sources. No other 2026 FID is named in public sources.
5.4 Mitsubishi-Led Consortium: Japan Round 1, Outside the US, UK and EU
FINDING: The Mitsubishi-led consortium won all three Round 1 sites in 2021 (1.7 GW per IEEFA; 1.76 GW per Baird Maritime) with bids of JPY 11.99–16.49/kWh against a JPY 29/kWh ceiling, withdrew in August 2025, and took a JPY 52.2bn impairment.
SO WHAT: Bids roughly 40–60% under the ceiling, followed by a reported doubling of costs, fit a winner's-curse pattern (inference), and Japan's new price floor targets exactly that.
NOW WHAT: When reading an auction result, compare the winning bid with the ceiling and check whether the contract indexes for cost inflation.
Rival bids ran at JPY 17–24.5/kWh, and total investment was above JPY 1 trillion. The government keeps JPY 20bn (USD 136m) of deposits and bars the consortium from future rounds. Reforms followed: January 2025 changes allow up to 40% cost inflation in the price and raise bid bonds for delays from JPY 13,000/kW to JPY 24,000/kW, while November 2025 measures give 20-year capacity revenue to zero-premium projects in Rounds 2 and 3. The Mitsubishi loss figures are taken from IEEFA, not from a company disclosure, and no source shows whether the Round 1 re-auction has launched or been awarded. Aegir Insights expects Round 4 to be "unlikely to launch before well into 2027".
5.5 What the Four Cases Show Together
FINDING: Two of the four sponsors, Ørsted and RWE, rely on selling equity to partners for 50% stakes in 2.9 GW and 3.1 GW projects respectively, while a third, Mitsubishi, left a market where it had won at 40–60% under the ceiling.
SO WHAT: Across the cases, capital follows protected revenue and a funded partner, and it leaves where low bids meet cost inflation without indexation (inference).
NOW WHAT: Apply the same two questions to any sponsor: who buys the equity, and how does the contract treat cost inflation?
Vattenfall's single FID shows the route works when a sponsor has both. This synthesis is analyst reasoning; it adds no new facts.
Company | Project or market | Verified figure | Date | Source |
Ørsted | Hornsea 3 sale; group balance sheet | DKK 39bn for 50%; about DKK 46bn divested; net debt DKK 21.96bn (H1 2026) | 3 Nov 2025; 13 Aug 2026 | Ørsted announcements |
RWE | AR7: five projects | 6.9 GW; Norfolk Vanguard £91.20/MWh; net capex forecast €9–11bn | 14 Jan 2026; 10 Aug 2026 | RWE press release; RWE H1 2026 report |
Vattenfall | Nordlicht I and II | 1.6 GW; 112 turbines; operation 2028 | 13 Jan 2026 | Vattenfall press release |
Mitsubishi-led consortium | Japan Round 1 | Bids JPY 11.99–16.49/kWh vs JPY 29 ceiling; impairment JPY 52.2bn | 2021; Aug 2025 | IEEFA |
T6. Case study summary. Mitsubishi loss figures are from IEEFA, not a company disclosure. Ørsted figures are in DKK because dollar rights-issue values conflict between sources.
6. Friction, Risk & Systemic Bottlenecks
DIRECT ANSWER
The biggest offshore wind investment risks between 2026 and 2030 are cost of capital, offtake and auction design, grid and supply-chain constraints, US policy, and an unexplained FID shortfall. Europe has started to correct auction design with two-way CfDs. The sources show no equivalent correction for borrowing costs, and the reason only 2.1 GW has reached FID in 2026 remains partly unknown. Curtailment risk is unquantified for offshore wind because the sources hold no offshore-specific data.
6.1 Risk Hierarchy by Materiality
FINDING: Ten risk categories carry sourced evidence, and three sit in the high band: policy and auction design, returns and cost of capital, and US political and legal risk.
SO WHAT: Grid, supply chain, developer balance sheets and permitting sit in a medium band, and curtailment is lower only because no offshore-specific data exists, not because it has been shown to be small.
NOW WHAT: Start diligence with the high-band risks, and for each ask whether the evidence is a primary document, a company disclosure or a press relay.
Rebecca Williams, Deputy CEO of the Global Wind Energy Council, wrote in GWEC's 9 June 2026 report release:
"The planning and grid connection process can be painfully slow and beset with risks for developers and investors."
In Europe, consenting is not the binding constraint: TGS | 4C forecasts 22.2 GW consented in 2026, the second-highest on record. Permitting bites harder in the US, India, South Africa and the Philippines. Curtailment and negative-price data cover all renewables and are solar-led: 3.9 TWh curtailed in 2025 across ten European countries (Montel via pv magazine, 3 February 2026) and 1,223 negative-price hours in the EU-27 in Q1 2026 against 593 a year earlier (Ricardo via pv magazine, 8 May 2026). We found no offshore-specific capture-rate data.
Band | Risk | Sourced evidence | Where it bites | Evidence quality |
High | Policy and auction design | Zero-bid rounds in Germany (Aug 2025), Netherlands (Oct 2025) and Denmark (Dec 2024); two bids in Taiwan Round 3.3; Philippines GEA-5 suspended 4 Jul 2026 | Europe (now being fixed by CfDs); Asia-Pacific | Regulator pages and press |
High | Returns and cost of capital | 9–11% LCOE per point; borrowing near 15-year highs; Ørsted DKK 1.2bn US impairment; IEA cut its five-year forecast by more than 25% | Global | Single-source publisher plus company disclosure |
High | US political and legal | Stop-work orders struck down, but $3.9bn of lease buyouts, a tax-credit deadline of 4 Jul 2026 and a pending Section 232 tariff decision; nine attorneys general and seven states have sued over buyouts | US | Press relay and court reporting |
Medium | Grid | NESO Gate 2 offers phased to January 2027; DESNZ "severe HVDC shortfall in 2025–26"; India's grid rests on PGCIL; Philippines grid agreements mandatory | UK, Germany, India, Philippines | Official study plus press |
Medium | Supply chain | Steel up 53%; Vestas offshore order intake 0 MW in Q2 2026; Siemens Gamesa orders down 78.5%; 3–5 mills supply heavy plate | Europe, US | Company disclosure; official study |
Medium | Developer balance sheets | Ørsted impairments DKK 3.633bn (FY2025) and DKK 2.537bn (H1 2026); RWE net capex forecast €9–11bn; SSE net debt £10.1bn | Europe, US | Company disclosure |
Medium | Permitting and consenting | GWEC: process "painfully slow"; Gagasi application withdrawn; Interior moved to revoke the approved plan for a Maryland project | US, India, South Africa, Philippines | Association release and press |
Lower | Curtailment and negative prices | 3.9 TWh curtailed; 1,223 negative-price hours; not offshore-specific | Europe (offshore effect unquantified) | Press relay of market data |
Lower | Turbine reliability and contractor disputes | Vineyard Wind complaint against GE Vernova claims $853m of costs from defective blades and $308m withheld; allegations, not findings | US | Press report of a court complaint |
Lower (counter-evidence) | Supplier recovery | Vestas EBIT margin before special items 9.4% in Q2 2026 (1.5% a year earlier); Siemens Gamesa profit before special items EUR 75m | Europe | Company disclosure |
T11. Risk hierarchy by materiality. Sources: Windtech International and TGS | 4C (22 September 2026); DESNZ supply chain readiness study (2026 update); Roll Call (30 September 2026); Vestas (12 August 2026); Siemens Energy (5 August 2026); Utility Dive (8 April 2026).
METHODOLOGY NOTE
The ordering is structured analyst judgement. Each risk was assessed on four criteria: how many markets it affects, how directly the evidence ties it to a missing or delayed FID, the quality of the evidence (primary document, company disclosure, or press relay), and whether policy can reverse it within the 2026–2030 window. Risks are placed in three bands (high, medium, lower), not scored, and the report states where the evidence is weakest.
6.2 The Unresolved Issue: Why Only 2.1 GW Has Reached FID
FINDING: TGS | 4C reports that only 2.1 GW has reached FID so far in 2026 against more than 12 GW due to enter operation, with 5.6 GW expected by year-end, and no second publisher reports a 2026 year-to-date figure.
SO WHAT: This is the only measure of new commitment we found, and its project list, FID definition and treatment of China are unpublished, so it shows direction better than it shows a precise count.
NOW WHAT: Use the 2.1 GW as a direction marker, and wait for a second source or the TGS | 4C project list before building a model on it.
What is verified
TGS | 4C's quarterly Global Market Overview, published on 21 September 2026, says only 2.1 GW has reached FID so far in 2026, and that by year-end 5.6 GW is expected, which would make 2026 the second-lowest year for offshore wind investment decisions since 2020. Ocean News and Windtech International repeat the figures, which does not make them two-source. The 7.1 GW "at FID" in the same report is a stock of FID-stage projects not yet in construction, which is a different measure from the 2.1 GW of 2026 decisions. Ivar Slengesol, Managing Director and VP of TGS | 4C, said on 21 September 2026: "general market uncertainties are making financing more challenging."
The pipeline that exists but has not converted
Against 22.2 GW of forecast consents and about 25 GW awaiting FID outside China (GWEC), the only 2026 FID named in public sources is Vattenfall's 1.6 GW Nordlicht. The UK awarded 8.4 GW on 14 January 2026, and RWE's Norfolk Vanguard (3.1 GW) was targeted for FID in summer 2026, but RWE's report of 10 August 2026 still shows the KKR deal as open. Windtech International reported on 22 September 2026 that developers remain cautious about committing capital to the next generation of projects. No source reconciles GWEC's 25 GW, or the 8.4 GW of awards, with TGS | 4C's 7.1 GW at FID.
Why capital is waiting: five sourced reasons
1. Returns do not clear internal hurdles at current rates. TGS | 4C cites 9–11% LCOE per percentage point, borrowing near 15-year highs and steel up 53%. Ørsted suspended its 1.4 GW Incheon bid in South Korea because the project no longer met internal investment criteria (Ocean Energy Resources, 28 August 2026). An Ørsted white paper (2025), developer-authored, puts the share of European capacity at risk at about 45% of capacity commissioned from 2029, which is one developer's view.
2. Offtake design. Germany, Denmark and the Netherlands saw no-subsidy or negative-bid rounds fail. The same white paper puts a CfD's effect at about 2 points of WACC. Europe is correcting this, with effect from 2027.
3. Partner sell-down comes before commitment. RWE pairs FID with KKR's closing, Ørsted funded itself with a rights issue and about DKK 46bn of divestments, and Vestas booked a firm 1,380 MW order before FID.
4. US policy has removed a large demand pool. Inside Climate News (17 August 2026) relays a BNEF cut to its US offshore forecast to 2035 from 39 GW to 5.9 GW and reports 12 offshore leases cancelled between March and August 2026.5. Supply chain and grid add cost and delay, though the sources quantify them poorly.
Inferences (analyst reasoning, not published forecasts)
Commissioning in 2026–2029 is largely supported by the 41.2 GW under construction and 7.1 GW at FID, so the FID shortfall bites later. Two timings are possible: a dip from about 2028–2029, consistent with Wood Mackenzie's qualitative 2028 warning, or a thinner 2029–2031 based on project timelines such as Nordlicht (FID January 2026, operation 2028). To reach 5.6 GW, about 3.5 GW of FID must close in Q4 2026 (author arithmetic: 5.6 less 2.1). AR7 projects are the likeliest candidates, but no AR7 FID was verified. The binding constraint is conversion, not consenting or awards. FID volume would be expected to rise after the AR7 partner closings due in H2 2026; this is inference, not a published forecast. GWEC's 24% compound growth for 2026–2030 depends on FID rising well above the 2026 level.
What is not known
We do not know the project list and FID definition behind 2.1 GW and 5.6 GW, whether China is included, how WindEurope's 0.6 GW financed in Europe in H1 2026 squares with a 1.6 GW FID in January, or whether Norfolk Vanguard, Dogger Bank South or Awel y Môr reached FID between summer and 30 September 2026. No independent 2027–2030 commissioning forecast built from the 2026 FID count exists in our sources.
6.3 Counter-Evidence and Conflicts Investors Should Weigh
FINDING: Vestas reported an earnings before interest and tax (EBIT) margin before special items of 9.4% in Q2 2026 against 1.5% a year earlier, and Siemens Gamesa returned to a profit before special items of EUR 75m, while TGS | 4C reports steel up 53%.
SO WHAT: Wood Mackenzie (November 2025) expected costs to stabilise in 2026 and TGS | 4C (September 2026) shows pressure, so cost direction is contested, and the later data point is TGS | 4C's.
NOW WHAT: Read supplier profit alongside order intake, since Vestas booked 0 MW of offshore orders in the same quarter.
Ben Backwell, CEO of GWEC, said on 14 January 2026:
"A strike price of £91 is 40% lower than the alternative cost of building and operating a new gas plant in the United Kingdom."
BNEF's benchmark, by contrast, shows offshore wind at $100/MWh against $102/MWh for gas. The two claims use different geographies, price bases and gas cost assumptions, and no source reconciles them, so treat GWEC's claim as an association claim, not a cost comparison.
Continue reading Part II:
