Published: October 22, 2026 | Updated: October 22, 2026
POLICY & REGULATION
France closed bids on 12 October 2026 for AO10, its largest offshore wind tender: eleven projects and about 10 GW, roughly half of it floating. The headline figure is a €100/MWh target for the weighted average of winning bids, and it cannot be read against a Japanese bid price. In France, the transmission operator builds the grid connection and recovers it through network charges, about €20 billion for this round. The state also compensates output lost to shutdowns it orders, grid outages, and wake from neighbouring projects. In Japan, developers pay for their connection and absorb curtailment losses, and we found no mechanism for wake. Before using foreign auction results as a yardstick for Japan, check which risks sit inside the price.
👉 Japan’s Offshore Wind Policy & Regulatory Framework Explained
→
Execution Reality
→
Bankability Test
RTE, the French transmission operator, estimates AO10’s connections at about €20 billion, paid through network charges. A Japanese bid has to cover its own connection.
Shutdowns ordered for environmental or aviation reasons are compensated at 95%. Grid outages beyond set limits are compensated. Wake losses from designated neighbouring projects are compensated for up to 30 years. In Japan, curtailment is uncompensated, and we found no wake compensation in the auction rules.
From its next auction, Japan links the FIP base price to inflation. Grid cost, curtailment and wake from neighbours stay with the developer.
€100/MWh Covers the Wind Farm. About €20 Billion of Grid Sits Outside It
AO10 bundles eleven projects off Normandy, Brittany, the South Atlantic coast and the Mediterranean. Seven are floating, with a combined 4,500 to 5,250 MW. Four are fixed-bottom, with 5,200 to 5,450 MW. Bids closed at noon Paris time on 12 October. The government plans to award the projects in February 2027, after CRE has analysed the offers.
Each winner signs a 25-year contract for difference, a support contract that pays the gap between a fixed reference tariff and the market price. The tender sets no cap on any single project. The tariffs of the selected bids, weighted by each project’s minimum capacity, must average €100/MWh or less, in January 2024 prices indexed to the bid date. If the average comes out higher, CRE drops projects in a set order until the rest fit. CRE noted that such an average target had never been used in a French offshore tender, and warned it could make bid strategy harder to build.
In the larger group of eight projects, fixed-bottom and floating bids enter the same average. In our reading, that means low fixed-bottom tariffs at Fécamp leave room for higher floating tariffs elsewhere in the group. The floating projects do not each have to clear €100/MWh on their own.
| Group | Projects | Foundation | Capacity (MW) |
|---|---|---|---|
| PPE2 (extensions of earlier rounds) | A South Brittany, B Occitanie, C PACA | Floating | 1,300 to 1,650 |
| PPE3 | 1 to 3 Fécamp Grand Large | Fixed-bottom | 4,050 to 4,200 |
| PPE3 | 5 Oléron 1 | Fixed-bottom | 1,150 to 1,250 |
| PPE3 | 4 Brittany North-West, 6 and 7 Gulf of Lion Centre 1, 8 Gulf of Lion Centre 2 | Floating | 3,200 to 3,600 |
The grid is where the French and Japanese numbers part ways. RTE builds the connection for every AO10 project and recovers the cost through TURPE, the network tariff all French grid users pay. It puts this round’s connections at about €20 billion. CRE notes that connection could account for close to 40% of French offshore wind costs by 2040, and criticised a selection method whose €100/MWh average leaves those costs out.
In Japan, the connection is part of what the developer prices. In the fixed-bottom auction outside the Sea Area Act, connection cost is a separate step in the ceiling calculation. Under the Sea Area Act, where grid capacity has already been secured for a zone, the winner takes it over and pays an objectively calculated price for it.
Three Output Losses the French State Compensates
The AO10 rules go further than the grid. They move three sources of lost output from the developer to the state or to RTE.
Shutdowns ordered by the authorities. If an authority orders turbines stopped or slowed to protect birds or bats, or for aviation safety, the generator receives a premium. It equals 95% of the reference tariff applied to the output lost, estimated from the plant’s capacity factor that year (Article 5.5).
Grid outages. If RTE’s connection falls below its contractual availability rate, RTE compensates the generator under the connection contract. Planned outages caused by connecting another user to the same offshore substation are capped at 21 days over four years. Beyond that, the support payment is adjusted to cover the energy that could not be exported.
Wake from neighbours. A wind farm upwind slows the wind reaching the turbines behind it, the wake effect. AO10 names, for each project, the neighbouring projects that could later cause it. Once a named neighbour operates, the generator can claim a support adjustment for the lost wind resource (Article 5.2.5). Independent experts estimate the loss, and compensation can run for up to 30 years from the start of the support contract. For the zone next to Oléron 1, the state intends to charge part of the cost to the neighbour’s future owner.
Japan treats these risks differently. Since April 2021, solar and wind generators applying to connect anywhere in Japan must accept the “unlimited, uncompensated” curtailment rule: no cap on how much output can be cut and no payment for it, for FIT and FIP generators alike. The triggers differ. Japan’s rule covers cuts to balance supply and demand or to relieve congested lines, while France compensates ordered stops and failures of the grid itself. In both cases the generator did not cause the loss. On wake, we found no compensation mechanism in the January 2026 draft of Japan’s auction guideline, although two promotion zones lie along the same stretch of coast off Noshiro in Akita.
Lenders size debt on the downside energy case, often P90, the output a project should exceed in nine years out of ten. State-ordered stops, grid outages and neighbour wake all widen the gap between that case and the central estimate. Under AO10 each of them is compensated, so less of that gap falls on the debt. AO10 also resets the reference tariff when interest rates are fixed, in both directions (Article 5.7), which moves rate risk between bid and financial close away from the sponsor. We found no Japanese equivalent, so these risks stay in the sponsor’s case. For the same project, a Japanese bidder needs either a higher price or less debt.
25-Year Contracts and 5 GW of Floating: What France’s Regulator Warned About
The French design has critics, and the most detailed is CRE itself. It opposed extending the support contract from 20 to 25 years: the change shifts market price risk to the state, and for floating projects, with less certain operating conditions beyond 20 years, the financing benefit is less evident.
Scale is the second concern. CRE called the number of projects and the capacity in one procedure unprecedented, and warned that competition on each project could weaken and costs rise. AO7, the earlier tender for Oléron, received no bids, and AO8 drew very few candidates. CRE also wants the procedure redesigned before AO11, either with the state obtaining permits and handing them to winners, as in the Netherlands, or with auctions for projects already authorised, as in the United Kingdom.
Where Japan Has Moved Toward France, and Where It Has Not
Japan’s auction design has been changing since Round 1. In Rounds 2 and 3, a single bid at the zero-premium level, ¥3/kWh, forced every other bidder down to the same level to win. The new design METI set out in January 2026 adds an expected supply price range below the ceiling and, for now, gives a bid at the ceiling 100 of 120 price points. A price adjustment scheme links the FIP base price to inflation from the next auction, and Japan’s centralized method already provides state-led site surveys.
| Item | France AO10 | Japan, Sea Area Act auctions |
|---|---|---|
| Inflation | Target indexed to the bid date; tariff indexed before the contract starts and partly each year during it | FIP base price linked to inflation from the next auction |
| Interest rates | Tariff reset when rates are fixed (Art. 5.7) | Not addressed in the documents we reviewed |
| Site data | State-run wind, geophysical, geotechnical and environmental studies | State seabed survey after a zone becomes promising |
| Grid connection cost | Built by RTE, paid through network charges | Carried by the developer |
| Curtailment and ordered stops | Ordered stops compensated at 95%; grid outages compensated | Unlimited, uncompensated curtailment rule |
| Wake from neighbours | Compensated for designated neighbours, up to 30 years | No mechanism found in the January 2026 draft |
Japan has moved where the risk is cost inflation, which Round 1’s withdrawals made impossible to ignore. It has moved much less on lost output and the cost of reaching the grid. For floating wind the gap matters more. Japan aims to form more than 15 GW of floating projects by 2040, and floating sites lie further out and deeper, so connection cost and energy yield risk both tend to rise.
An auction price records how much risk the state has already taken off the table.
AO10’s tariffs, due in February 2027, will be quoted in Japan as evidence of what floating wind costs. Read them with the contract attached. A French floating tariff excludes the grid, is averaged with fixed-bottom bids, and is backed by compensation for ordered stops, grid outages, wake and interest rate moves. As far as we can find, a Japanese bid carries these risks itself.
For developers, the useful comparison is risk by risk, and the headline price says little on its own. Consider using AO10’s clauses as a checklist when pricing a Japanese floating bid: each item France compensates must be covered in Japan with contingency, insurance or equity. Exercise caution with reading the French result as a price Japan should already be able to match. France buys part of its lower price with risks now held by its budget and grid users, and CRE has questioned whether that trade is worth it.
Related DeepWind Articles
- Japan’s Marine Renewable Energy Act: how zones are designated and operators selected
- Japan’s fixed-bottom auction outside the Sea Area Act and its cost gap
- Where Japan’s offshore wind faces curtailment in FY2031
- Floating offshore wind in Japan: a structural guide to markets, costs and policy
DeepWind Weekly tracks Japan’s offshore wind market beyond headline announcements, focusing on execution risk, cost structure, project viability, supply-chain constraints, and policy implications.
Subscribe to receive weekly intelligence on Japan’s offshore wind market.
Why Isn't Floating Wind Bankable in Japan Yet?
The synchronized supply-chain bottleneck, quantified node by node, with a full sensitivity ranking and a bankability ladder of scaling conditions. Includes Simulator Professional (β) access to reproduce every figure on your own assumptions.
See the full analysis →