Last Call for REC Multipliers: Korea's Final Offshore Wind Auction

Five rounds, and what they added up to

Competitive fixed-price bidding for offshore wind opened in 2022. Five rounds later Korea has selected 18 projects totalling 5.89 GW. One of them did not reach a contract after its award, so 17 projects and 5.14 GW still hold one. Written as one number it reads like steady progress.

Written year by year it is not steady at all.

Cumulative 5.89 GW awarded across 18 projects. 5.14 GW across 17 projects still holds a contract. Capacity in MW beside each project name. Volumes are capacity awarded at each round, not capacity commissioned, and projects that failed to convert an award into a contract are not removed. Wando Geumil and Anma were each selected in two phases and are counted once. Firefly floating, awarded 750 MW in 2024, failed to conclude a certificate sale agreement twice and was barred from bidding for two years from January 2026. Source: KWI Master Database.

Awarded capacity by round. The 2026 first half is the orange bar.

The same five rounds seen by lead developer rather than by year.

Awarded capacity by lead developer, 2022 to 2026 first half, in MW. Each project's awarded capacity is assigned in full to one lead developer, so the bars add up to exactly the 5,891 MW awarded. Jeonnam P1 is entered at the 99 MW it was awarded, not the 96 MW installed, and under SK innovation E&S, which holds 51 percent to CIP's 49. Taean is led jointly by Vena Energy and CIP, which hold 37.5 percent each to KOWEPO's 25, and is shown as one bar. Southwest Phase 2 is developed by Korea Offshore Wind Power, jointly owned by KEPCO and the six generation companies, and is entered under KEPCO. Source: KWI Master Database.

CIP leads two awarded projects, Haesong 3 and Haewoori 2, totalling 1,036 MW. It also co-leads Taean with Vena Energy and holds 49 percent of Jeonnam P1, so it is named on 1,635 MW in all, 28 percent of everything awarded. Three of the eight fixed-bottom candidates for the second half are projects CIP sits in: Jeonnam P2, Jeonnam P3 and Haesong 1. State-owned utilities and public corporations lead 1,110 MW, through KOEN at Wando Geumil, KEPCO at Southwest Phase 2 and Jeju Energy at Handong Pyeongdae. Counting the stakes they hold without leading, a public entity is named on at least ten of the eighteen awarded projects and 3,237 MW, a little over half of everything awarded, counting only the shareholdings the KWI ledger records. KOMIPO is the public partner at Geumodo, which won the public-led track, and holds 29 percent of Gochang; KOSPO holds a major stake at Yawol. The public-led track turns on equity, not on who operates the project. A project qualifies when public bodies jointly hold more than half of it, or one holds 34 percent or more, and the track carries a preferential ceiling price. That is why KOMIPO sits in the Geumodo consortium alongside DL Energy.

The shape matters more than the total. A thin round is not a defeat for a developer who can wait, and on volume alone waiting has often been the rational move. Awarded volume has swung by more than a factor of two between adjacent years, the ceiling price is reset at every round, and the evaluation structure has been rewritten each time. KWI counted five rounds and five different structures in an earlier analysis. A project that judged the terms unfavourable could sit out and return into something different.

Waiting has only recently started to cost something, because the ceiling has only recently started to fall. It stood at 176.565 won per kilowatt hour in the first half of 2025. In the first half of 2026 it fell for the first time since the scheme began, to 171.229 won for fixed-bottom, 3.02 percent lower. The same notice gave floating a ceiling of its own for the first time, at 175.100 won. So a developer weighing whether to sit this round out is doing it in the first cycle after the price cap turned downward, and in the last cycle that carries the multiplier. Both arguments for waiting have gone at once.

Two of this year's winners show the same mechanism from the other side. Haesong 3 and Hanbit both failed in the 2025 first half and both were selected in 2026, each returning without shrinking its project. Nothing about the projects had to change. The round changed.

That option is what the second half removes, and the rest of this piece is about what happens to a market when the option to wait is taken away.

What the first half actually did

The 2026 first-half round offered about 1,800 MW split across three tracks. Nine projects bid, carrying 3,625 MW between them. Five were selected, totalling 1,786 MW.

Capacity in MW. Bid totals are as reported at the 12 May submission deadline. The results notice cites 3,656 MW of submissions, 31 MW above the May figure.

The headline figure of 2.0 to 1 describes none of the three markets it averages. Floating drew 1,467 MW against roughly 400 MW on offer and ran at 3.7 to 1. The public-led track drew 560 MW against roughly the same volume and ran at 1.4 to 1. A bidder's odds depended less on the quality of its project than on which of the three doors it walked through, and the volume behind each door is set by the notice rather than by the developer.

The public-led track is the one worth reading twice. Two bidders, 560 MW between them, against about 400 MW on offer. Geumodo took 160 MW of it. Roughly 240 MW of the track went unawarded, and Jangbogo, the other bidder, was not selected into any of it. That outcome is the single most informative thing the first half produced, and section 06 returns to it.

The project-level outcome is where the second half is already visible.

The four projects that were not selected are all expected back in the second half. Haewoori 3 bid 560 MW against a licensed capacity of 510 MW in the KWI ledger. Bid capacity and licensed capacity are different measures and they do not have to agree.

Four of the twelve names now circulating for the second half are returning bidders from the first. Haesong 1, Jangbogo, Haewoori 3 and East Blue Power were all in the room in May and all left without a contract. Between them they carried 1,839 MW into the first half, which is more than the entire volume awarded in the round they lost.

This is not a new market opening. It is a resit. The difference is that a candidate sitting it has one fewer attempt left than it had in June, and after this sitting the paper itself changes.