Denmark-based Ørsted has filed early environmental paperwork for Dadu II, a proposed offshore wind project of up to 2,002 MW roughly 38 kilometres off Changhua County in the Taiwan Strait, according to filings with Taiwan’s Ministry of Environment reported this week. The project would cover about 234.6 square kilometres of sea, use turbines rated between 14 MW and 22 MW, and connect into Taipower’s substation in Changhua. A public consultation meeting was held on 13 August, with the proposal now moving through Taiwan’s formal environmental review.
It is the second live proposal in the area from Ørsted Offshore Wind Taiwan alongside the roughly 924 MW Dadu I, and comes as the Danish developer is simultaneously finishing construction of the 920 MW Greater Changhua 2b and 4 project nearby, expected to reach full commercial operation in the third quarter of 2026. That project carries genuine significance beyond its size: its entire output is contracted for 20 years to Taiwanese chipmaker TSMC under what was, at signing, the world’s largest corporate offshore wind power purchase agreement.
Taiwan’s offshore wind programme has run competitive award rounds since 2018, giving developers multiple opportunities rather than a one-shot tender. Anchor demand from export-driven, decarbonisation-pressured manufacturers such as TSMC has proven willing to sign long-tenor, fixed-price offtake, de-risking the revenue side of projects in a way merchant or short-tenor contracts cannot. A story being repeated by data centre demand in other markets like the US and even India now.
Ørsted has also used partial divestments (Cathay Life Insurance bought 55% of Greater Changhua 2 in a deal worth roughly DKK 5 billion) to recycle capital and bring in institutional co-investors, part of over DKK 33 billion raised through its 2025–26 partnership programme. None of this has been friction-free, with Taiwan’s local-content rules now the subject of a World Trade Organization dispute with the EU, and the island still trails its own 2025 renewable-share target. But the fundamentals that get bids submitted are in place.
What India’s tenders got wrong
India’s Ministry of New and Renewable Energy cancelled both of its first offshore wind tenders in August 2025 — a 500 MW project in Gujarat’s Gulf of Khambhat and a seabed lease for up to 4,000 MW off Tamil Nadu by citing minimal developer interest. A post-mortem by climate think tank Ember pointed to structural causes rather than mere cost: the tendered capacity utilisation factor assumption of roughly 36% was well below international benchmarks and below Tamil Nadu’s own subsequently measured potential of 45–60%, while the auction offered developers no compensation if grid connectivity from India’s Central Transmission Utility was delayed, unlike European transmission operators such as TenneT, which absorb that risk themselves.
India is now resetting: a new viability-gap-funding scheme worth roughly ₹7,500 crore backs a first phase of 1 GW split evenly between Gujarat and Tamil Nadu, and a year of floating LiDAR data has confirmed Tamil Nadu’s capacity factors are competitive. Revised tenders, either two 500 MW auctions or a single 1 GW award, are expected in the second half of 2026, with Ørsted, RWE and Adani Green among those reportedly circling.
However, even after viability gap funding, the levelised cost of India’s first offshore auction is roughly ₹5/kWh above the fixed tariff on offer a gap the tender simply left uncovered.
The design gap, not just the subsidy gap
The comparison is not simply that Taiwan pays more or subsidises more generously. It is that repeated award rounds let a domestic supply chain and an operations-and-maintenance ecosystem , case in point being Ørsted’s O&M hub at the Port of Taichung, actually take shape, while fixed-price corporate offtake and transmission-risk-sharing remove exactly the uncertainties that sank India’s first attempt. As India’s revived tenders approach, matching Taiwan’s realism on capacity factors and its willingness to share transmission-delay risk may matter more to bid turnout than the size of the VGF cheque on offer.