Singapore taps Indonesian hydrogen imports to feed data-centre turbines due in 2029
PacificLight's 670-MW hydrogen-ready plant needs cross-border fuel supply that lacks binding contracts, storage infrastructure or a proven delivery chain.
Singapore data centres are already importing Indonesian hydrogen, Bloomberg reported, turning a city-state with no land and limited renewables into an offtaker for a neighbour with both. The arrangement looks clean on paper. In practice, the fuel supply chain behind PacificLight's 670-MW hydrogen-ready plant, scheduled for completion in 2029, does not yet exist at the volumes a plant that size requires.2
The turbine technology is settled. Singapore's hydrogen-ready gas turbines are proven at 30% hydrogen compatibility, according to an Asian Power report published on 19 August 2026. The hurdles named by industry analysts are fuel supply, storage, land availability and the economics of higher hydrogen blends — none solved by a turbine specification.2
The 2029 PacificLight deadline is close. Cross-border hydrogen chains are not built quickly, and permitting risk on the Indonesian side — storage, land, grid connection — runs on a different schedule from turbine hall construction in Singapore.2
The hydrogen arbitrage sits inside a broader Southeast Asian power story that is mostly about LNG. Data-centre capacity across the region is forecast to reach 9.4 GW by 2035, and Wood Mackenzie expects that growth to lift average annual LNG demand growth by 16% over the same period. Singapore, Malaysia and Thailand are positioned to capture much of that demand, the consultancy said.4,6
For those markets, the fuel arrives on existing LNG infrastructure. For Singapore's hydrogen ambitions, new infrastructure is required on both sides of a national border. The LNG forecast is a projection against a market that already functions. The hydrogen import flow is a market being built from scratch.4,2
What makes the offtake side attractive is the counterparty profile. Md Fadhlullah Omarali, principal analyst at Wood Mackenzie, said data-centre demand is interesting from an LNG perspective because the buyers are large, creditworthy off-takers with power needs that remain stable regardless of economic cycles. That stability is the strongest commercial argument for the hydrogen-import model. But creditworthy offtakers do not remove the supply-side risk: fuel not yet contracted, storage not yet built, Indonesian permitting not yet resolved.3,5
Wood Mackenzie's verdict on generation technology reinforces why Singapore is pursuing gas turbines at all. CCGTs are the most viable technology for meeting data centres' 24/7 reliability requirements at scale, while grid-scale battery storage remains commercially immature across the region through the mid-term.4 Batteries are not a near-term substitute. Indonesian hydrogen feeding a Singapore turbine is, in that framing, the only credible path to low-carbon firm power for the data-centre load — which is why the 2029 date matters.
Grid capacity is the third constraint. Data centres, EVs and green industrial clusters are expected to add around 100 TWh of incremental power demand in Southeast Asia by 2030, but slower grid infrastructure development could delay the rollout, according to a Bain & Company and Standard Chartered report. Of roughly $540 billion in announced green capital expenditure across Southeast Asia's power and EV value chains to 2030, only around $315 billion is on a credible deployment path under current conditions.1
The same report found that de-bottlenecking grids and connecting Southeast Asian countries via cross-border electricity trade could unlock an additional $70 billion of clean energy and grid-related capital. Better interconnection cuts both ways for hydrogen: it reduces the need for molecule-by-molecule cross-border trade, but dedicated hydrogen pipelines and storage would compete for the same capital pool as grid spend.1
The incremental data-centre and industrial demand expected through 2030 is roughly triple the approximately 30 TWh of additional demand the region absorbed in the five years to 2025. Southeast Asia has 24 to 36 months to resolve its infrastructure gaps before that capital begins to move elsewhere, the Bain and Standard Chartered report said.1
The first binding supply contract for the PacificLight volumes will test the model. A contract that locks in Indonesian production validates the geographic arbitrage. One that leaves the source flexible suggests Singapore's hydrogen strategy is still searching for its counterpart at the other end of the pipe.2