SINGAPORE, 22 AUG 2026 — Singapore's Energy Market Authority granted conditional approvals for 900MW of electricity imports from Peninsular Malaysia, generated by solar and battery storage in Johor. Sembcorp Utilities holds approval for 300MW and Southern Solar Alliance, a subsidiary of the Malaysian developer Ditrolic Energy, for 600MW.
Commercial operation is expected around 2029. The state that will export the power is the same state where Malaysia's data centre construction is concentrated, and that is the part worth thinking about.
What was approved
Both developers must still obtain approvals in every relevant jurisdiction, conclude power purchase agreements with buyers, secure financing and complete project development milestones before reaching financial close. Singapore has now issued conditional approvals or conditional licences to thirteen import projects across Australia, Cambodia, Indonesia, Malaysia and Vietnam.
Conditional is the operative word, and thirteen is the number to hold
A conditional approval establishes that a project may proceed if everything else falls into place. It is a regulatory permission, not a contract, not financing and not a megawatt.
Thirteen projects across five countries is the more informative figure than 900MW, because it describes a pipeline where the binding constraints sit downstream of the approval. Cross-border power requires an exporting government willing to license the export, transmission capacity at both ends, a bankable offtake agreement and lenders comfortable with sovereign and currency risk across two jurisdictions. Any of those can stall a project that already has its Singapore approval.
A 2029 commercial operation date for an approval granted in 2026 is a three-year construction and financing runway, which is a reasonable estimate rather than a schedule. Judge this on financial close, not on the announcement.
Johor exports power and consumes it
The location makes this more than a simple procurement story.
Johor has absorbed the bulk of Malaysia's data centre growth, on the strength of cheap land, lower electricity tariffs than Singapore and physical proximity to Singapore's customers. Projections have put the state at around sixty per cent of national data centre capacity by 2030. Malaysia moved in February to restrict new non-AI data centre investment specifically because of electricity and water demand.
So the same state is simultaneously building loads that consume electricity at scale and contracting to export 900MW of generation across the strait. Those are not contradictory — new solar and storage is additional capacity built for export, not diverted from domestic supply — but they compete for the same land, the same grid connection queue and the same construction and financing capacity.
For any new energy project in Johor, the question is which side of the border its output will serve — and whether its grid connection comes at the expense of the other.
Why Singapore is buying rather than building
Singapore's constraint is not capital or willingness. It is land.
Domestic solar at scale is bounded by available surface area on an island of roughly 730 square kilometres with high building density and competing uses. Rooftops and reservoirs have been extensively deployed already. There is no configuration in which Singapore meets a material share of demand from domestic renewables.
Importing is therefore not a preference but the only route to renewable supply at scale. It converts a physical constraint into a set of commercial and diplomatic ones — a favourable trade for a country with strong credit and little land.
It also creates a dependency that did not previously exist. Electricity imported across a border is subject to the exporting country's policy, and export licensing is a policy instrument that governments retain the ability to adjust.
Solar plus storage is what makes this dispatchable enough to matter
Pairing solar with battery storage is a technical detail, but it carries most of the commercial weight.
Unfirmed solar delivers power when the sun is available, which is a poor match for a system that must meet demand continuously and has no hydro reservoir or large domestic gas storage to lean on. Storage converts an intermittent resource into something closer to a scheduled one, and it is what allows imported solar to be contracted rather than merely accepted when available.
A substantial share of the project cost sits in the batteries, making the economics sensitive to storage pricing between now and financial close. A project underwritten on 2026 storage costs and delivered in 2029 is exposed to that curve in both directions.
Imported electrons and where the emissions land
An import programme raises an accounting question that matters more than it sounds, because it determines whether this reduces emissions or relocates them.
Under standard territorial greenhouse gas accounting, emissions are attributed to the country where generation occurs. Importing solar-generated electricity therefore lowers Singapore's reported emissions by displacing domestic gas generation, and adds nothing to Malaysia's, because the generation is renewable. In this specific case the reduction is real on both sides of the ledger rather than shifted across the border.
That is a genuine distinction from the harder version of this argument. Importing power from a coal-heavy grid would relocate emissions while improving the importer's national figures, which is the criticism levelled at electricity trade generally. New solar and storage built for export does not have that property.
The remaining question is additionality: whether these Johor projects would have been built anyway to serve Malaysian demand, in which case the export contract redirects clean generation rather than creating it. Nothing in the announcement establishes that either way, and it is the question that separates a decarbonisation outcome from a decarbonisation transfer.
What this says about the ASEAN Power Grid
The regional grid has been discussed for decades and progresses through bilateral arrangements like this one rather than through the multilateral framework its name implies.
That is not a failure. A bilateral deal has two governments, one border and a single commercial structure to negotiate; a multilateral grid has all of those multiplied. Singapore's thirteen approvals across five countries are the practical form regional integration is taking, and each one that reaches financial close makes the next easier to underwrite.
A set of bilateral import contracts into one buyer is not a grid, however. It is a hub-and-spoke arrangement with Singapore at the centre, which delivers decarbonisation for the buyer without necessarily building the interconnection that would let Malaysia, Thailand and Vietnam trade with each other.
What remains unconfirmed
No power purchase agreements have been concluded and no financing is secured for either project. Malaysian export approval is not established as granted, and the projects require approvals from all relevant jurisdictions before financial close.
Neither the tariff nor the contracted price is public, the specific sites in Johor are not named in the available reporting, and the storage capacity accompanying each solar installation is not stated. The material reviewed does not disclose how many of the thirteen conditional approvals have reached financial close — the key figure for judging if the pipeline is converting permissions to projects.
What to watch for
The first milestone is Malaysian export licensing. Approval on the Singapore side is the easier half, and the exporting government's position is what has historically determined whether cross-border projects proceed.
The second is financial close on either project, which is the point at which 900MW stops being a permission. Watch for it well before 2029.
The third is Johor's grid connection queue. If domestic data centres face lengthening delays while export projects proceed, the politics of exporting power could change quickly.
View author profile → · Editorial policy
