Keppel’s Indonesian solar project looks like a renewable-energy story. But the more important investment angle may be what happens after the electricity reaches Singapore: Keppel could be building an integrated power-and-data-centre ecosystem that turns electricity availability into a competitive advantage.
Keppel’s latest energy plans contain a deceptively simple idea.
Build solar power in Indonesia.
Connect it to Singapore through a subsea cable.
Bring the electricity into Jurong Island.
Use existing generation infrastructure.
And potentially supply the rapidly expanding data-centre ecosystem next door.
On the surface, this is another Singapore clean-energy project.
For Keppel investors, however, the more interesting question is whether the company is gradually assembling something much more valuable:
an integrated platform linking low-carbon power generation, electricity imports, flexible gas generation, renewable fuels and AI-ready data centres.
That distinction matters.
Singapore’s biggest constraint in expanding data-centre capacity is not simply finding land or attracting customers.
It is power.
And as AI dramatically increases electricity requirements per data-centre facility, securing reliable, scalable and increasingly low-carbon power could become a competitive moat.
Keppel is positioning itself across both sides of that equation.
The real asset may be the connection, not the solar farm
Keppel and a private fund are developing up to 2 GW of solar capacity in Indonesia’s Riau Islands, integrated with battery storage, together with a subsea transmission cable to Singapore.
Keppel has previously received a conditional licence to import 300 MW of low-carbon electricity from Indonesia.
The important development is where that electricity is intended to land.
The cable will connect to the Keppel Merlimau Cogen plant on Jurong Island, which sits next to Keppel’s planned data-centre development.
That creates an unusual strategic combination:
Indonesia solar → battery storage → subsea cable → Jurong Island power infrastructure → data centres.
Each individual component is difficult enough.
Putting them together could be considerably more valuable.
Why?
Because the value of electricity to a data-centre operator is not determined only by its price.
It also depends on:
- reliability;
- availability;
- scalability;
- carbon intensity;
- regulatory compliance;
- connection infrastructure;
- speed of deployment.
Keppel potentially has exposure to all of these.
AI is changing the economics of Singapore’s power market
This is where the data-centre strategy becomes particularly important.
Singapore recently awarded 200 MW of additional data-centre capacity to four operators, including Keppel.
Keppel plans to invest more than S$1 billion in an AI-ready data centre on Jurong Island.
That investment is significant.
But the real constraint is what happens after construction.
An AI-ready data centre requires enormous amounts of electricity.
The industry therefore faces a fundamental bottleneck:
You can build the building faster than you can create new electricity supply.
Singapore has limited land and limited domestic renewable-energy resources.
That makes electricity imports increasingly important.
Singapore aims to import around 6 GW of low-carbon electricity by 2035.
Keppel is one of six companies that have received conditional licences to collectively import 3 GW from Indonesia.
This creates an important strategic advantage for companies already developing cross-border power infrastructure.
Keppel is not merely a potential consumer of Singapore’s energy transition.
It is trying to become part of the infrastructure that enables it.
The hidden Keppel advantage: monetising the same power multiple ways
This could become one of the most interesting aspects of the strategy.
Suppose Keppel brings low-carbon electricity into Singapore.
It does not necessarily have to make money only from selling electrons.
It can potentially create value across several layers.
Layer 1: Power generation
Solar assets can generate electricity and potentially produce long-duration contracted cash flows.
Layer 2: Transmission
The subsea cable becomes strategic infrastructure connecting Indonesian generation with Singapore demand.
Layer 3: Power optimisation
Battery storage can help manage intermittency and potentially improve the value of imported renewable power.
Layer 4: Generation flexibility
KMC provides existing power-generation infrastructure at the Singapore landing point.
Layer 5: Data centres
Keppel can potentially sell data-centre capacity to customers that increasingly value reliable low-carbon power.
This is the key difference between owning a solar farm and owning an integrated energy-and-digital infrastructure platform.
The data centre could be the highest-value customer
This is where investors should pay close attention.
Keppel has been transforming itself from a traditional conglomerate into a more focused infrastructure and asset-management business.
Data centres are central to that strategy.
But data-centre economics depend heavily on power.
If Keppel can offer customers a combination of:
land + data-centre infrastructure + reliable electricity + renewable-energy pathways + potentially lower-carbon power
it may be able to compete for customers on more than simply rent or facility specifications.
That could become increasingly valuable as hyperscalers and AI companies face pressure to manage the environmental footprint of their rapidly expanding computing infrastructure.
In other words, Keppel’s renewable-energy investments could indirectly increase the attractiveness of its data-centre assets.
That is the second-order effect the market should be watching.
Biomethane adds another piece to the puzzle
Keppel is also investigating imports of biomethane derived from palm-oil waste, including palm-oil mill effluent from Malaysia and Indonesia.
This might initially appear unrelated to solar power.
It isn’t.
Biomethane gives Keppel another potential pathway for producing lower-carbon electricity using existing gas infrastructure.
That could become valuable because Singapore’s energy transition is unlikely to move directly from natural gas to 100% renewable electricity.
The transition is more likely to involve a mixture of:
- solar;
- imported renewable electricity;
- batteries;
- natural gas;
- biomethane;
- hydrogen;
- other emerging technologies.
For an infrastructure operator, flexibility across multiple energy pathways can itself be an asset.
Keppel’s existing power plants potentially give it an advantage because new fuels can be integrated into infrastructure that already exists.
Hydrogen is the longer-term option — but investors should be patient
Keppel Sakra Cogen is another important component.
The 600 MW plant is Singapore’s first combined-cycle power plant designed to be compatible with hydrogen and can initially co-combust up to 30% hydrogen with natural gas.
But there is an important caveat.
The plant has not started using hydrogen.
And that matters.
Hydrogen economics remain challenging because low-carbon hydrogen production, transportation, storage and distribution are expensive.
Keppel itself has indicated that wider adoption depends on achieving sufficient economies of scale.
Investors should therefore treat hydrogen as an option on future energy infrastructure, not as a current earnings driver.
This distinction is important.
There is a temptation to attach a large future valuation to every “hydrogen-ready” asset.
The more disciplined approach is to assign value only when:
- supply becomes commercially available;
- costs become competitive;
- customers are willing to pay;
- regulation supports deployment.
Until then, hydrogen readiness provides strategic flexibility rather than immediate earnings.
The real investment thesis is vertical integration
Keppel’s strategy starts to make more sense when viewed as a chain rather than a collection of projects.
Energy supply
Indonesia solar and other renewable sources.
↓
Energy transportation
Subsea cables and cross-border electricity infrastructure.
↓
Energy management
Battery storage and existing generation capacity.
↓
Lower-carbon fuels
Biomethane and eventually hydrogen.
↓
Digital infrastructure
AI-ready data centres.
↓
Asset management
Potentially recycling mature infrastructure assets into funds and investment vehicles.
That final step is particularly important given Keppel’s broader business model.
If Keppel can develop infrastructure, stabilise cash flows and then recycle assets into private funds or other investment vehicles, it potentially gets paid twice:
once through development economics and again through recurring management or platform economics.
That is a much more interesting long-term business model than simply owning power plants.
But there is a major risk: regulation
The biggest risk may not be technology.
It may be government policy.
Cross-border electricity projects involving Singapore and Indonesia require multiple layers of regulatory approval.
Indonesia’s requirement for clean-energy export permits to be renewed every five years has already been identified as a concern for potential investors and customers.
That creates a financing problem.
A data centre is typically a long-duration infrastructure investment.
Investors want visibility over decades.
A five-year regulatory renewal cycle introduces uncertainty into the underlying power supply.
This could affect:
- financing costs;
- contract duration;
- customer commitments;
- project valuations;
- required returns.
The economics of the solar project therefore cannot be assessed purely by looking at solar-generation costs.
Regulatory durability is part of the project’s economic value.
Another risk: Keppel could overbuild the ecosystem
There is also a capital-allocation question.
Solar farms, subsea cables, batteries, power plants and data centres all require significant capital.
Keppel’s transformation is partly about becoming an asset manager and infrastructure owner rather than simply accumulating capital-intensive assets.
That means investors should ask:
How much capital does Keppel actually need to deploy itself?
and:
How much can it develop and subsequently recycle into third-party funds?
The answer will determine whether the strategy produces attractive returns on equity or simply creates a larger collection of infrastructure assets.
For investors, return on capital matters more than headline asset growth.
The market may be underestimating the value of power access
This may ultimately be the most important point.
The conventional way of analysing data centres is to focus on:
- occupancy;
- rental rates;
- contracted capacity;
- hyperscaler customers;
- development pipeline.
But as AI increases electricity demand, another variable becomes increasingly important:
power availability.
A data-centre operator that has secured land but cannot obtain sufficient electricity has limited economic value.
Conversely, an operator that controls or has privileged access to reliable power can potentially unlock scarce data-centre capacity.
That means the economic moat may migrate from:
real estate → power infrastructure.
Keppel’s Jurong Island strategy puts it unusually close to both.
That is potentially more significant than the individual solar project.
Bull case: Keppel becomes an integrated AI infrastructure platform
The optimistic scenario is relatively straightforward.
Singapore continues expanding data-centre capacity.
AI increases electricity demand.
Low-carbon power becomes increasingly valuable.
Indonesia becomes an important source of renewable electricity.
Cross-border transmission infrastructure develops successfully.
Keppel’s solar, cable and power infrastructure becomes operational.
Its Jurong Island data-centre capacity attracts high-quality customers.
And Keppel ultimately recycles some mature assets into funds while retaining attractive management economics.
Under that scenario, Keppel could evolve into something considerably different from its historical identity.
Not simply:
a conglomerate with power assets and data centres.
But:
an integrated infrastructure platform connecting capital, energy and digital demand.
Bear case: attractive projects, mediocre returns
The downside scenario is more subtle.
The projects could all work technically while still producing disappointing shareholder returns.
Why?
Because infrastructure is capital intensive.
If competition for data-centre customers pushes up development costs, or if renewable-power projects require excessive capital, returns can be diluted.
Regulatory uncertainty could increase financing costs.
Hydrogen could remain uneconomic for longer than expected.
Power-import projects could face delays.
And data-centre demand could ultimately grow more slowly than today’s AI investment assumptions imply.
In that scenario, Keppel would still own useful infrastructure.
But useful infrastructure does not automatically mean high shareholder returns.
What investors should watch over the next 12–24 months
Rather than simply tracking Keppel’s renewable-energy announcements, investors should focus on five measurable developments.
1. Data-centre customer commitments
Who actually signs contracts for the Jurong Island capacity?
Customer quality and contract economics matter more than the headline MW figure.
2. Power-import approvals
Watch whether the Indonesia project progresses from conditional approvals toward final regulatory and commercial commitments.
3. Project capital requirements
Investors need visibility on how much equity Keppel itself contributes versus third-party capital.
4. Asset recycling
Can Keppel successfully transfer mature infrastructure into private funds while retaining attractive recurring management economics?
5. Returns on capital
Ultimately, the question is whether the strategy raises sustainable returns — not simply whether Keppel’s infrastructure portfolio becomes larger.
Investment conclusion: Keppel is increasingly selling access to scarce infrastructure
The most interesting part of Keppel’s Indonesia solar strategy is not the solar panels.
It is the connection between scarce power and scarce digital infrastructure.
AI is creating enormous demand for data-centre capacity.
Singapore has limited domestic energy resources.
Indonesia has significant renewable-energy potential.
Keppel sits between those two markets with existing power infrastructure, cross-border energy ambitions and a growing data-centre business.
That creates a potentially powerful strategic position.
But investors should distinguish strategic optionality from current earnings.
The Indonesian solar project, biomethane and hydrogen initiatives could become important components of Keppel’s future infrastructure platform, but regulatory approvals, project economics, capital requirements and customer contracts will determine whether that optionality ultimately becomes shareholder value.
Over the next 12 to 24 months, the critical question is therefore not simply whether Keppel announces more renewable-energy projects.
It is whether the company can demonstrate that power infrastructure is increasing the economic value of its data-centre platform while generating attractive returns on the capital deployed.
If it can, the market may eventually need to value Keppel less like a traditional conglomerate and more like an integrated owner and manager of energy and digital infrastructure.
That is the potentially transformative part of this story.