Why EQT Is Backing Contracted, Fast-to-Energise Power Assets

Developed-market electricity demand is growing for the first time in more than twenty years. Over the same period, coal’s share of global generation is slipping below one-third. Together, these two signals change which power assets deserve capital, and the shift goes well beyond a simple story about AI needing more power.

EQT, the Swedish alternative asset manager listed on Nasdaq Stockholm, has built a distinct energy transition strategy around this convergence. Its roots run back about 150 years to the Wallenberg family, long-time core shareholders in the business that became ABB. That company built the first commercial high-voltage direct current link in 1954.

The view outlined here comes from a presentation by an EQT speaker, so treat it as EQT’s framing rather than settled fact. For Mining & Energy investors, though, institutional capital flows are a useful signal of where power infrastructure returns are expected to land.

Here is a working framework for reading where that capital is heading, and the points where the thesis could break.

Why is electricity demand rising after two decades of stagnation?

The global numbers surprised even the forecasters. In January 2024, the International Energy Agency (IEA) expected electricity consumption to grow 2.5% that year. By July it had lifted that forecast to 4%. The final figure came in higher again.

The demand signal Global electricity consumption rose 4.3% in 2024, the largest absolute increase on record, according to the IEA.

The EQT speaker links this to data centres, fleet electrification and heavier air conditioning use as temperatures rise. The speaker also framed it as a security issue, claiming 75% of the global population faces energy security risk because it depends on imported energy rather than domestic generation.

The IEA’s upward revisions point to a wider pattern in which grid bottlenecks, rather than generation alone, increasingly decide where new load can connect and where capital can earn a return.

The NEM as a local test case

Australia’s National Electricity Market (NEM) shows the turn clearly. The EQT speaker described a decade of flat consumption, at roughly 170-176 TWh a year recently. In June 2026, AEMO’s Integrated System Plan (ISP) Step Change scenario projected underlying consumption nearly doubling from about 205 TWh to around 390 TWh by 2050. The speaker expects the NEM to reach about 250 TWh by the mid-2030s.

The two baselines measure different things. The lower figure is probably grid-supplied power, while AEMO’s figure counts underlying consumption, which includes energy generated on site, such as rooftop solar. Neither source states this explicitly.

Data centres are about 2% of grid-supplied NEM use today. AEMO projects about 25% annual growth, which would take them to almost 10% of underlying demand by 2050 (about 34 TWh), alongside roughly 35 TWh from hydrogen. The speaker’s separate claim of a 13-13.5% data-centre share by the mid-2030s exceeds AEMO’s 2050 figure. That gap likely reflects a different metric, such as capacity rather than consumption, or a different scenario, so the two numbers should not be read as one trajectory.

When a flat decade gives way to a near-doubling forecast, the riskiest input in any power asset model becomes the grid planning assumption built on historical load. AEMO frames the data-centre growth as a scenario, so you should hold it with matching caution.

What is replacing coal in the generation mix?

If demand is rising, something has to supply it. The IEA’s Electricity 2025 report expects coal’s share of global generation to fall below 33% for the first time this century. The EQT speaker frames this as a reliability and policy story: ageing coal plants are failing the reliability needs of new loads, and US policy has been volatile.

The falling share does not mean falling volumes everywhere, because the coal consumption paradox shows absolute use staying high in some developing markets even as coal’s proportion of global generation declines.

The replacement is a stack of technologies rather than a single one. Each plays a different role:

• Solar: supplies large volumes of cheap energy, but only when the sun shines.
• Batteries: store that energy and release it later, smoothing out solar’s variability.
• Gas: provides firm capacity, meaning power that can be dispatched on demand regardless of weather.

Solar and batteries

According to the IEA, solar generated about 2,000 TWh in 2024, around 7% of global electricity, up from 5% in 2023 (these figures have not been independently confirmed). The EQT speaker cited 600 TWh of solar added in a recent year, the largest single-year addition by any technology on record.

Batteries are the fastest-growing power technology, according to the speaker, and Australia is a particularly strong market for both solar and storage. AEMO’s Step Change scenario reportedly calls for around 40 GW of storage and about 120 GW of grid-scale wind and solar to 2050, although these figures are unverified.

Why gas is still in demand

Firm grids still lean on gas, and the constraint is equipment rather than appetite. The EQT speaker says gas turbines are sold out for the next decade.

A sold-out turbine market means an investor with secured equipment and contracted offtake holds a scarce asset. An investor relying on future turbine access carries execution risk that may not show up in the headline thesis.

How does EQT decide where to put power capital?

With demand rising and firm supply constrained, EQT’s screening logic follows directly. The speaker framed the opportunity in large numbers.

EQT’s capital estimate Capital investment in EQT’s focus transition subsectors is projected to rise from about US$100 billion to about US$425 billion, roughly 20% annual growth, according to the EQT speaker. No independent sourcing for this estimate was found.

The four criteria

1. Speed to power: businesses that deliver firm power to customers faster than the grid can gain a structural advantage.
2. 15-20 year take-or-pay contracts: long revenue agreements give the asset infrastructure-like characteristics.
3. Proven development track record: experienced builders can sign offtakers earlier, which lowers development risk.
4. Avoiding undifferentiated merchant exposure: capacity selling into wholesale markets carries cyclical revenue.

Speed matters because new transmission has long lead times. On-site assets such as gas engines, fuel cells and solar-plus-storage can often be permitted, financed and built in months rather than years.

A take-or-pay contract obliges the customer to pay for contracted capacity or energy whether or not it uses it. A merchant asset sells into spot or short-term markets.

Scale Microgrids as the test case

In January 2025, EQT Transition Infrastructure acquired Scale Microgrids, its first North American investment. Scale develops, owns and operates microgrids (small local power systems that can run independently of the main grid). At acquisition it had about 250 MW operating or under construction and a 2.5 GW near-term pipeline.

The same on-site logic is already visible in microgrids in mining operations, where remote sites cannot wait for grid connections and pair solar, storage and backup generation to secure reliable power.

Scale combines solar, batteries, gas generators, fuel cells and combined heat and power for customers including data centres and EV fleets. It builds on site, which delivers speed, and contracts output over long periods, which delivers revenue certainty. Its delivery record supports early contracting, and its contracted model sidesteps merchant risk. EQT maps integration and growth over three to five years.

The structure is mainstream. Brookfield, Blackstone and KKR broadly position around long-duration contracted power, although no specific recent transactions were retrieved. If you are assessing any power infrastructure name, listed or private, ask two questions: is revenue contracted for 15-20 years, and can the asset be energised before the grid can reach it?

Where could the thesis break?

The same features that make the model appealing also concentrate risk. Five principal risks stand out:

1. Demand overstatement: AEMO treats 25% data-centre growth as a scenario with wide uncertainty bands, and earlier EV load scares did not fully materialise in the near term. Revenue built on peak forecasts may arrive later than modelled.
2. Counterparty concentration: take-or-pay removes volume risk but ties returns to a few offtakers. Credit quality becomes the key variable.
3. Stranded gas: the IEA’s World Energy Outlook 2024 reportedly discusses under-utilisation risk for fossil assets in accelerated-transition scenarios (unverified). Merchant gas is most exposed, while contracted back-up roles are less so.
4. Policy volatility: regulators are revisiting connection rules, capacity markets and planning frameworks. Tariffs or tighter climate policy could alter microgrid economics.
5. Equipment and system integration: turbine sell-outs reward developers who hold equipment but add execution risk. Heavy behind-the-meter reliance could also fragment the system and complicate resource-adequacy planning.

The concentration point A contract is only as strong as the customer signing it. If AI demand disappoints, renegotiation and refinancing risk remain even under take-or-pay terms.

For you as an investor, the larger risk is less that demand disappears and more that contracted cashflows depend on a handful of counterparties whose own AI spending is not guaranteed. The speaker’s reframing of sustainability as security, consistent with the 75% energy-insecurity figure, suggests demand may persist even if the political language around it changes.

Readers interested in the sustainability-as-security argument can explore our detailed coverage of energy security capital allocation, which shows how security concerns now steer record investment decisions.

What EQT’s approach signals, and what it does not

The argument holds together in three parts. Demand is rising structurally, the generation mix is shifting toward solar, batteries and firm gas, and capital is favouring contracted, fast-to-energise assets over merchant exposure.

Three questions travel well to any power infrastructure asset. Is revenue contracted long term? How fast can it deliver power? Who is the counterparty?

This remains one manager’s framing from a presentation, and the data-centre projections are scenarios rather than forecasts. For Australian readers, AEMO’s ISP updates and turbine supply are the leading indicators to watch.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results. Financial projections are subject to market conditions and various risk factors, and forward-looking statements are speculative and subject to change.