The Case for Small Modular Reactors: Capacity, Certainty and Portfolio Relevance

Source: International Atomic Energy Agency

Global electricity demand is entering a structurally faster phase. The International Energy Agency expects electricity use to grow more quickly through to 2030, driven by data centres, artificial intelligence, transport electrification and industrial activity. The Agency has also highlighted an unprecedented near-term requirement of roughly three thousand five hundred terawatt hours of additional generation by 2027, which is comparable to Japan’s annual consumption.

Policy is moving in the same direction. More than thirty countries have now endorsed a declaration to help triple global nuclear capacity by 2050, signalling a firmer financing and permitting environment for new nuclear build. Small modular reactors, which allow serial deployment and easier siting than traditional large reactors, are central to that agenda.

Where theory becomes investable reality

The investment case turns on evidence that modular nuclear is moving from design to delivery. The early proof points are concentrated, but they are real.

China is furthest along the deployment curve. The high temperature gas cooled reactor known as HTR PM has entered commercial operation following inherent safety demonstrations. In parallel, the ACP100, also known as Linglong One, completed cold functional testing in 2025 and is targeting commercial operation during 2026 at the Changjiang site on Hainan.

Russia continues to operate the Akademik Lomonosov floating plant for the remote Chukotka grid and recently passed 1B kilowatt hours generated, which shows that small reactors can support isolated systems at scale.

North America is derisking licensing and first builds. Canada has approved construction of the BWRX 300 at Darlington, the first grid scale small modular reactor in the region, with commercial service targeted for 2029. In the United States, NuScale holds Nuclear Regulatory Commission approvals for its 50 megawatt and its uprated 77 megawatt modules, which keeps the option open for clients to reference these designs in future combined licence applications. Advanced reactor licensing is also moving: Kairos Power received a construction permit for the Hermes test reactor in December 2023 and permits for Hermes 2 in November 2024, with safety related construction already underway.

In the United Kingdom, the national programme has narrowed the field to four viable designs and moved into final tendering, an inflection that will determine industrial participation and order books through the early 2030s.

The new buyer of firm power

A structural change is underway in the demand for firm, clean electricity. Hyperscale technology companies are contracting nuclear power directly, which accelerates early projects and tightens supply chains.

Microsoft and Constellation have agreed a 20-year power purchase arrangement that supports the restart of Three Mile Island Unit 1 by 2028, matching 835 megawatts of carbon free output to Microsoft’s data centre load in the PJM region.

Amazon, Energy Northwest and X energy are advancing a 320 megawatt Xe 100 project in Washington with an option to expand to 960 megawatts, supported by Amazon’s investment of about USD 500M in X energy.

Google and Kairos Power have agreed a framework that targets about five hundred megawatts of advanced nuclear by 2035, with the first unit expected by 2030 to provide round the clock carbon free power near Google’s data centres.

These contracts are long dated and binding, and they indicate that firm power is becoming a strategic input for the digital economy rather than a commodity purchased on a purely opportunistic basis.

What will constrain, and therefore price, the opportunity

Fuel availability

Many advanced designs require high assay low enriched uranium. The United States Department of Energy has begun to rebuild the domestic fuel cycle with a series of production scale task orders worth 2.7 billion US dollars, while Centrus Energy has extended its production contract through to June 2026. This remains a scarce and strategic input, and timelines will depend on enrichment and deconversion build out.

Licensing familiarity

Platforms that regulators already understand, such as the BWRX 300 and NuScale’s modules, enjoy a practical time to market advantage that matters for both procurement and finance.

Grid and balance of plant

Switchgear, transformers, interconnection works and thermal systems are capacity constrained, and the International Energy Agency’s 2026 report emphasises grid flexibility as a central challenge for the remainder of the decade. This supports earlier and more predictable cash flows for suppliers in these categories.

Uranium supply

The market remains tight. Kazatomprom plans to reduce output in 2026 by around ten per cent, equivalent to about five per cent of global primary supply, and Cameco has also flagged variability at key assets. Tightness in the fuel cycle is supportive for upstream pricing and is a cost line that developers need to hedge.

We favour three areas:

Fuel and enrichment

Exposure to enrichment and conversion capacity that benefits from policy support and supply scarcity. The Department of Energy’s multi year orders and the live Centrus production line provide clearer visibility than at any time in the last cycle.

Licensing ready platforms

Designs with proven regulatory progress and identifiable first sites, such as the BWRX 300 at Darlington and NuScale’s certified modules, where customers can reference approvals in their own applications.

Critical balance of plant and grid suppliers

Companies that deliver transformers, switchgear, interconnection works and thermal management, where orders precede reactor commissioning and cash conversion is earlier in the project cycle.

We maintain a measured diversifier.

Selective uranium and conversion exposure can hedge the structural tightness in the fuel cycle, sized for volatility and balanced against project execution risk.

Bottom line for portfolios

Small modular reactors will not replace large nuclear plants. They will complement them where grids are constrained and demand is growing fastest. The investment edge lies in owning the paid bottlenecks and the platforms with regulatory and customer traction, while using milestones to govern entry and size. The direction of travel is clear. The discipline is to allocate where capacity is scarce, cash flows arrive earlier and policy is already at work.