Why the Nuclear Power Revival Also Depends on Building a Reliable Uranium Supply

Nuclear power is back in the energy conversation, but the discussion often begins with reactors rather than the fuel they will eventually need. Electricity demand is also rising in new ways. The International Energy Agency reported that data-centre electricity use grew 17% in 2025, while consumption at AI-focused data centres rose even faster. Its latest outlook still sees total data-centre electricity demand roughly doubling by 2030.

That matters because nuclear expansion creates a parallel question about uranium supply. The OECD Nuclear Energy Agency and International Atomic Energy Agency reported in September 2026 that 418 commercial reactors operating at the start of 2025 required about 64,500 tonnes of uranium annually. Global mine production reached 61,924 tonnes in 2024, its highest level since 2016, but still below annual reactor requirements.

The same report does not suggest the world is running out of uranium. Identified recoverable resources exceed 8.1 million tonnes. The challenge is turning those resources into dependable supply at the right time. The NEA and IAEA estimate annual reactor-related requirements could rise to between about 84,800 and 143,900 tonnes by 2050, depending on how quickly nuclear capacity expands.

Mining timelines make that distinction especially important. Natural Resources Canada notes that mineral exploration can take seven to ten years, followed by another five to ten years of development. Uranium projects also move through environmental and nuclear-safety reviews. In Canada, the Canadian Nuclear Safety Commission describes separate licensing stages for site preparation and construction, operation, decommissioning, and eventual abandonment. These processes are designed around safety and environmental oversight, but they also mean future supply cannot be created instantly.

Michael Collins, CEO and director of Generation Uranium Inc., believes this timing issue deserves more attention in discussions about nuclear growth. Generation Uranium operates at the exploration end of the supply chain, where discoveries can precede production by many years. "There are a number of deposits that have been identified, but it takes a long time to permit them and bring them into production," Collins says.

That distinction becomes more relevant as new sources of electricity demand develop. Collins points to AI processing and data centres as one factor that could support additional nuclear generation. The IEA expects nuclear power to play a larger role in supplying data centres toward the end of this decade and beyond, with the first small modular reactors in its base-case outlook coming online around 2030. Collins sees that possibility as another reason to consider fuel availability well before new capacity is operating.

Small modular reactors, or SMRs, could add to future uranium requirements if deployment scales. Their eventual effect will depend on how many projects are completed, when they enter service, and which reactor designs are used. Collins believes the market can focus too heavily on resources in the ground without separating them from production that is actually available. "There are enough resources to fill the current deficit," he says.

There are signs that the uranium sector is responding. The 2026 Red Book reports that global uranium exploration and mine-development spending exceeded $1.78 billion across 2023 and 2024, about 46% more than in the previous two-year period. The same report says the production increase came mainly from restarted capacity and higher output at existing mines. It also notes that no new uranium mining projects began production during the period, even as several advanced through approvals and development.

For Collins, that reinforces the case for treating exploration as part of long-term energy planning rather than as a response to an immediate shortage. His view is that supply preparation has to begin before the market feels the full effect of new demand. Discoveries, technical studies, permitting, financing, construction, and commissioning all take time, and not every identified resource will become a producing mine.

The broader evidence points to a simple planning issue. Reactor approvals, construction schedules, financing, and grid demand will remain central to the nuclear debate, but uranium availability belongs in the same conversation. The more useful measure of preparedness may be how much material can be responsibly developed and delivered when utilities need it. As nuclear ambitions grow, the question is no longer only how many reactors can be built. It is also where their uranium will come from.