Small Modular Reactors: Why the Technology That Should Have Worked Is Finally Working

NuScale's VOYGR cancellation was the headline. But behind it, a different class of SMR projects is making real progress. The technology that seemed permanently stuck is moving — and the reason is manufacturing economics, not physics.

## The NuScale Lesson

**NuScale** was the first SMR to receive NRC design approval — a genuine milestone. But the Carbon Free Power Project collapsed in November 2023: projected cost per megawatt-hour had risen from $58 to $119 over five years of development. The lesson wasn't that SMRs don't work. It was that the economics of the light-water reactor approach don't improve with smaller size.

A conventional large reactor benefits from economies of scale. When you make it smaller, you lose those economies without gaining sufficient offsetting advantages.

> ⚡ The VOYGR cancellation wasn't a failure of nuclear physics. It was a failure of assuming smaller = cheaper without changing the manufacturing paradigm.

## The Factory Manufacturing Thesis

The SMR projects gaining traction in 2026 share one core premise: factory fabrication beats site construction.

**GE Hitachi's BWRX-300** (300 MWe, boiling water reactor) is under active development with Fermi Energy in Finland and OPG in Canada. The design uses 90% fewer components than a conventional BWR — not by being simpler, but by eliminating redundant safety systems through passive safety design. Gravity, convection, and compressed gas replace pumps and valves. Fewer components mean less factory time and dramatically lower construction risk.

**Kairos Power's KP-FHR** uses fluoride salt as coolant instead of water — operating at high temperature and low pressure. The low-pressure operation eliminates the heavy pressure vessel that dominates conventional nuclear construction costs.

## The Data Center Demand Factor

The Microsoft-Constellation deal (September 2023) to restart Three Mile Island Unit 1 — a 835 MWe plant — was the clearest signal that tech companies will pay nuclear premiums for reliable, carbon-free baseload power. Data center power demand is growing 15–20% annually. Solar and wind cannot guarantee 24/7 supply at the scale required.

The **AP300** (Westinghouse, 300 MWe) targets this market directly: a scaled-down AP1000 with the same proven design. First commercial operation target: early 2030s.

> ⚡ At 300 MWe, a single SMR can power a hyperscale data center campus indefinitely. The math works for Microsoft, Google, and Amazon — even at $100/MWh.

**X-energy** (Xe-100, 80 MWe pebble bed HTGR) won an $800M DOE award under the Advanced Reactor Demonstration Program. Amazon Web Services signed a deal in 2023 to deploy multiple units. The pebble bed design uses TRISO fuel particles — each fuel kernel coated in ceramic layers that cannot melt, even in worst-case accident scenarios.

## The Regulatory Timeline Realities

Regulatory approval remains the long pole in the tent. The NRC's Part 52 process for new reactor designs runs 5–8 years. Canada's CNSC has shown faster timelines for some projects, which is why several SMR developers are targeting Canadian first-of-a-kind deployments before US.

The UK's Small Modular Reactor competition selected Rolls-Royce (470 MWe) as a frontrunner — a different scale than most "SMR" definitions but using factory-manufactured modules. The first UK unit targets 2035.

South Korea's APR+ and other established designs are pursuing SMR variants in the 300–500 MWe range, with export ambitions to Eastern Europe and Southeast Asia.

## The Bigger Picture

The SMR story is not about one design winning. It's about whether the combination of passive safety, factory fabrication, and growing demand from hyperscale computing can break the 50-year cycle of nuclear cost escalation. The projects still standing after 2023–2024 are the ones with credible answers to the manufacturing economics question. The timeline to commercial operation for leading designs is 2030–2035. That's not fast enough for the current data center buildout — but it represents real progress after decades of stagnation. Nuclear energy's second chance is being funded by the same companies that built the internet.

Small Modular Reactors: Why the Technology That Should Have Worked Is Finally Working

// COMMENTS

ON THIS PAGE