Trump's 5 GW reactor restart push collides with NRC's untested Part 57 rules
Washington's nuclear expansion targets 5 GW from old reactors, but the regulatory pathway for reopening them remains unproven.
President Trump's May 2025 executive order "Reinvigorating the Nuclear Industrial Base" set two targets: construction of 10 new large reactors and 5 GW of additional power from existing plants by 2030. The uranium ETF URA slipped 1.17% to $44.48 as of 2026-08-10, as investors continued weighing that ambition against the regulatory mechanics of actually reopening shuttered units.6
The 5 GW restart target is the harder half. Restarting a disused reactor is not a matter of flipping a switch; rigorous safety standards apply to every system that has sat idle, from coolant loops to control electronics. The NRC's traditional licensing framework was built for light-water reactors that never stopped operating, not for units coming back from years of cold shutdown.6
The NRC has been rewriting those rules. Part 53, the agency's risk-informed licensing pathway, pioneered a different approach: instead of mandating the same safety protocols required for conventional light-water reactors, it sets performance targets that developers are free to meet however they choose. The agency is now extending that logic through Part 57, a rule central to how new and revived reactors get licensed.2
Part 57's risk-informed philosophy cuts both ways for restart projects. In principle, it lets operators argue that a reactor's existing safety systems are adequate for its specific risk profile, avoiding the cost of retrofitting to standards designed for different designs. In practice, no developer has yet taken a shuttered US reactor through this pathway. The first application will tell the industry far more than the text of the rule itself.2
The economics of fuel supply add a separate layer of difficulty. Accident-tolerant fuels are gaining traction as a safer alternative to conventional uranium dioxide, but producers say the cost hurdle is demand, not technology. "It doesn't take a lot to make the economics work," one industry executive told finance.yahoo.com. "The problem has just been needing enough demand booked" to justify dedicated factory construction.3
That is the chicken-and-egg problem facing the broader buildout. Fuel fabricators will not build capacity without firm orders, and utilities hesitate to commit to fuel contracts for reactors that have not cleared licensing. The executive order's 10 new reactors and 5 GW of restarts are supposed to provide that demand signal, but the timeline to 2030 leaves little room for regulatory delays at any stage.3
International experience shows what can go wrong when external events stress nuclear safety systems. The UAE's Barakah plant passed what IAEA chief Rafael Grossi called a "test by fire" after a May 17 (2026-05-17) attack disrupted external power supplies, with safety systems activating automatically. "When you have good professionals and good procedures in place, immediately the nuclear power plant is a very safe place," Grossi said.4
But Barakah is a new plant operating under a single, coherent design. Restarting older US reactors means reviving systems designed decades ago and operating them under modern threat assessments, including cybersecurity. Lancaster University has unveiled a £2 million control room simulator to train operators on SMRs, pressurised water and fusion reactors, with a specific focus on cyber threats — a discipline that barely existed when most shuttered US units were designed.5
The UK is pursuing its own nuclear expansion, aiming to quadruple capacity to 24 GW by 2050. That target depends on SMRs arriving fast enough and cheaply enough to matter. "The honest question is — can they really be built fast enough, cheaply enough, and in large enough numbers," one analysis noted, a question that applies with equal force to US restart economics.1,5
For US restarts, the concrete signal to track is whether any operator files a Part 57 application for a shuttered reactor. URA at $44.48 as of 2026-08-10 prices in demand growth, but that level reflects policy expectations rather than confirmed regulatory progress. A Part 57 application that clears quickly would reset timelines for the 5 GW target; one that stalls pushes the 2030 deadline into doubt and puts uranium demand forecasts under pressure.6,2