Summary

• Nuclear Turbines has raised £15 million for a compact reactor and high-temperature gas turbine concept.
• The company targets industrial and datacentre sites seeking dedicated power behind the meter.
• Engineering validation, fuel, licensing, manufacturing, finance, and delivery time remain substantial barriers.

A British engineering startup has raised £15 million to develop a compact nuclear power system for industrial sites and datacentres, entering a market where access to electricity is becoming as important as access to computing hardware.

Nuclear Turbines, a Manchester-based company spun out of BAE Systems, has emerged from stealth with a reactor concept built around high-temperature turbine technology more commonly associated with aircraft engines and gas-fired generation.

The foundational round was led by IQ Capital, with participation from Rhapsody Venture Partners, Zero Carbon Capital, and Empirical Ventures. Funding will support detailed engineering, large-scale test rigs, recruitment, and preparation for manufacturing initial fuel elements.

Former BAE Systems principal engineer Jeremy Owston founded the company with University of Manchester professor Tim Abram. Its intended markets include factories, manufacturing facilities, microgrids, and datacentres requiring reliable electricity close to the point of consumption.

The turbine shapes the reactor

Most existing nuclear stations use reactor heat to produce steam, which then drives a turbine. Nuclear Turbines argues that steam systems require boilers, heat exchangers, cooling infrastructure, and high-pressure water equipment that are difficult to reduce in size without damaging the economics.

Its proposed design would instead heat compressed air for a high-temperature gas turbine, creating a more direct conversion system. Owston has described the approach as starting with the cheapest way to generate electricity and designing a reactor around it.

Beginning with mature turbine technology does not remove the difficult parts of nuclear development. The company must still validate reactor materials, heat transfer, fuel behaviour, containment, maintenance, control systems, and performance under fault conditions.

A system operating at temperatures suitable for efficient gas turbines also creates demanding materials and engineering requirements. Results will have to be demonstrated through physical testing rather than inferred from the established use of gas turbines in other industries.

Computing sites are looking beyond the grid

Datacentre demand gives the proposal a clear commercial setting because new facilities can wait years for grid connections, while AI clusters require large, continuous loads that may exceed the power available at preferred sites.

Operators are already examining on-site generation, long-term power agreements, battery systems, gas generation, renewable projects, and nuclear technologies. Dedicated supply can provide greater certainty, but it also turns a datacentre developer into a participant in energy infrastructure with additional financing, regulatory, and operational responsibilities.

A compact reactor cannot solve the immediate capacity shortage. Nuclear Turbines is targeting an initial system in the early 2030s, and that timetable depends on engineering progress, regulatory approval, fuel availability, manufacturing capacity, and customer finance.

Computing hardware also changes more quickly than nuclear infrastructure. A plant designed to operate for decades must remain useful through several generations of accelerators, cooling systems, and computing architecture, while customers need enough confidence in long-term power demand to finance the asset.

Behind-the-meter nuclear concentrates risk

Conventional nuclear projects are generally planned around utility-scale generation and national grids. Smaller deployments beside industrial consumers would create a more direct relationship between plant performance and the economics of a factory or computing campus.

Energy security may improve when supply sits beside the customer, although an outage at the dedicated plant could remove a large share of the site’s power. Grid backup, redundant generation, maintenance schedules, and refuelling plans must therefore be built into the commercial model.

Licensing will remain decisive. British regulators have experience with large reactors and are developing approaches for small modular designs, but a novel combination of reactor, fuel, materials, and turbine will still require extensive evidence.

Claims about low generation costs cannot be established until the design, construction method, financing, insurance, waste handling, and operating model become clearer. Nuclear projects frequently encounter costs outside the reactor itself, particularly when first-of-a-kind engineering enters construction.

Britain’s wider industrial capacity will also influence delivery because nuclear engineering, datacentre development, and sovereign computing ambitions all depend on specialist components and skilled workers. A credible design will require a manufacturing ecosystem as well as venture capital.

The £15 million round gives Nuclear Turbines enough capital to test its central engineering proposition, rather than finance a commercial fleet. Progress should therefore be assessed through regulator engagement, test results, manufacturing partnerships, and credible customer commitments.

Power constraints have created space for unconventional approaches to computing and industrial energy. Nuclear Turbines has identified a genuine problem, while its longer task is proving that the proposed system can leave the drawing board, pass nuclear scrutiny, and generate dependable electricity at a price customers can finance.