Britain’s Fusion Gamble Is No Longer Just About Electricity

Britain has spent decades promising that nuclear fusion will change the energy system, usually on a timetable just distant enough to avoid immediate accountability. Now a British company is attempting something more audacious: using fusion not to power the national grid first, but to move spacecraft through the solar system. In March, Pulsar Fusion produced and confined plasma inside the exhaust architecture of its proposed Sunbird propulsion system in Bletchley, a small but real engineering milestone that places Britain in a new contest over the practical uses of fusion.¹

The achievement does not mean Britain has built a fusion rocket. It does not mean astronauts are about to reach Mars in a matter of weeks. It means an early test has moved an idea from diagrams and promotional films into hardware. That distinction matters. Fusion has acquired a fashionable vocabulary—revolution, limitless energy, the holy grail—while the difficult questions remain stubbornly ordinary: how to contain the plasma, protect the machine, obtain the fuel, manufacture the magnets, regulate the reactor and pay for years of failure.

Those questions are now converging in Britain. The government has committed more than £2.5 billion to fusion research and development over five years, while the private sector is building companies around magnets, robotics, power systems and space propulsion. The country’s bet is not simply that one reactor will work. It is that the industrial capabilities developed around fusion will create a technological base that can serve energy, medicine, manufacturing and space.

The test was modest, but it was not imaginary

The March demonstration took place at Pulsar Fusion’s facility in Bletchley and was shown during a technical session at Amazon’s MARS conference in California, an event associated with Jeff Bezos and the overlap between artificial intelligence, robotics and space development. The company’s Mark I Sunbird exhaust test system generated plasma using krypton, then used electric and magnetic fields to guide charged particles through the exhaust channel.¹

That description is less cinematic than “a fusion rocket has been lit”, but it is more accurate. Krypton was a test propellant. The experiment demonstrated initial plasma confinement and the physical arrangement needed to study how particles move through a prospective exhaust system. It did not demonstrate fusion reactions, net energy production or sustained thrust. No spacecraft was propelled. No crew-rated engine exists.

Yet the difference between a simulation and a working plasma is not cosmetic. Plasma is an electrically conducting gas that must be created, controlled and kept away from the walls of the machine. If it strikes those walls, energy is lost and the components are exposed to extreme heat and particle bombardment. A reactor or engine must keep the plasma in the right place while extracting useful energy from it. That is a control problem before it is a fuel problem.

Pulsar says the next stages will involve detailed measurements of thrust and exhaust velocity, further magnetic upgrades, radio-frequency and rotating-magnetic-field heating, and studies of how neutron radiation could affect reactor walls and magnets. The company has also described a plan to move towards rare-earth high-temperature superconducting magnets and, later, experiments with more demanding fusion fuel cycles.¹

Each step could expose a failure that the previous step could not. A plasma can be confined for a short test without being stable enough for an engine. An engine can generate thrust without producing enough power to justify its mass. A reactor can achieve a promising plasma condition without surviving the repeated thermal and radiation loads of a commercial system. First plasma is therefore a beginning, not a verdict.

That is precisely why the test is important. The fusion industry has been damaged by its own language. Every new experiment is presented as a breakthrough, even when it closes only one small part of a chain containing hundreds of unsolved problems. The proper significance of the Bletchley test is narrower and stronger: a British team has demonstrated part of the physical architecture required for fusion-based space propulsion, and it now has to show that the architecture can perform useful work.

Why put fusion in a spacecraft?

The attraction of fusion propulsion is not that it sounds advanced. It is that conventional propulsion forces spacecraft to choose between power and efficiency. Chemical rockets can produce enormous thrust, which is why they remain indispensable for leaving Earth and making rapid manoeuvres. But they consume propellant at a prodigious rate. Electric propulsion uses propellant far more efficiently and can achieve high exhaust velocities, but its thrust is so low that acceleration takes months or years.

Fusion is intended to occupy the uncomfortable middle ground: high exhaust velocity without abandoning useful thrust. In theory, a fusion system can heat and accelerate particles to very high speeds while producing a sustained push. That would change the economics of deep-space transport. A spacecraft would not need to carry all the energy required for its journey in chemical propellant launched from Earth. It could reach orbit using an ordinary launch vehicle, dock with a reusable transfer vehicle and use the fusion system for the longer journey.

Pulsar’s proposed Sunbird is a dual-output machine. Its Dual Direct Fusion Drive concept is designed to provide propulsion and electrical power rather than treating the reactor as an engine alone. The company says the system could deliver specific impulse in the range of 10,000 to 15,000 seconds and as much as 2 megawatts of power to a payload.² Those are design targets and modelling claims, not flight results, but they reveal the intended market: not a faster launch from Earth, but a reusable orbital tug capable of moving equipment and spacecraft between destinations.

The distinction is central. Fusion propulsion would not replace the launch industry at the bottom of the gravity well. It would change what happens after launch. The rocket that carries a payload to low Earth orbit would no longer need to perform every manoeuvre required to send that payload to Mars, Jupiter or an asteroid. A specialised tug stationed in orbit could handle the interplanetary leg.

That model has advantages beyond speed. A power-rich transfer vehicle could support high-bandwidth communications, refrigeration, scientific instruments and industrial equipment far from the Sun. Solar power becomes less attractive as a spacecraft travels into the outer solar system, where sunlight is weaker. A reactor that provides both thrust and electricity could allow a probe to carry instruments that would otherwise be limited by its power budget.

It could also make missions more modular. Instead of designing a new propulsion system for every probe, a space agency or commercial operator could hire a tug. Cargo could be assembled in orbit, attached to the transfer vehicle and sent onwards. The same machine could return, be serviced and fly again. That is the logic behind reusable launchers translated into the interplanetary environment.

But the logic depends on the tug being reliable, serviceable and cheaper than the alternatives. A machine that can cross the solar system but requires a new reactor after every flight is not a transport system. A vehicle that produces megawatts but weighs too much to launch is not a transport system. The commercial proposition will be determined by mass, maintenance, fuel supply, shielding and operating life—not by the elegance of the plasma physics.

Britain’s real advantage is the work nobody sees

The most persuasive case for Britain’s fusion programme is not that British scientists have discovered a shortcut around physics. It is that Britain has concentrated on the supporting systems that ambitious fusion announcements often leave in the background.

Fusion requires powerful magnets, heat-resistant materials, remote maintenance, fuel handling, specialist manufacturing and software capable of controlling an unstable plasma. A reactor is not a single invention. It is a collection of industrial problems that must all be solved at once. Britain has accumulated expertise in many of those areas through decades of work at the UK Atomic Energy Authority and the Culham Centre for Fusion Energy.

The country is now trying to turn that expertise into an industrial ecosystem. Culham has become a base for public research and private companies, while the government’s STEP programme is intended to build a prototype fusion power plant at West Burton in Nottinghamshire, on the site of a former coal-fired station. The official target is operation around 2040, with the plant expected to demonstrate net energy, fuel self-sufficiency and a workable route to maintenance rather than merely produce another laboratory result.³

The timetable is not thrilling. That is one reason it is more credible than the usual claim that commercial fusion is only a decade away. A prototype power plant must be designed, permitted, built, operated and repaired. It must prove that it can breed or obtain enough fuel, remove heat, replace damaged components and deliver electricity without turning every maintenance cycle into a bespoke research project.

Researchers working on STEP have described a design aiming for at least 100 megawatts of net electrical output and self-sufficiency in tritium production. The details are not decorative. Tritium is scarce, radioactive and difficult to handle. Any serious power plant must create more of it than the fusion reaction consumes, or else depend on an external supply that cannot support a global fleet of reactors.⁴

That is why the supposedly dull parts of fusion matter. A machine that cannot be maintained is a museum piece. A machine that cannot breed its fuel is a demonstration. A machine that requires exotic components to be hand-built by a handful of specialists is not an industry. Britain’s opportunity lies in showing that the technology can be engineered, manufactured and repeated.

The same reasoning applies to Pulsar. A fusion rocket may attract attention because it promises a shorter journey to Mars, but its credibility will be judged by the quality of its test stands, magnets, diagnostics and supply chain. The first company to build a plausible fusion engine will not necessarily be the company that wins. The winner may be the one that can make ten engines, inspect them, repair them and operate them under an agreed safety regime.

Magnets may be the bridge from science to industry

Among the technologies emerging from Britain’s fusion programme, high-temperature superconducting magnets are the clearest example of a capability with uses far beyond one reactor design.

Superconductors carry electrical current with very low resistance when cooled below a critical temperature. In fusion machines, they are used to create the powerful magnetic fields needed to contain plasma. High-temperature superconductors do not operate at room temperature—Tokamak Energy’s magnets still require cryogenic conditions—but they can operate at temperatures less demanding than older systems and can support stronger fields in more compact designs.

That matters because the size of a fusion machine is strongly affected by the strength and geometry of its magnets. Stronger fields can allow a smaller device to achieve useful plasma conditions. Smaller machines may be faster and cheaper to build, although compactness does not remove the need for shielding, heat removal or maintenance.

Tokamak Energy, founded as a spin-out from the UK Atomic Energy Authority, has developed a business around both fusion and high-temperature superconducting technology. The company says it has more than 300 employees, more than 300 live patents and $335 million in investment. In 2026 it announced a £70 million contract to develop the magnet system for the UK’s STEP prototype plant through 2029.⁵

Its acquisition of Ridgway Machines is revealing. Ridgway is not a glamorous start-up; it is a long-established Leicester engineering business specialising in machinery for taping, winding and insulating electrical systems. That kind of equipment is precisely what a country needs if it intends to manufacture superconducting cables and magnets at scale. The acquisition joins scientific design to industrial production.

The phrase “high-temperature superconductor” can mislead the public because the operating temperatures remain far below freezing. The comparison is not with ordinary weather. It is with the extreme cooling burden of older superconducting systems. Every reduction in cooling complexity can affect cost, reliability, size and maintenance. In a fusion plant, those gains may determine whether a magnet survives repeated operation. In a spacecraft, they may determine whether a propulsion system can be made compact enough to launch.

There are terrestrial markets too. Superconducting systems could improve electricity distribution, medical imaging, motors, generators and industrial equipment. The commercial value of fusion research may therefore arrive in fragments long before a power plant supplies a national grid. A company that earns revenue from magnets or manufacturing services can continue developing fusion without waiting for the final reactor.

This is the sensible version of the fusion story. The technology does not have to leap from laboratory to limitless electricity in one move. It can create businesses along the way. That reduces the financial risk, but it also changes the political question. Public investment is not only a subsidy for a distant reactor; it is a decision about whether Britain wants to own the supply chains that will support a future energy and space industry.

The fuel problem is fusion’s inconvenient secret

Fusion is often described as a way to turn seawater into energy. That slogan is not false, but it conceals the difference between a fuel that exists in nature and a fuel that can be delivered to a machine in the required quantity, purity and form.

For terrestrial fusion power, the leading fuel cycle has been deuterium and tritium. Deuterium is found in seawater. Tritium is much rarer and radioactive. A future power plant must breed tritium from lithium inside a blanket surrounding the fusion chamber, then recover and recycle it. That is a major engineering challenge, not a minor detail.

Space propulsion introduces a different temptation. Pulsar’s long-term concept points towards aneutronic fuel cycles involving deuterium and helium-3. The attraction is clear: the reaction can produce charged particles that may be directed by magnetic fields, while reducing the damaging neutron output associated with deuterium-tritium fusion. That could make direct conversion of fusion energy into thrust more attractive and reduce some forms of reactor damage.

But helium-3 is not sitting in warehouses waiting for a rocket industry. It is scarce on Earth and expensive to obtain. Lunar mining is often proposed as the answer because the solar wind has deposited helium-3 in the Moon’s surface material over geological time. The proposal is physically interesting and economically unproven. There is no mature lunar mining operation, no established market price for industrial quantities and no infrastructure for transporting the material from the Moon to an orbital fuel depot.

The danger is that the fuel question gets pushed into the future because it spoils the promotional narrative. A design may work on paper with helium-3 while remaining commercially impossible until someone builds the mining and transport system. That does not make the design worthless. It means the timetable must be tied to a supply-chain plan, not only to a plasma milestone.

There are possible interim strategies. A propulsion system could begin with easier-to-source fuels for testing or use a different reaction while the technology matures. A space tug might first operate in cislunar space, where missions are shorter and servicing is more feasible, before attempting voyages to Mars or the outer planets. The early market may be moving cargo around the Moon rather than carrying crews across the solar system.

That would be a less dramatic story, but a better business. Transport between low Earth orbit, lunar orbit and future stations could create repeated demand. Repeated demand creates operating data, which creates engineering confidence. Engineering confidence attracts capital. A system that earns its way outward is more credible than one that begins with a promise to revolutionise Mars.

Britain is funding a race, not a guarantee

The government’s £2.5 billion commitment is large by British research standards, but it should not be mistaken for a guarantee that fusion power will arrive on schedule. It is a wager that the country can buy a strategic position in a technology whose commercial value could be enormous if the central engineering problems are solved.

The stated public case is broader than electricity. The government presents STEP as a means of creating thousands of jobs, reviving industrial sites and building a domestic supply chain. UK Fusion Energy is expected to work with engineering and construction partners, while the wider programme supports facilities, skills and research. The policy is an attempt to turn scientific leadership into industrial leverage.

That approach is more realistic than pretending that Britain can outspend every country pursuing fusion. The United States has deep venture capital markets and major private companies. China can mobilise state resources at scale. Europe has ITER, the international project in France, whose complexity shows how difficult it is to assemble a fusion machine involving many nations and millions of components. ITER’s own planning has included data systems capable of handling at least 50 gigabytes per second during full deuterium-tritium operation, a reminder that a fusion facility is also a vast computing and information-management project.⁶

Britain’s advantage, if it has one, is speed and concentration. It can connect a national laboratory, universities, start-ups, manufacturers and government procurement more directly than a multinational project. It can use public money to create demand for components, then allow private firms to sell the resulting expertise elsewhere. But concentration creates its own risk. If the flagship programme slips, costs rise or technical assumptions fail, the entire national strategy is exposed.

The procurement model will matter. STEP has shortlisted industrial groups for engineering and construction partnerships, with the stated aim of creating a public-private integrated team and a British supply chain.⁷ Such partnerships can spread risk and bring industrial discipline into a scientific programme. They can also bury responsibility under layers of contractors if targets are unclear.

A serious programme therefore needs public milestones that are more demanding than headlines. It must publish cost and schedule performance, explain changes in design, report failures without treating them as public-relations disasters and separate company forecasts from independently validated results. Fusion will not be made credible by repeating the word “breakthrough”. It will be made credible by showing that each machine performs better than the last one and that the gains survive outside the laboratory.

There is a political temptation to describe fusion as a clean substitute for fossil fuels and a source of national renewal at the same time. Those claims may eventually be compatible, but they are not automatic. Industrial renewal requires training, procurement and patient capital. Clean energy requires a power plant that can run, maintain itself and compete with alternatives. The government has funded the attempt; it has not bought the outcome.

The space business may arrive before the power station

One of the most interesting features of Britain’s fusion push is that the first commercially useful product may not be electricity from a reactor. It may be a component or service derived from fusion research.

That possibility is particularly strong in space propulsion. A rocket engine does not need to solve every problem faced by a grid-connected fusion plant. It may not need to breed its own tritium. It may operate for a shorter mission. It can accept a specialised fuel and a high price if the performance advantage is large enough. A space customer may pay for speed, payload capacity and power in a way a national electricity market will not.

This does not make the space version easy. It makes the commercial test different. A terrestrial reactor must produce electricity at a cost that competes with established generation. A space propulsion system can command a premium if it reduces launch mass, shortens a mission or enables an otherwise impossible scientific payload. That premium could fund further development.

Pulsar has described Sunbird as a transfer vehicle that would wait in orbit while conventional launchers deliver payloads to it. The company’s own mission concepts include cargo transport to Mars, outer-planet probes, lunar supply, asteroid missions and deep-space telescope deployment. Its published modelling includes a target of moving a roughly 1,000-kilogram spacecraft to Pluto in four years, but such figures should be read as design studies rather than schedules.⁸

The first customers, if the system ever reaches operation, are more likely to be governments and large institutions than ordinary commercial operators. Space agencies buy reliability and scientific capability before they buy low prices. A nuclear-powered tug could supply power to a probe after it reaches a weak-sunlight environment. It could support radar, communications and instruments that would otherwise be restricted by solar generation. It could also reduce the time during which a crew or payload is exposed to the hazards of deep space.

Those benefits must be weighed against the cost of launching and operating a nuclear system. Every reactor placed in orbit requires safety analysis, launch approval, public confidence and an international framework for accidents and end-of-life disposal. The reactor may be harmless while cold and unstarted, but regulators will still examine the launch vehicle, failure modes, radiation shielding and what happens if the spacecraft loses control.

A credible market could therefore begin with uncrewed missions. A tug could move cargo or scientific equipment before anyone asks it to carry people. That would allow engineers to measure degradation, validate maintenance and test the propulsion system in the environment where it is meant to work. The space industry has learned, sometimes painfully, that flight heritage is worth more than a persuasive animation.

Britain has an opportunity here because its space sector does not need to build the entire launch stack. It can specialise in propulsion, power systems, test infrastructure and mission services, then work with international launch providers. That is a more plausible ambition than attempting to replicate every capability of the United States or China.

Fusion’s old problem is now a credibility problem

Fusion research has endured because the underlying prize is immense. A successful fusion plant could provide low-carbon electricity from relatively abundant fuels, with different waste and safety characteristics from conventional fission. But the field has also cultivated a habit of turning every intermediate result into a prophecy.

There is a difference between scientific optimism and commercial certainty. A scientist can reasonably say that a new magnet improves the odds of achieving a useful plasma. An investor can reasonably fund a company on that basis. A headline that promises cheap electricity or rapid transport before the engineering chain is complete is making a different claim.

Britain’s current programme should be judged by whether it confronts that distinction. STEP has a target around 2040, not next year. Its published design work addresses net electricity and tritium self-sufficiency, not just temperature. Tokamak Energy is building a business around magnets and manufacturing as well as reactors. Pulsar’s public test announcement identified the result as an early development step and described the next experiments needed to measure performance. Those are signs of a programme moving from slogans towards evidence, although company statements still need independent verification.

The wider private sector has changed the financing landscape. The Fusion Industry Association was created to represent private fusion companies and argued that public-private partnerships and regulatory certainty would be necessary to move from research to demonstration.⁹ The industry now contains multiple technical approaches: tokamaks, stellarators, inertial systems, field-reversed configurations and other designs. Competition may accelerate progress, but it also makes comparisons difficult. A company can claim an advantage by choosing a metric that favours its own design.

The same problem appears in fusion propulsion. Specific impulse is important, but it is not the only measure. Thrust-to-power ratio matters. Reactor mass matters. Shielding matters. Fuel availability matters. The lifetime of magnets and first-wall materials matters. A drive with extraordinary exhaust velocity may still lose to a less spectacular system that can be launched, operated and serviced.

Journalists and politicians should resist both easy enthusiasm and easy dismissal. Fusion is not a fantasy in the sense that the physical reaction is unknown; scientists have created fusion plasmas for decades. Nor is it a near-term product simply because a plasma has appeared inside a new machine. The honest position is more demanding: the physics is established, the engineering is unfinished and the economics remain unproved.

The next milestone will be less glamorous and more decisive

For Pulsar, the next meaningful question is not whether another presentation can make Sunbird look like a spacecraft from the future. It is whether the company can produce a set of measurements that engineers outside the project accept as useful and repeatable.

That means measuring thrust, exhaust velocity, plasma density, temperature, stability and power consumption. It means showing how the system behaves as the magnetic field is increased, how the components respond to heat and how performance changes over time. It means identifying the point at which the machine stops working and explaining why. A test that finds a limit can be more valuable than a test that produces a flattering headline.

The company has stated that later experiments will add stronger magnets, more sophisticated heating systems and a dedicated thrust balance. A future in-orbit demonstration has been discussed. If it occurs, the result will have to answer questions that no ground test can settle: how the system survives launch vibration, how it operates in vacuum, how it rejects heat, how it controls its attitude and how it handles a fault when there is no workshop nearby.

Britain’s energy programme faces a parallel test. STEP must translate design studies into procurement, construction and operation. The public will need to see not only a plasma but an industrial project that can control costs and schedules. The supply chain must develop enough skilled workers to build specialist components, and the regulator must provide rules that protect the public without making innovation impossible.

These efforts are connected. A country that learns to make superconducting magnets, remote-handling systems, heat-resistant materials and high-power electronics for a fusion plant is also building capabilities useful to space propulsion. A company that learns to test a fusion exhaust system may create knowledge relevant to terrestrial plasma devices. The relationship will not be automatic, but it is real.

The danger is fragmentation. If public funding is spread across too many projects without a route to hardware, Britain may produce impressive research and little industrial power. If funding is concentrated on one design and that design fails, the country may lose a generation. The answer is not to eliminate risk. It is to make risk visible, preserve competing approaches where they are genuinely different and demand evidence at each stage.

Fusion will eventually be judged in the least romantic places: the maintenance bay, the procurement contract, the insurance policy, the electricity meter and the launch approval file. That is where Britain’s claims will either become durable or collapse.

What Britain is really trying to build

The popular version of the fusion race is a contest to build a miniature sun. The British version is more ambitious and more prosaic. It is an attempt to build an industrial system around a technology that has not yet reached commercial maturity.

That system includes the national laboratory at Culham, the STEP project at West Burton, private reactor developers, magnet companies, specialist manufacturers, robotics teams, universities and the space firms testing new propulsion ideas. It also includes planning rules, export controls, safety standards, training programmes and investors willing to wait longer than a normal technology cycle.

If the programme succeeds, Britain will not merely own a power station. It will own knowledge about how to design, build, inspect and repair fusion machines. It will have companies able to export magnets, control systems and manufacturing equipment. It may have a space-propulsion business capable of selling transfer services to international customers. The economic prize would come from the network, not from one reactor alone.

If the programme fails, the failure will not necessarily prove that fusion is impossible. It may show that one design was too expensive, one fuel cycle too difficult or one schedule too optimistic. The responsibility of government and industry is to learn from failure without hiding it or pretending that a new press release has erased it.

The Bletchley plasma test therefore deserves neither ridicule nor worship. It is evidence of progress, and evidence of how far the programme still has to go. Britain has shown that it can build a machine capable of producing and controlling plasma inside a proposed fusion exhaust. It has not yet shown that the machine can make fusion propulsion economical, durable or safe in orbit.

That gap is not an embarrassment. It is the whole story. Fusion becomes credible when the distance between the demonstration and the product is measured honestly. Britain has chosen to enter that distance with public money, private companies and a national industrial strategy. The result will depend less on the magic of the word “fusion” than on whether the country can keep doing the difficult work after the cameras have gone.

References

1. Pulsar Fusion, “In a World’s First, Pulsar Fusion Demonstrates Its ‘Sunbird’ Nuclear Fusion Rocket’s ‘First Plasma’ at MARS Conference Hosted by Jeff Bezos”, GlobeNewswire via Yahoo Finance, 25 March 2026. ↩
2. Pulsar Fusion, “Sunbird Fusion Propulsion”, company description of the Dual Direct Fusion Drive, specific impulse, power output and orbital-transfer concept. ↩
3. UK Government, “Fusion energy powers UK’s Industrial Strategy”, description of the £2.5 billion commitment and STEP’s target of a prototype fusion power plant at West Burton by 2040. ↩
4. UK Atomic Energy Authority, “Conceptual Design Workflow for the STEP Prototype Powerplant”, design aims including net electrical output and tritium self-sufficiency. ↩
5. Tokamak Energy, “From fusion innovation to commercial impact”, company information on HTS magnets, STEP partnership, investment, patents and Ridgway Machines. ↩
6. ITER Organization, “Partnerships help prepare for operation”, discussion of ITER’s data infrastructure and expected data rates during deuterium-tritium operation. ↩
7. STEP Fusion, “Shortlist announced for STEP’s industry partners”, public-private delivery model and industrial supply-chain objectives. ↩
8. Pulsar Fusion, “Sunbird Fusion Propulsion”, published modelling and mission concepts for cargo, outer-planet exploration and a 1,000-kilogram spacecraft to Pluto. ↩
9. Fusion Industry Association, “Fusion Industry Association Announces Launch”, statement on private fusion companies, public-private partnerships and regulatory certainty. ↩