The promise of fusion

Excitement in the fusion community is palpable. Jokes about fusion forever being 30 years away have grown stale. Investment – which once crawled at a snail’s pace – is flooding in as some private companies say they’ll demonstrate successful fusion by the mid-2020s with power on the grid by the 2030s. Weight for weight, fusion fuel releases ten million times the amount of energy of coal.

Oil and gas firms are joining private investors, and the likes of Google, Jeff Bezos and PayPal founder Peter Thiel are backing the sector.

Now the UK is feted as the most fusion-industry-friendly country in the world, with the prospect of supply chain and regulation certainty on the horizon, as well as playing host to future experimental pilot plants. Companies have been founded around the world, from India to China to Australia, as well as North America and Europe.

“In one fell swoop, the investment in private fusion companies has doubled – it’s really exciting,” says plasma physicist Dr Melanie Windridge, founder of Fusion Energy Insights. She compiled a survey of 30 private fusion firms worldwide, published in November by the Fusion Industry Association (FIA), which revealed 18 firms had attracted $1.8bn (£1.3bn) of investment, rising shortly afterwards to $2.4bn (£1.8bn). This was trumped by the announcement that US-based Commonwealth Fusion Systems – a spinout from MIT – had raised a further $1.8bn to build and operate a pilot plant, with construction under way in Massachusetts and predicted to run by the end of 2025.

Over the last decade, says Windridge, the fundamental science has matured after years of publicly funded research – previously only the large state-funded facilities could pursue fusion with confidence as billions was invested. Now smaller companies look poised to sidestep vast state experiments with smaller, cheaper devices. Some are achieving temperatures ten times hotter than the sun.

“That’s not to say it’s all solved, but private companies are building on a good base of mature plasma physics,” says Windridge. “People no longer think it’s decades away.” Advances in supporting technology in the form of more powerful magnets, high-performance computing, machine learning, and advanced manufacturing techniques have whipped up enthusiasm.

“The private sector is moving faster than many inside and outside of fusion expected,” says Andrew Holland, chief executive of the Washington-based FIA. “Fusion has moved from being something that was done in national labs to a potential business.”

There are, he notes, many ways to achieve fusion, and light-footed companies are trying them all. With each milestone passed, more funding is released, and investors’ expectations are high. “Our livelihoods depend on delivering – or we’re gone,” says Michl Binderbauer, chief executive of TAE Technologies, which has operated since 1998.

“The vast majority of [FIA] member companies are using alternative approaches, there’s little technological overlap,” says Holland. “These are mostly things left on the cutting room floor by national programmes in the 1970s and ’80s.”

Nuclear fusion sounds deceptively simple – smashed together with enough energy, two lightweight atomic nuclei fuse to create a single larger nucleus with slightly less overall mass, releasing extra energy. This is the reaction that keeps the sun and stars shining, controlled by gravity. While nuclear fission – the source of nuclear power today – is easy to start and hard to stop, fusion is the exact opposite – hard to ignite and, to date, impossible to sustain on Earth.

Challenges are complex – California-based TAE Technologies has filed 1,800 patents on supporting technologies alone. “The complexity of the machines is beyond anything we’ve designed,” says Dr Arthur Turrell, a plasma physicist and author of ‘The Star Builders’. “There are tens of millions of parts [to a fusion reactor].”

Now some companies say they have reached or expect shortly to reach high enough temperatures – more than 100 million degrees – for matter to bind.

But controlling and containing the unruly, electrically charged hot plasma where fusion occurs for any length of time has so far defeated scientists – plasma typically lasts for millionths of a second. There are, says Turrell, two main methods of controlling the ionised particles. One is magnetic confinement – levitating the plasma in a vacuum within a tightly woven and powerful magnetic field, most commonly within a doughnut-shaped chamber: a tokamak.

“Once this hot ethereal material hits the edge of the can, it will cool off and it’s game over,” says Binderbauer. TAE Technologies is working on a less conventional linear reactor with rotating plasma.

Another method, inertial confinement, squeezes the plasma to extreme densities to achieve fusion before it collapses and leaks energy. In the US, the National Ignition Facility last summer declared it was on the cusp of igniting inertial confinement fusion when high-intensity lasers were fired into a tiny amount of fuel. Inertial confinement is also the method explored by University of Oxford spinout First Light Fusion, which uses shockwaves to compress plasma.

But generating more energy than goes in is key to unlocking fusion as a viable source of energy – and no one is there yet. It was the European experimental facility JET (Joint European Torus), based at the Culham Centre for Fusion Energy in Oxfordshire, that came closest. Briefly it generated 67 per cent of energy input in 1997. Other milestones, such as controlling high-​temperature plasma for increasing seconds to minutes, have been passed. Eventually, says Turrell, fusion pioneers want to generate 30 to 100 times the energy that goes in.

And then there’s fuel. Most models use deuterium and tritium, which are isotopes of hydrogen. While the former is plentiful and found in seawater, tritium is scarcer and mildly radioactive, has a half-life of 12.3 years, and can be created from lithium within a fusion reactor.

This year Canadian company General Fusion will begin work on a new experimental plant in the UK – chosen after a three-year global search – a pilot that should be operational by 2025. “There was a unique set of conditions that lined up to support what we wanted to do,” says Jay Brister, a veteran of nuclear power and now chief business development officer at Vancouver-based General Fusion. Culham, where the demonstration plant will be built, is home to the Atomic Energy Agency (UKAEA), the body in overall charge of UK fusion research.

The UK government has begun encouraging a supply chain and establishing how to regulate the nascent industry. “We wanted the right host with the right set of characteristics to move forward with our demonstration facility,” says Brister. “Globally I think the UK is way ahead on the regulatory front.” The demonstration plant will cost £295m and the government has invested an undisclosed sum.

In 2020, the government declared its ambition for the UK to be the first country to commercialise fusion technology, with plans for a new prototype fusion energy plant – STEP (Spherical Tokamak for Energy Production) – to contribute to the grid by 2040; five potential sites have been shortlisted in England and Scotland.

And in a business park near Didcot, British start-up Tokamak Energy is building a spherical tokamak equipped with high-​temperature superconducting magnets which it predicts will be deployable in the early 2030s, having declared good progress to reaching 100 million degrees Celsius.

By contrast, General Fusion, which has raised at least $200m (£150m) of public and private funds, is using the inertial confinement method – holding a ball of plasma within a liquid metal shell. Pistons will squeeze the ball to compress the plasma. As fusion occurs, high-energy neutrons escape into the liquid lithium. The heat from this will be used to generate electricity and create tritium. “CFS and General Fusion are probably the two companies in the private sector who will have something operational by 2025,” says Brister.

Last year, US start-up Helion Energy attracted $500m (£368m) of funding for its planned plasma accelerator, which aims to generate electricity directly rather than via a conventional power plant, and says it has reached over 100 million degrees Celsius during testing. Now on its seventh prototype, Helion says this will produce a small amount of net electricity by 2024.

Back in 2014, Lockheed Martin announced it had made breakthroughs with a compact fusion reactor, although little has been announced since and the project is famously secretive within the sector.

Google-backed TAE Technologies, which has raised $880m (£650m), predicts its reactors will be commercialised within ten years, and in experiments has already reached more than 50 million degrees Celsius. A new prototype will operate at more than 100 million degrees Celsius.

TAE is currently partnering with Japan’s National Institute for Fusion Science for the first large-scale experiments with hydrogen and boron. There are many upsides, says Binderbauer, not least that these fuels are plentiful, cheap and not radioactive so won’t degrade material or require shielding. TAE’s design fires neutral beams into a central chamber to rotate and heat the plasma. As it spins around an axis, it gains stability without the need for strong magnetic fields.

Ultimately, TAE is aiming for temperatures ten times higher to fuse hydrogen and boron – up to a billion degrees Celsius, which Binderbauer acknowledges sounds preposterous. TAE’s design borrows from experiments at CERN, where particle physicists have advanced understanding of accelerating and colliding particles. “CERN operates at the equivalent of trillions of degrees,” says Binderbauer. TAE is mixing particle accelerator physics with plasma physics, with a goal of higher temperatures and greater density plasmas.

Binderbauer says his company, in partnership with Google, has made strides in the use of machine learning and artificial intelligence to learn how to control plasma with split-second adjustments.

“The plasma reactor timescale is short, so electronics, the efficiency of software, everything has had to evolve. Now we can control things that even five years ago were difficult and 10 years ago absolutely impossible – it would have been science fiction… we have enough technical tools to meet the challenge,” he says. “We’re going to crack the nut, there’s no doubt.”

In October 2021, the UK declared its ambition to create a world-leading fusion industry through support for the private sector. A 2021 discussion paper set out proposals to regulate the industry, with a response expected early this year. In April last year, the UKAEA awarded a contract to create the world’s first tritium research centre, H3AT, where scientists will investigate the processing, distribution, storage, recycling and disposal of the fuel.

“There’s a high degree of economic and uncertainty because of regulatory burdens,” says Binderbauer. “The UK is the first to establish a fusion licensing framework, and I salute the country for creating certainty.”

But how private and public sectors work together in the future must be clarified, says the FIA’s Holland. “We’d love to see countries around the world emulate the UK – they are definitely leaders,” he says.

The cost and size of predicted future fusion energy plants varies. Smaller companies are looking at creating 50-100MW plants, and some could be suitable for off-grid power for mining operations, or for data centres, or even shipping and space travel. But the vast majority focus on the electricity market and aim for 200-300MW plants, says Holland. “Nobody other than maybe the Chinese wants a multiple-gigawatt-sized power plant. It’s just too big for demand.”

Last year for the first time, fusion made the agenda at the latest COP meeting in Glasgow. But even if we make great strides in renewables, 86 per cent of the world’s energy is still generated by fossil fuels and by 2050 industry predicts the world will need half as much energy again as we do today.

“If we’ve learned anything from the pandemic,” says Holland, “[it’s that] if the will is there to do something, there are ways to accelerate it. But it has to be a priority.”