The UK government has a very ambitious plan for nuclear fusion, which I don’t think is widely enough known about. The plan is to build a pilot nuclear fusion plant able to deliver electrical power to the grid by 2040 – the Spherical Tokamak for Energy Production (STEP). The project was launched in 2019, and the current government has guaranteed funding for it at the very significant level of £500m a year for five years.
At a time when many people from different political positions agree that a big problem of the UK state is its inability to deliver big projects, this is a huge investment to build state technological capacity.
This post is a brief introduction to the STEP project. Nuclear fusion does generate some reflexive scepticism – we all know the jokes: “it’s twenty years in the future, and always will be”. I want to get beyond that, while still being realistic about the huge challenges this programme faces. I’ll describe some of the technological and engineering issues, and the approaches being proposed to overcome them.
I’m only going to discuss STEP; I make no attempt at a comparison with other international and national projects, such as ITER and the the planned Chinese pilot plant (China Fusion Engineering Test Reactor), or with the various private sector approaches in the the USA and elsewhere.
The Spherical Tokamak for Energy Production (STEP)
The UK’s civil nuclear fusion programme has been going many decades, run by UKAEA, the UK Atomic Energy Authority. Its main activity for many years was hosting the Joint European Torus (JET) at its research centre at Culham, near Oxford – this was a European Union project which for many years held the record for the closest approach to energy breakeven. UKAEA has also developed an alternative design for a fusion reactor – the spherical tokamak – commissioning a large scale research facility, MAST (the multi ampere spherical tokamak), which was substantially upgraded with a 2020 launch as MAST-U.
The Spherical Tokamak for Energy Production (STEP) project aims to build a larger spherical tokamak, using the lessons learnt in MAST-U and JET, and incorporating new technologies such as high temperature superconductor magnets and a liquid lithium tritium breeding blanket. The tokamak will be integrated in a prototype power plant at a scale which would allow it to export of order 100 MW of electrical power to the grid, and which will generate its own fuel. The geometry and characteristics of the spherical tokamak means that it should be possible to build it at a significantly lower cost than conventional tokamaks like the international collaboration ITER, and the goal is to deliver power to the grid by 2040.
The STEP project was announced in 2019, with ~£200m of funding for a 5 year design phase. The current government has announced funding of £500m for the next five years, to work up detailed plans. A site has been found at West Burton, Nottinghamshire, at an old coal-fired power station, and the decision has been made that the plant will be regulated as a standard industrial site, rather than as a nuclear site like a fission reactor.
STEP won’t be a commercial power plant – even if it all works to plan, it’s not big enough to be cost-effective. But it should demonstrate and validate all the technologies that are needed to generate electrical power from fusion: it’s an engineering project, not a science experiment. The goal is that the experience gained in designing, building and operating STEP will be sufficient to design a series of production nuclear fusion power plants.
Fusion Basics
If the nuclei of light elements such as hydrogen can be brought together in very close proximity, they can fuse to form a heavier nucleus like helium, with the release of substantial energy. This is the process that keeps the Sun and other stars hot – and is also the source of the huge destructive power of modern nuclear weapons. If the process could be reliably controlled and exploited at reasonable cost, it would provide an essentially unlimited source of (relatively) clean, zero-carbon energy.
That’s a big “if”, though. The nuclear forces that glue protons and neutrons together in an atomic nucleus, and are the origin of the energy released during nuclear fusion, are very strong, but operate only at very short ranges. But nuclei are positively charged, and so electrostatic forces mean that nuclei will repel each other. These electrostatic forces are much weaker than nuclear forces, but they operate at much longer ranges. There is an energy barrier that needs to be overcome before two nuclei can be brought close enough together to fuse – to overcome this barrier, the materials have to be raised to temperatures of hundreds of millions of degrees.
At this temperature, matter forms a plasma, in which the atoms are completely ionised, with the electrons are completely separated from their nuclei. In order to achieve significant rates of fusion, the plasma needs to be confined at a relatively high density for a significant period of time. There are two ways of confining high temperature plasmas on Earth enough to achieve nuclear fusion (in the Sun, it’s gravity that confines the plasma, but you need a star’s worth of matter for this to work). In a thermonuclear weapon, material is compressed by an intense burst of x-rays arising from a fission bomb. Inertial confinement fusion, as carried out in the US National Ignition Facility at Lawrence Livermore, reproduces this process in a controlled way on a much smaller scale through an array of very high powered lasers.
STEP will use magnetic confinement. The charged particles of a plasma can be trapped by a suitable magnetic field. In the most usual arrangement – a tokamak – the plasma is confined in a doughnut shaped vacuum chamber, with an arrangement of magnets that produce a field pointing around the ring of the doughnut (with a slight twist). Charged particles are constrained by the Lorentz force to move in spirals around the lines of the B-field. The bigger the magnetic field, the tighter the spiral – and the more effective the confinement.
What elements are candidates for a fusion reaction? Three elements are relevant for STEP – hydrogen, helium and lithium. Because we’re talking about nuclear physics, we need to consider each isotope individually. Hydrogen has three isotopes: normal “light” hydrogen, with a nucleus consisting of a single proton; deuterium (or heavy hydrogen), with a proton and a neutron; and tritium, with a proton and two neutrons. Deuterium is stable, and naturally occurring hydrogen contains about 0.02% deuterium. Tritium is unstable and radioactive, decaying with a half-life of about 13 years to form helium-3. Helium-3, with two protons and one neutron, is one of the two isotopes of helium – it is stable, but very much rarer than helium-4.
In the easiest fusion reaction to exploit (i.e. the one that needs the lowest temperature) a deuterium nucleus and a tritium nucleus fuse to create helium-4, a neutron and 17.6 MeV of energy. The downside of this “D-T” reaction is that tritium is hard to come by – it is currently only created artificially in fission reactors. So widespread use of the D-T reaction for power generation relies on fusion reactors being able to generate their own tritium. This is possible by exploiting a neutron capture reactions in the light metal lithium. Lithium has two isotopes, both naturally occurring and relatively common. Lthium-6 has 3 protons and 3 neutrons, and lithium-7 has 3 protons and 4 neutrons. Lithium-6 will absorb a neutron to produce helium-4 and tritium.
The motivation for D-T fusion is that 1 kg of deuterium, reacting with the tritium produced from 3 kg of lithium, generates as much energy as burning 35,000 tonnes of coal, while producing no carbon dioxide: it is a route to clean energy abundance.
In its last experiments, the Joint European Torus at Culham demonstrated D-T fusion in a tokamak, producing 69 MJ of fusion energy in a 5 second burst. However, JET has still not demonstrated “energy break-even” – to obtain this amount of fusion energy out, roughly three times this much energy needed to be fed in to heat up the plasma.
To state the obvious, to be commercially viable, a fusion reactor needs to generate more energy than it consumes. STEP’s task is to demonstrate this at an engineering level, producing a surplus of electricity over that which is needed to run the plant. This will involve the efficient harvesting of all the energy produced by the fusion reaction – including both the energy carried by the helium nucleus, which heats up the plasma, and the 14 MeV neutron, which will pass through the walls of the reaction vessel.
In addition, STEP needs to be self-sufficient in the production of the rare isotope tritium, so that the only fuel inputs are deuterium and lithium. The entire plant needs to be run safely; fusion reactions don’t produce the very long-lived nuclear waste that fission reactors do, but tritium itself is radioactive, and needs to be handled carefully, while the high neutron fluxes produced in the reaction could induce radioactivity in some of the plant components, necessitating careful selection of materials.
Confining the plasma in STEP
STEP will use magnetic confinement, but instead of using a conventional doughnut-shaped torus, it will use a spherical tokamak to confine the plasma. The spherical tokamak is still toroidal in topology, but rather than being doughnut shaped, it is more like a sphere with a narrow core down the middle. A spherical tokamak can be smaller and more compact than a torus – and thus cheaper to build. UKAEA’s MAST-U is a spherical tokamak, so the lessons learnt there about how effectively to confine a plasma, avoiding the dangerous instabilities that confined plasmas are susceptible to, can be transferred to STEP.
A key new technology that makes this more compact design possible are new types of magnets using high temperature superconductors. The higher the field one can generate inside the tokamak, the higher the density of the plasma; the fusion power density varies with the fourth power of the magnetic field – so if you can double the field, you can increase the rate of fusion by a factor of 16.
The technology of making conducting ribbons from rare-earth/barium/copper oxide (REBCO) ceramic superconductors has developed to the point where these materials can carry useful current densities and have high enough critical fields to meet the needs of plasma confinement. “High temperature” is a relative term when it comes to superconductors; these still need to operate at 20K, with associated cryogenics. A huge amount of energy is stored in these very high magnetic fields, and it is very important to avoid “quenches” – runaway thermal events which lead the coil to stop being superconducting, heating up even more by resistive heating, damaging or even destroying the coil.
A general issue for all the components close to the plasma is their susceptibility to damage from the very high flux of energetic neutrons generated by the fusion reactions, including both the REBCO itself and the insulator that the coils are wound on. Much research will be needed on the radiation tolerance of all the materials used in STEP; the neutron flux is considerably higher than materials are exposed to in fission reactors, and the neutrons have higher energies than those produced in fission.
Breeding tritium
Nuclear fusion as a scalable energy generation technology is a non-starter unless it can generate its own tritium. The solution is to use the neutrons from the fusion reaction to react with lithium, producing tritium that can be extracted and injected into the plasma. At first glance, this looks very challenging – each fusion reaction uses up a single tritium atom, and produces a single neutron.
In the most useful reaction to produce tritium, a single neutron is absorbed by a lithium-6 nucleus, to produce a single tritium atom. But it seems unlikely on the face of it that one could capture every single neutron, as one would need to replace all the tritium used in the fusion reaction. In fact, given likely losses in the system, one will need to have a breeding ratio (i.e. the ratio of the number of tritium atoms produced to the number used up) greater than unity – STEP will be aiming for 1.25.
STEP plans to use a breeding blanket of liquid lithium. The other materials used to need to be as transparent as possible to neutrons, so as few as possible are wasted. The liquid lithium is likely to be enriched in lithium-6; lithium-7 does absorb neutrons too, and usefully it doesn’t use up a neutron in the process of making tritium, but the likelihood of the reaction is smaller. Other materials that effectively multiply the number of neutrons may need to be included too.
The engineering involved in handling the liquid metal, and extracting the tritium from it, is likely to be challenging. The materials in contact with the lithium need to be resistant to an environment of molten lithium, with an operating temperature of 650 °C and a high flux of neutrons. Much heat will be generated in the metal – this is where much of the energy of the neutrons will be dumped, and the lithium-6 neutron capture reaction itself releases energy. The breeding blanket will need to be cooled, using helium gas.
Materials for managing heat and neutrons
In earlier phases of fusion research, the tricky physics of controlling plasmas was the limiting factor. As the STEP design is being developed, much more is known about plasmas – arguably, the limiting factors now are related to material properties, particularly in compact designs like the spherical tokamak. The very large magnetic stresses put high demands on the structural materials, while those parts of the reactor close to the plasma will undergo huge energy loads and very high exposure to high energy neutrons.
To focus on heat, the STEP design anticipates that the fusion reaction will generate 1750 MW of thermal energy. Added to this is 417 MW from the heat that has to be put into the plasma, together with heat generated in the molten lithium from the tritium breeding reaction.
The walls will be be exposed both to radiative heat, and heat generated from the charged particles in the plasma hitting the walls. A particularly challenging part of the design is the “diverter”. This is the tokamak’s exhaust system, through which some of the plasma needs to be bled off at the top and bottom of the reaction vessel, so the helium generated in in the fusion reaction can be removed, and tritium and deuterium recycled. One of the achievements of the MAST-U programme so far has been test new designs for the diverter which minimise the heat and radiation burden on the materials.
The main coolant for the reactor will be helium – operating temperatures will be high, to support efficient conversion of energy into electricity. Some parts of the reactor – the core and the diverter – will be water-cooled: this needs to be heavy water (with hydrogen replaced by deuterium) to minimise neutron absorption. It’s very important to keep water away from molten lithium.
The planned overall energy balance is that the 2280 MW heat generates 925 MW of electricity. But 775 MW of this electricity needs to go back to run the plant, leaving us with about 150 MW net electricity output to go onto the grid.
Can AI solve the problem of nuclear fusion?
No, not in the Sam Altman sense that we just create a superintelligence and ask it how to harness nuclear fusion. But high performance scientific computing – both physics-based simulation and machine learning – will be crucial for materials design, optimising the magnetic confinement, and controlling the plasma.
Success isn’t guaranteed, but the UK should do it anyway
This is just a sketch of a few of the issues that will need to be solved in order for the STEP project to meet its ambition of putting electrical power on the grid by 2040, and setting the stage for a new generation of commercial fusion reactors. Even this brief and incomplete account should make it obvious that this project presents a huge array of engineering and scientific problems. Everything needed for STEP to work is possible in principle, but nothing is easy, and success is far from guaranteed.
STEP is a really substantial engineering project that will build project capacity and drive much research and development, and build innovation capacity, in areas like materials science, robotics, and large scale computer modelling, here in the UK. We are already seeing a cluster of private sector fusion companies growing up near the UKAEA headquarters in Culham, attracted by the skilled people and the climate of innovation in related areas of technology.
The location of STEP in West Burton should significantly broaden the geographical impact of UKAEA. The location is in commuting distance of Doncaster, Scunthorpe and Sheffield, communities that are still struggling with the impact of deindustrialisation, but which are building new capabilities in advanced manufacturing. UKAEA already has a facility on the Advanced Manufacturing Park in Rotherham, close to the Advanced Manufacturing Research Centre, which has done so much to nucleate a new cluster of firms focused on advanced manufacturing and advanced materials.
The UK’s history means that it does have a genuinely world competitive position in fusion research, strengthened enormously by its hosting of the Joint European Torus, and evidenced by the progress of the MAST-U spherical tokamak project. UKAEA has demonstrated its capacity to do complex and sophisticated engineering and science at the technology frontier. There are not many parts of the UK state that one can say that about.
What if STEP does succeed? If it does demonstrate an economically viable route to generating electricity, we might expect next generation plants, on a much bigger scale than STEP, providing firm power to a fossil-free grid. What it won’t do is contribute significantly to net zero by 2050. It’s a project that will deliver in the second half of the century – a gift from our generation to our children and grandchildren.
Having committed to this course of action, I think we should generate a national consensus to see this project through. There are two dangers to avoid. One is the enduring tendency of the UK state to prevaricate and delay, justifying cost-cutting with a self-fulfilling scepticism. The other is a danger of success, of trying to find a private, often overseas based, buyer for the project too soon. This should be a flagship project for UK state technological capacity, over which the UK should maintain ownership and control.
Sources
I’ve drawn heavily on the collection of articles in the Theme issue of the Philosophical Transactions of the Royal Society: “Delivering Fusion Energy – The Spherical Tokamak for Energy Production (STEP)”