Engineering the future of fusion

How we’re overcoming the many challenges of turning a laboratory experiment into a large-scale industrial process

The generation of power from nuclear fusion is not a new concept. For more than three-quarters of a century, we’ve known that stars such as the sun compress hydrogen atoms together to create helium. When these light atoms are fused together, energy is released. Nuclear fission, which is the process used by the current nuclear reactors connected to the grid, generates power by splitting heavy atoms apart to release stored energy. While fission results in nuclear waste, fusion is much cleaner. With our growing energy supply concerns, fusion power has the potential to provide abundant clean energy for centuries. However, we have yet to achieve the continuous and controlled fusion process necessary to efficiently derive power from it.

Fusion power is no longer a physics problem, but an engineering one. Controlled nuclear fusion has been achievable since 1951, when the world's first successful fusion experiment was performed at the Los Alamos National Laboratory using a Z-pinch machine. However, transforming a laboratory experiment into an industrial process is difficult at the best of times. This is especially true when you’re essentially extracting energy from an artificial sun.

A typical fusion reactor compresses deuterium and tritium together, heating them up to 150,000,000°C in vacuum to create plasma, which is held in place by a powerful magnetic field. These magnetic fields are not unlike those seen on the warp core of the Starship Enterprise.

Meet the JET sun

The Joint European Torus (JET) at the Culham Centre for Fusion Energy (CCFE) is a purely experimental tokamak reactor. Built in 1977 in conjunction with EUROfusion, JET is the first stage in realising a fusion power plant.

"ITER is expected to be three times larger than JET and weigh 23,000 tons – around the weight of 115 blue whales."

Many of the findings from JET are informing the design of ITER (Latin for “the way”), which is the next generation of experimental fusion reactor, and is currently under construction in France. ITER is expected to be three times larger than JET and weigh 23,000 tons – around the weight of 115 blue whales. The findings from ITER will, in turn, lead into the creation of the first demonstration fusion power plant, called DEMO.

However, before fusion power plants are fully realised, there are several key issues that need to be resolved. “We have performed fusion on JET: the plasma physics is largely solved,” says Damian Brennan, the active operations department head of CCFE. “Now come the engineering tasks.”

Fusion reactors are colossal machines. The reactor for JET is the size of a large house, and is contained within a building the size of an aircraft hangar. These reactors are also incredibly complex: more than 2,500 control cubicles are required to manage JET. Should one cubicle fail, the process will not work.

"CCFE have to avoid running JET at times of peak electricity demands such as the advert breaks during popular television programmes, to avoid placing undue load on the National Grid."

It takes 16 weeks of preparation to start JET, thus shutdowns are costly in both time and money. Restarts require long periods of leak testing to ensure that the vacuum pressure of 1x10‑8mbar, required for plasma generation, is maintained. This is followed by a stepping up of power for the neutral beam systems and plasma, until the optimum operating conditions are achieved.

At the moment, JET can only create bursts of plasma for up to 30 seconds every 20 minutes and requires 1% of the National Grid’s capacity for each burst of plasma. CCFE have to avoid running JET at times of peak electricity demands (called “pulse avoidance periods”), such as the advert breaks during popular television programmes, to avoid placing undue load on the National Grid.

"The materials that will be exposed to the fusion process must be able to resist neutron damage, have a low activation, and be sufficiently robust that they can last for years rather than months."

Fusion power creates enormous wear on reactor components. As well as the colossal heat and pressure required for fusion, the plasma emits fast neutrons, which embed themselves in the protective tiles encasing the reactor core. Over time, these cause irradiation embrittlement, leading to accelerated degradation of the reactor’s structural materials.

Thus the materials that will be exposed to the fusion process must be able to resist neutron damage, have a low activation, and be sufficiently robust that they can last for years rather than months. Materials development is therefore one of the key engineering challenges.

For a fusion power plant to become viable, it would need to be simple, compact and reliable. Furthermore, such a reactor would need to provide a “long low hum” of continuous power generation. ITER, for example, will be designed to pulse for up to an hour, with superconducting magnets enabling the extended runtime.

Greener than coal, but safer than fission

"Despite the challenges, fusion is a highly tempting form of power generation. It doesn’t produce greenhouse gases, has a potentially abundant fuel supply, and presents a vastly reduced risk of radiation when compared with nuclear fission."

Despite the challenges, fusion is a highly tempting form of power generation. It doesn’t produce greenhouse gases, has a potentially abundant fuel supply, and presents a vastly reduced risk of radiation when compared with nuclear fission. The radioactive waste from a fusion reactor will be safe to recycle within a century; fission waste is an environmental burden for thousands of years.

Fusion is also inherently safer than fission, as it doesn’t have the problem of stored energy. If the reactor core is ruptured, even before the crack is visible, the plasma field will collapse due to the loss of vacuum. All that will remain in the reactor is tritium, which can be safely recovered through tritium-filtration systems.

However, plasma is a highly unstable state of matter: in JET, plumes of so-called edge-localised modes (akin to solar flares) can often be seen arcing from the plasma field. Such instabilities in the plasma cause disruptions that can be so strong they cause the reactor to move.

“A long time ago [JET operators] were seeing what they could do, and having a pretty good go of it, and they had a big disruption, so they shut down early and went home,” recalls CCFE’s Brennan. “The next morning they had a phone call from one of the Oxford University seismic stations saying ‘What the hell did you lot do last night at 8.45pm? We registered an event on your site!’”

“Sun in a bottle”

“The next morning they had a phone call from one of the Oxford University seismic stations saying ‘What the hell did you lot do last night at 8.45pm? We registered an event on your site!’”

Plumes of plasma dump huge amounts of energy into the reactor walls, corroding the tiles. Frozen pellets of deuterium-tritium, fired into the plasma in time with any “rattle”, can dampen the oscillations. Another method is to use resonant magnetic perturbation (RPM) coils around the plasma to dampen the disturbances.

One of the areas under the greatest stress through cyclic heating is the divertor, which removes waste heat from the plasma while the reactor is operating. This can experience a power load of up to 30MW/m². For perspective, a space shuttle’s tiles receive “only” 10MW/m² during atmospheric re-entry. Scientists are looking at using a gas buffer composed of nitrogen to absorb a proportion of the heat before it hits the divertor, thus reducing thermal loads.

Currently, the protective tiles of JET’s first wall are made from beryllium and tungsten. Previously the fusion core was protected by carbon tiles, which have a similarly low atomic mass, but these absorbed the deuterium and tritium fuels too easily, adversely affecting efficiency.

While deuterium is relatively abundant on Earth (it can be extracted from seawater), the finite supply of tritium needs to be addressed. New technology is being developed that can breed kilogram quantities of tritium from lithium blankets wrapped around fusion reactors. “Theoretically it is possible, and mathematically quite easy, but you still need to engineer the blankets,” says Brennan.

"Despite the numerous technical issues facing fusion power, it is expected that we will have an economically viable working fusion power plant within 50 years."

CCFE’s Mega-Amp Spherical Tokamak Upgrade (MAST-U) is the UK’s latest fusion experiment and the next stage in fusion design. Smaller and simpler than JET, MAST-U uses a cored-apple shape. This will potentially give a far more efficient design, as it needs a lower magnetic field for the same level of performance. It will also incorporate two divertors, which can potentially distribute the thermal load. The results from MAST-U will further inform the design of ITER and allow more plasma experiments to be conducted.

Despite the numerous technical issues facing fusion power, it is expected that we will have an economically viable working fusion power plant within 50 years. “We have to,” concludes Brennan. “Because of energy resources and global warming – we have got to do something different. Now we have the sun in a bottle, let’s do the engineering!”

READ NEXT: Taking on climate change with technology

Photos: Peter Gatehouse

Read more about: