This picture shows a section of plasma (pink - authentic colour) superimposed on JET vessel equipped with the ITER-like wall. In actuality the plasma  would not be confined to a small section of the vessel as in this image, but would extend the whole way around. Credit: EUROfusion Consortium

This picture shows a section of plasma (pink - authentic colour) superimposed on JET vessel equipped with the ITER-like wall. In actuality the plasma would not be confined to a small section of the vessel as in this image, but would extend the whole way around. Credit: EUROfusion Consortium

Infrastructure  |  Focus On  |  Sustainability

Everything we know (so far) about nuclear fusion

Profile of a possible new energy source, which would have a limited environmental impact. Scientific research has been at work since the 1950s, but the first power plants may only be realised as of 2050. And meanwhile the world's energy needs are growing

Nuclear fusion is an experimental methodology aimed at obtaining a possible, potentially almost unlimited source of energy for civil purposes, using the same physical principles that power stars such as the Sun. The aim is to trigger an artificial 'fusion' between nuclei of atoms of light elements to generate energy, to such an extent that the reaction is able to be self-sustaining, producing more than is consumed to activate and maintain the process itself. The process is the reverse of nuclear fission, which, in order to release energy, causes heavy chemical elements to decay.

The context

Global energy consumption is expected to grow by 50% by 2050, compared to 2020 (Eia International energy outlook, 2021), and at the same time the world is increasingly looking for non-climate-altering energy sources.

According to the framework designated by the International Energy Agency (Iea), in a 'Net Zero Emissions' scenario by 2050, 90 per cent of electricity will be provided by renewables and 8 per cent by nuclear power.

Although this share is expected to fall from 10% today, nuclear power will have to double its generated power from 413 GW today to 812 GW in 2050 and its energy production from 2,690 TWh in 2020 to 5,500 TWh. Although 90 per cent of this growth will be concentrated among emerging economies, the global acceleration is expected to be greater than in the last three decades, when nuclear capacity increased by about 15 per cent. The development of fusion is therefore seen as an opportunity by governments, companies and supranational research organisations, due to numerous advantages over fission itself.

Raw materials

  • The fuel for nuclear fusion is heavy isotopes of hydrogen, namely deuterium and tritium
  • Useful metals for various components will be niobium, tin and titanium, as superconductors within the magnets to control the plasma, as well as lithium to self-generate closed-loop tritium within the reactor.

Date of invention

The first concept for a magnetic fusion facility was the 'stellarator', invented by Lyman Spitzer at Princeton University in 1951.

The second model was the 'tokamak', built in 1958 by Natan Yavlinsky of the Kurchatov Institute of Atomic Energy (USSR).


The tokamak (a Russian acronym meaning 'toroidal chamber with magnetic coils') is the most promising type of experimental facility for harnessing fusion energy with magnetic confinement of the plasma (it is adopted by 45% of companies in the field).

The world's largest will be ITER in France, a project in which the EU, USA, Russia, China, Japan, India and South Korea are participating. Weighing 23,000 tonnes (equivalent to 3.5 Eiffel Towers), it will bring the plasma to a temperature of 150 million degrees (10 times the core of the Sun) in a volume of 840 cubic metres. The tank will have an external diameter of 19.4 metres and will be 11.4 metres high.

The core of a tokamak is an empty doughnut-shaped chamber inside which, under the influence of extreme heat (the Chinese system, East, reached 120 million degrees for 101 seconds, in 2021) and enormous pressure, hydrogen becomes plasma.

Here, deuterium and tritium particles collide with each other, overcoming the electromagnetic repulsion between nuclei. Fusion generates helium nuclei, neutrons and large amounts of energy and heat, which can be used to power a power plant.

Feeding energy into the grid will be the goal of DEMO, Iter's prototype successor planned for 2050 and expected to produce 300-500 MWatts of energy, equal to the annual needs of 1.5 million households. To this end, the consortium Eurofusion (26 EU member states, Switzerland, UK and Ukraine), which in February 2022 with the Joint European Torus facility in Oxford produced a record 59 megaJoules (power of 11 megaWatts) of energy, maintaining fusion conditions for five seconds, will collaborate.

The plasma can be controlled by huge coils of superconductors placed around the container, which generate an electromagnetic field helping to keep the unstable, red-hot raw material away from the walls of the structure (in the case of the JET facility, the magnetic field is 10,000 times stronger than on Earth). The Divertor test tokamak by Enea under construction in Frascati will use, for example, 26 kilometres of niobium and tin cables and 16 km of niobium and titanium cables at 269 degrees below zero.

Magnetic plasma confinement is also employed by the stellarator which, instead of inducing electric currents within it, uses external coils to generate a rotating magnetic field. This method is in use in the Wendelstein 7-X reactor (Max Planck Institute, Germany). The construction of such facilities requires millimetric precision in building such cable 'shafts'.

Inertial confinement of plasma is another modality conceived in 1972 and used for example at the National Ignition facility in Livermore, in California. A millimetre-sized spherical shell filled with deuterium and tritium is rapidly compressed by high-power laser irradiation. The metal fuel capsule reaches very high temperatures and pressures, capable of triggering nuclear fusion and emitting sufficient energy, before expanding again. NIF first achieved an amount of energy (1.3 MegaJoules) almost equal to the energy spent to compress and heat the plasma (1.9 MJ), in July 2021.


  • The fusion process does not emit greenhouse gases, only helium nuclei and neutrons. It is believed that the radioactive waste (neutron activation) of a fusion plant can have a lifetime of about a century, as opposed to the thousands of years of fission
  • Deuterium is an abundant element that can be extracted from heavy water or sea water
  • One gram of deuterium undergoing fusion produces the same amount of energy as 30 tonnes of coal.

Weak points

  • Tritium does not exist in nature and is very rare, currently only 20 kilograms are available worldwide. It also decays in a rapid time (about 12 years).
  • One gram of tritium costs $30,000, but saving one kilogram of tritium for the needs of the DEMO plant would cost 2 billion.
  • Deuterium-tritium fusion remains the 'easiest' to achieve at the 'lowest' temperatures (over 100 million degrees).


  • Tritium is a waste element from nuclear fission power plants and only Canada, South Korea and Romania make it available
  • The availability of power plants for commercial purposes is not expected until 2050 (ITER).
  • Cost estimation is difficult at experimental stage. The Iter Project, for example, rose from an initial forecast of EUR 4.9 billion (2001) to EUR 13 billion (2016). When the roadmap set the first plasma in 2025 (a date later postponed), EUR 4 billion was added to the valuation.


  • Increased funding for fusion companies ($2.8 billion in 2022 alone, out of a total of $4.8 billion)
  • Increase in companies established in the last decade worldwide (23 out of 33 in total)
  • Possibility of integrating Artificial Intelligence to control magnetic coils and maintain plasma configurations (DeepMind at the Swiss Plasma Center of the Swiss Federal Institute of Technology in Lausanne).
  • Possibility of producing tritium from plasma neutron collisions with a lithium coating or insertions inside the reactor (among DEMO's objectives).
  • Development of new reactor models, such as Colliding beam fusion (Tae Technologies) or magnetic target fusion (General Fusion-Ukaea).
  • Possibility of optimising fusion using hydrogen and boron, as an alternative to deuterium-tritium fusion. The boron makes the plasma more stable, helping to confine it by keeping the temperature high.

The expert

“Nuclear fusion can be a desirable solution in supporting an effective energy transition in the context of the climate crisis, the worst of the problems of our time,” explains Paola Batistoni, head of the nuclear fusion development and promotion section at Enea, Italy's lead national agency for projects in this field. “The vast majority of energy in 2050 will be from renewable sources, but we will still need the 10-20% of continuous and programmable base power that nuclear power can provide. Fission will still preponderate in 2050, but fusion can replace it in the long term”.

“Of course, a new industrial chain will have to be created, but the expected benefits are so great that many states wanted to participate in the research,” Batistoni continues. “Fusion poses no risk of nuclear proliferation, the fuel is practically inexhaustible and evenly distributed on Earth, it leaves no waste, it does not cause accidents with major long-term impacts, it is possible to stop the process at any time, and of course it does not emit greenhouse gases”.

“The project with the most important results so far is undoubtedly Eurofusion's JET, because it achieved a record in fusion power and duration, working in deuterium-tritium,” Batistoni concludes. “Every project in the world studies different aspects and this technology appears almost mature, so much so that private initiatives are multiplying. However, I do not believe that in the next decade someone will already be able to feed fusion power into the grid, with energy gain and tritium self-sufficiency, one of the biggest challenges along with research into low-induced radioactivity materials. JET confirmed the forecasts for ITER, which will have to ascertain scientific and technological feasibility, to be followed by DEMO, which will have to demonstrate the safety and cost-effectiveness of fusion. The horizon, once all the technological challenges have been met, is 2050 for us”.

Daniele Monaco - Freelance journalist, he has collaborated with Ansa, QN-Il Giorno, and Wired Italia, where he writes about Economics and digital issues. He works alongside press offices and communication agencies as a copywriter and consultant for the production of content related to digital transformation, innovation, sustainability, Industry 4.0, for companies, associations, public bodies, consortia and startups. Professional since 2010, he holds a dual degree in Communications from the University of Milan, the city where he resides and obtained a master's degree in Journalism from the Catholic University. His website is

More like this