• A prototype of an “International Thermonuclear Experimental Reactor” - or ITER - is currently in the design phase, according to our partner The Conversation.

  • The ITER project is an extraordinary one: it is the result of a partnership between more than thirty countries for a budget of around 20 billion euros over several decades.

  • The analysis of this phenomenon was carried out by Emmanuel Franck, researcher at the National Institute for Research in Digital Sciences and Technologies (Inria).

Clean energy, without the risk of serious accidents and virtually unlimited? This dreamlike goal has been associated with the principle of nuclear fusion for decades. So much so that in 1986, despite their strong differences, Ronald Reagan and Mikhail Gorbachev proposed cooperation to test this approach. This project, joined since by the European Union, India, China, Japan and South Korea, took the name of "ITER", for "International experimental thermonuclear reactor". Finally initiated in 2006, it will provide a proof of concept that fusion can be a source of energy. The next step, an industrial-scale prototype, is in design and is called Demo - it would certainly be implemented in practice if ITER demonstrated that energy can be drawn from such a reactor.

What principles are behind this now well-known name?

How to develop a project of such a scale in practice?

What is nuclear fusion?

Nuclear energy is known because it allows us, especially in France, to supply us with energy, but also of course because of the risks associated with it and which led to famous accidents, in particular those at Chernobyl in 1986 and Fukushima in 2011.

Example of a nuclear fission of uranium © Fastfission / Wikipedia CC BY-SA

This energy is based on an atomic reaction,

nuclear fission

, which involves breaking a heavy atom into two lighter atoms. In current reactors, it is the uranium which is cracked by collision with a neutron, for example forming one atom of krypton and a second of barium or even strontium and xenon. During this reaction, part of the initial mass is converted into energy which is recovered and the reaction also releases a neutron which, itself, will “break” another atom and so on. Unfortunately, it is possible to lose control of these successive reactions, which can lead to accidents.

As early as the 1940s, scientists GP Thomson and M. Blackman postulated, by filing a reactor patent, that there was an approach which consisted rather of "fusing" two light atoms into a heavier one, again with a loss of mass. that can be recovered in the form of kinetic energy (fast particle), then in the form of heat.

This reaction is the one we find in the heart of our sun and most stars.

This research started seventy years ago and ITER will not start to be tested until 2025. Why is it taking so long?

A nucleus of deuterium and a nucleus of tritium merge into a nucleus of helium © Wykis / Wikipedia CC BY-SA

For a reason that is ultimately quite simple. To fuse two hydrogen atoms - the smaller and simpler of the atoms to be fused together, the hydrogen gas must be heated to several million degrees when the gas is very dense (several tens of times the density of the area). ), as is the case in the sun. For more reasonable densities, which are considered in practice in a reactor (several thousand times less dense than air), a temperature must be reached between 100 and 150 million degrees. Obviously, no known material can withstand such conditions. How then to heat a gas to this temperature and enclose it in a reactor without danger?

Fortunately, at this temperature the gases become electrically charged, the electrons separate from the atoms and we speak of "plasma". The brilliant idea of ​​Russian scientists Igor Tamm and Andrei Sakharov in the 1950s was to come up with a device called a “tokamak”. This involves enclosing the plasma in a chamber in the shape of a


, or "torus" in mathematical language, and forcing it to rotate inside the torus without ever approaching the edges using fields. Extremely powerful magnets which confine the gas to the center of the torus.

The objective of ITER is to find out whether this technology can be used for energy production.

For this, ITER seeks to produce more than 5 to 10 times of energy than that used to heat the plasma and to confine it for a few minutes.

If this is the case, we can move on to industrial prototypes by 2050 or 2060.

Plasma in its toric cavity.

A human indicates the ladder, bottom right © ITER

Although the principle of tokamaks has existed since the end of the 1950s, the construction of ITER did not begin until the end of the 2000s. Indeed, the technology of tokamaks is very complicated and we built, over time. over the years, increasingly complex and larger tokamaks until ITER, which should be the first to produce more energy than that used to make it work.

When digital technology makes it possible to better control risks and costs

One of the central difficulties for making nuclear fusion work is the management of “instabilities”: confinement of the plasma at 100-150 million degrees by the magnetic field will inevitably have small defects.

These defects can cause a portion of the plasma to "escape" towards the edge of the containment chamber, leading at best to a loss of energy, at worst to very heavy damage to the containment chamber (this type of damage would not give rise to a Fukushima-type nuclear accident, but would have a very significant financial cost).

Black smoke rises from Reactor No.3 at the Fukushima Nuclear Power Plant, Japan, March 21, 2011 © REUTERS / Ho New

A central issue is therefore to detect and predict these instabilities in order to control or avoid them. A containment chamber and most devices are very expensive and we cannot afford to directly test solutions for instability control in a real tokamak. Therefore, physicists are using mathematical and numerical models of plasma dynamics in the containment chamber to test potential control and detection methods. It is first of all a question of transcribing the problem in the form of mathematical equations (very complex, because coupling phenomena taking place at different scales of time and space),then to solve these equations using supercomputers which will make it possible to predict the evolution of the plasma and its response to a new control method.

This modeling / simulation process is in fact very frequently used in industry and in physics: in meteorology to forecast the weather, for tsunami prediction, in medicine for digital twins or in the automobile and aviation industry for test the prototypes.

This type of tool is used to separately simulate the different phenomena present in a tokamak, because we do not yet know how to model the complete operation.

This has already made it possible to propose several relevant avenues for controlling instabilities in recent years.

Our “Nuclear” dossier

Recently, these numerical models combine physical approaches (models of fluid mechanics and electromagnetism) and artificial intelligence methods, which make it possible to build predictive models from experimental data or data from simulations.

From a scientific point of view, ITER is an extraordinary project.

It is the result of a partnership between more than thirty countries for a budget of around 20 billion euros over several decades.

It mobilizes theoretical physicists, engineers, materials specialists, computer scientists, mathematicians, who work together to make this old dream a reality for the next generation for whom the energy issue will be central.


Sun: The study of neutrinos confirms that the energy emanating from them is 100% nuclear


Energy: How hydrogen could help store electricity on a large scale

This analysis was written by Emmanuel Franck, researcher at the National Institute for Research in Digital Sciences and Technologies (Inria).

The original article was published on The Conversation website.

Declaration of interests

Emmanuel Franck received funding from Eurofusion.

  • Digital

  • Fukushima

  • Chernobyl

  • Nuclear plant

  • The Conversation

  • Nuclear

  • Science