Surviving the Maelstrom inside ITER

How do you build something that can survive for twenty years in the harshest conditions ever created on Earth, with no chance of replacement or repair, and with no test facility to replicate this environment? Welcome to the challenge faced by the designers for ITER, an international effort to create a proof-of-principle nuclear fusion reactor, currently under construction in the south of France. When ITER starts full power operations in 2035, the conditions inside this donut-shaped vacuum chamber will be the most extreme on Earth, with huge magnetic fields, temperatures of hundreds of millions of degrees Celsius, beams of particles close to the speed of light, a bath of intense electromagnetic radiation, and thermonuclear, atom-fusing reactions.

Diagnostics: the senses of the machine

How can we peek inside such a maelstrom and try to understand what is happening? This is where our diagnostics, the senses of the machine, come in. In ITER there will be 50 to 60 diagnostics in many shapes and sizes: from tiny conical probes, to high speed cameras to capture the writhing plasma motion, up to large, powerful lasers to measure the temperature of the plasma.

But in ITER, we don’t simply want to know what the plasma is doing, we also need to use that information to control and stabilise the plasma so that we can sustain the fusion reaction. This is handled by CODAC, the ITER Control, Data Access and Communication system, which considers the vast amount of information coming from the diagnostics, and makes split second decisions on how to control the plasma to keep it stable, and how to protect the ITER tokamak device from plasma disruptions which can damage the machine. This places a heavy burden on the reliability and accuracy of the data provided by the diagnostics, made more challenging by the harsh conditions inside ITER which the diagnostics must withstand.

We don’t simply want to know what the plasma is doing, we also need to use that information to control and stabilise the plasma so that we can sustain the fusion reaction.


As part of a team at the Max-Planck-Institute for Plasma Physics, in Garching near Munich, I work on the bolometry diagnostic for ITER, a simple diagnostic that’s a bit like a tiny solar panel. Hot plasmas emit light - think about the glow of a red hot poker - and this light cools the plasma down. If the plasma cools too much, the fusion reaction will either not occur or stop. At its most basic, a bolometer is just a tiny thin strip of metal exposed to light from the plasma - the metal absorbs the light, and heats up. This changes the resistance of the metal strip, which we can measure, and from that small change we can work out how much light the plasma emits. There will be over 500 of these detectors surrounding the ITER plasma, so we can build up a picture of where plasma is brightest. CODAC can then use this picture to control the plasma stability and keep ITER running.

In order to get enough light onto our bolometers, we have to place them very close to the plasma - some will be peeking out between gaps between the blanket modules, which make up the first solid surface inside the machine that are exposed to the hot plasma. This exposes the bolometers to some of the harshest environments in ITER, but also makes them very difficult to replace. Once ITER reaches full power operation in 2035 and begins to burn deuterium-tritium fuel, it will no longer be possible for humans to enter the vacuum vessel. Any maintenance will have to be done remotely by robots, which makes repairing or replacing components very difficult.

Withstand hostile conditions

As the information from the bolometers is a key part of the control and safety of ITER, they must continue to function throughout the twenty years of expected operation, whilst sitting in a uniquely hostile environment, bathed in ultra-fast subatomic particles and mind-bogglingly high temperatures. This is difficult to guarantee - there is no environment that can adequately reproduce the conditions in ITER, and so we must test individual aspects of our diagnostic and build in redundancy to guard against failure of some of the bolometers.

For example, one aspect of the ITER environment is the neutrons. These subatomic particles have no electrical charge, so they can escape the strong magnetic fields which confine the rest of the plasma. Travelling at a reasonable fraction of the speed of light, they act like a bowling ball, colliding with the atoms in solid materials and rearranging them, leaving voids and weakening the impacted material. Even when hidden behind the blanket modules, our bolometers will be constantly bombarded by neutrons, which will cause them to weaken and fatigue. Even worse, neutrons can transmute elements - one traditional material used in bolometers is thin gold resistors, which can be turned into mercury by the neutrons. As mercury is a liquid, this is bad news - our bolometer could literally melt and drip away!

We can attempt to mimic the neutron damage expected in ITER using a nuclear fission reactor. We can place a sample inside a fission reactor and expose it to the neutrons there, and then study the damage. Although the conditions are not the same as in ITER, we can try to scale up the damage to understand how likely a given material is to survive. One possible solution is to use platinum instead of gold, which should be more resilient to the neutrons.

As well as these nuclear fission tests, we subject our bolometers to high temperatures under vacuum, compress and stress them to mimic the forces expected during ITER operation, blast them with electromagnetic noise to see how the accuracy of the signal is affected, and spray them with jets of steam to mimic the rupturing of water coolant pipes. These tests feed into our constantly evolving design.

Working on diagnostics for ITER is challenging, with a bewildering array of requirements to fulfil and tight deadlines to meet, set by the ever accelerating pace of construction. In the end, ITER is an experiment. We can do everything we can to try and predict how it will behave, but undoubtedly there will be some surprises! By careful design and prototyping, we make our diagnostic as reliable as possible. But the true test will come inside ITER. Here our bolometer will sit for twenty years while being baked, squeezed and bombarded with neutrons, steadily delivering the data to control and stabilise the plasma within the machine which is the next step on the path to a fusion power plant. It’s exciting to be able to play a small but essential part in such a huge project.