Liked the article?
Nuclear Fusion is the process the powers the sun and the stars, making life on Earth possible. One of the hardest scientific challenges of the 21st century is achieving such a reaction on our planet to use the heat it produces for electrical energy production.
The most promising kind of fusion reactor is called a “Tokamak”. It consists of a toroidal-shaped vacuum chamber able to create and contain plasma – the fourth state of matter - at 100 million degrees or more. In order for fusion to occur, the plasma must be kept at these temperatures for sufficient time. Once two particles fuse, the challenge is to harvest the released energy without losing control of the plasma. The plasma is said to ‘ignite’ when it generates enough heat on its own via fusion that it becomes self-sustaining and additional heating systems are no longer required.
As you can imagine, no existing materials can resist such high temperatures. So we must use other ways to “hold” the plasma. One method is to use very strong magnetic fields to stop the plasma from touching the so called ‘first wall’ (or inside wall) of the reactor. Superconducting coils surrounding the toroidal chamber induce strong magnetic fields that shape the plasma. Furthermore, in tokamak machines the plasma current (circular flow) is induced using a transformer that must be charged before an experiment and rapidly discharged during operation.
This sounds great, but how is it possible to provide the great amount of power needed to such a complex system?
I am actually working at the ASDEX Upgrade tokamak - Germany’s biggest tokamak. This machine is operated at the Max Planck Institute for Plasma Physics (IPP) in Garching where three flywheel generators are charged up before the start of a fusion experiment. A flywheel generator is composed by a motor, a big rotating mass (flywheel) and a generator: the motor is connected to the public grid and - taking power from there (up to 15 megawatts) for several minutes - it converts electricity into kinetic energy by accelerating the rotation of the heavy flywheel up to about 1.600 revolutions per minute. The flywheels essentially provide the 450 megawatts power needed to power a fusion experiment, using only the available 15 megawatts of the public grid. Just to give you an idea, this amount of power is equivalent to half of the average power consumption of the entire Munich area.
Ultracapacitors (a.k.a. Supercapacitors) have been around since the 1960s. They have attracted a lot of attention recently due to their growing use in electric vehicles. Ultracapacitors not only store relatively large amounts of energy but release this energy very rapidly when needed – their primary advantage. Fundamentally, an ultracapacitor consists of two conducting electrodes separated by an electrolyte (usually water), in which a porous separator is soaked. Because the ions that form in the electrolyte fluid can move freely through the separator, they move to the oppositely charged electrode (see Fig.2), forming two Helmholtz double layers. Here’s where it gets even more interesting: the surface of each electrode is not a smooth but is padded with activated (porous) carbon. This results in a surface area about 100.000 times larger than an ordinary capacitor.
For the charging process, due to their very high power density, ultracapacitors cannot just connect to the grid because they require a customized converter to limit the charging current to the desired value. One solution consists in using a ‘boost-converter’ that provides an output voltage greater than its input voltage and limits - through an inductance – the current’s variation.
Two features are the most relevant for fusion devices:
• Power Density: about 10 kilowatts /kilogramme (more than 10 times higher than Lithium-Ion batteries)
• Lifetime: about 1 million cycles against the average 1.000 to 10.000 cycles of Lithium-Ion batteries.
Now the question becomes: can ultracapacitors power a tokamak’s coils? For the discharging process of the ultracapacitors, a Modular Multilevel Converter-(MMC) based topology (schema) has been chosen, as shown in the figure below.
For a standard ASDEX Upgrade experiment lasting 10 seconds, the TF coils require about 1.400 Switching Modules which is about 60 tons of ultracapacitor modules. This sounds like a lot, but it is much, much less than the 220 tons of the actual flywheel used for powering such coils.
The modular structure makes it easily possible to replace any one of the Switching Modules with a new one, making this MMC-based topology much more reliable and lower-risk than having a single, big flywheel generator.
But no one has ever used this many ultracapacitors all at once before. Hence the purpose of my ultracapacitor PhD project is to realize a small scale prototype of this power supply to demonstrate its validity as a possible replacement of one of the huge flywheel generators of ASDEX Upgrade; such prototype will indeed require a very specific power electronics topology that can scale to (almost) any size in future, including future fusion power plants.
This is the first time that anyone tries to bring such a powerful ultracapacitors-based power supply to the fusion field: only small projects have been attempted in the past. Should its feasibility be successfully demonstrated, it could contribute to future fusion power plants possible even starting with DEMO, which will dwarf the ASDEX Upgrade tokamak. Through my work with ultracapacitors, I hope to help bring the fusion dream one step closer to reality.
This figure shows the structure of an ultracapacitor-based power supply for the ASDEX Upgrade tokamak’s toroidal field (TF) coils. The topology is composed of n x m (n=number of modules per arm, m=number of arms) equal modules (a.k.a. Switching Modules). Each module is composed of an IGBT (Insulated Gate Bipolar Transistor) Half-Bridge module to control the power flow, a supercapacitors module where the energy is stored and some passive components making this all work together. By turning the switches of the singles modules on or off it is possible to supply the magnetic coils with the desired electrical voltage and current.