It seems that for the last couple decades fusion is always 30 years away and by this point we’ve well passed the initial 30 and seemingly little progress has been made.
Is it just that it’s so difficult to make efficient?
Has the technology improved substantially and we just don’t hear about it often?
In: Physics
It’s a very hard problem. It’s “30 years away” because it’s previously been very difficult to even know what further challenges lay ahead until you get over some roadblock problem that’s holding you back right now.
I’m going to talk a little about how we’ve mostly been trying to do fusion over the last few decades – using a reactor shaped like a big donut, which is called a “tokamak”. It’s from a Russian acronym that means “toroidal chamber with magnetic coils”. “Toroidal” just means “donut shaped” and we’ll get to the magnetic coils in a bit.
It’s a hard problem because it relies on getting matter (gas/plasma) VERY hot (millions of degrees), and holding it together close enough for long enough (confinement) that fast moving particles of this energetic gas collide with each other, and combine to make heavier elements. When this happens the resulting combination (fusion!) releases a lot of energy as heat and radiation, which we want to harvest – the usual approach has been to collect some of the excess heat, then use that heat to heat water into steam, to turn a turbine so we can then use the rotational energy along with magnets to produce electricity. The steam turbine thing is also how “regular” nuclear power (fission power) works. Except in that application, you just gather up enough radioactive material together in one place, the material decays (which means tiny particles shoot off of it) producing heat.
The trouble with fusion is that when you get these particles so super hot, you a) can’t let them touch the walls of the vessel you’re containing them in, and b) they *really* want to repel each other. So you have to put a ton of energy in to get the reaction to work in the first place – think of squeezing some things together at really high pressure. That’s essentially what they’re doing, only it’s primarily with magnets. The donut-shaped chamber is wrapped with wires which, when you run electric current through them, produce specifically shaped/directional magnetic fields which work to contain the electrically charged plasma inside and “squeeze” it together.
The idea with a tokamak is that you have a “sustained” reaction where you put a lot of heat/energy in initially to get the reaction started, then the heat released from the reaction itself helps to keep it going over a long time period, and then you can extract some of that heat. This is called “ignition” and it’s really the hardest part to achieve. We can make fusion for short amounts of time but you have to really control the conditions to get a sustained reaction. The plasma is flowing around the ring and there is turbulence due to the changing magnetic fields as you go around, and understanding how every particle moves and how that affects the reaction is really, really difficult.
There are also issues with just how you actually extract the excess heat out of the system – if you let too much heat out through some part of your reactor, it’ll melt/vaporize and the whole thing will be destroyed (note – this would not be a nuclear explosion like a nuclear bomb – those need very specific conditions to happen). Also you need to keep enough heat in the system (inside the plasma, really) to keep the reaction going. Even if you can control the release of heat in the way you want to, there are issues with the inside of your reactor degrading over time due to the very high heat and radiation.
So it’s difficult because of the extreme conditions needed to sustain the reaction. There are physics challenges (understanding how it all works, how the reactions work etc.) and engineering challenges (what materials to use, how to build the devices, etc.) which all play into each other.
We can’t easily foresee the challenges and the solutions which will come up because we don’t know what advances in technology will be made which might help. There are several big advancements that have happened which will definitely help though. For one thing, there’s been a big advancement in superconducting magnets: [https://news.mit.edu/2021/MIT-CFS-major-advance-toward-fusion-energy-0908](https://news.mit.edu/2021/MIT-CFS-major-advance-toward-fusion-energy-0908)
This means that they can produce the magnetic fields needed to achieve the containment needed much more easily/efficiently, which means they can make the reactors smaller, which makes the whole thing easier.
Another thing is more gradual: Computer hardware. Simulation of the reactors takes a LOT of processing power, and of course since the 1980s computing power available has increased massively. This helps a lot in loads of different ways.
There are also huge advances in manufacturing technologies to build the reactors. Computer controlled machinery, 3D printing, advanced 3D CAD, materials technology advances, all sorts.
And finally, there has been a recent explosion in people trying fundamentally different technologies. We thought that the best way was with a big donut-shaped reactor but now there seems to be more funding available for fusion research, there are companies trying out different ways.
One company is essentially making a big cannon which fires a specifically-shaped bullet at a specifically-shaped target, which creates shockwaves which collapse and create the very high temperature and pressure needed for a fusion reaction – for a split second. They would do repeated shots (say, 10 per minute) inside a chamber and extract excess heat from the chamber.
Another is using a different type of simpler reactor and instead of extracting heat, using the radiation produced to directly produce electricity.
Another is trying to use big pistons to compress plasma to create the right conditions.
There are other approaches using ~100 massive lasers to heat a tiny fuel capsule which causes is to collapse and create massive temperature and pressure.
Basically there are lots of different approaches being tried now. Lots of people think they have an answer, and it’s entirely possible that some of them will work, relatively soon. However a large part of the whole “30 years away” thing is that you don’t really know if something will work, or what challenges you will face, until you build something and try it. And building the reactors, depending on the design and the approach being used, has led to some of the most complex engineering challenges, and largest projects ever conceived, in any field. The most famous one right now is ITER being built in France. It’s the largest tokamak reactor we’ve ever built and should be ready to start making plasma in 2025.
Latest Answers