You can’t talk about the future of nuclear energy without talking about nuclear fusion. Actually, you can’t talk about the future of energy without talking about nuclear fusion. In this blog I will dive into this subject, looking at how a fusion reactor works, the pros and cons, the misconceptions, and how modern design approaches can bring fusion closer to fruition.
Fusion is supposedly the future of clean, sustainable energy that will solve all the world’s problems. Supposedly. Proponents of fusion boast about its eco-friendly credentials, its safety, and the abundance of fuel. These are all true.
Nuclear fusion works by fusing small elements together, usually hydrogen to make helium. The electromagnetic force between elements repels their nuclei, but if this force can be overcome then the strong force can come into effect. As you might expect, the strong force is very strong, but it only acts over a small radius. Once two nuclei are within that range, then the strong force will pull them together, and in doing so releases energy called the nuclear binding energy. It is this energy that fusion reactors seek to extract.
You will notice that the electromagnetic force acts as a major obstacle that must be overcome. This is both a blessing and a curse. On one hand, this makes fusion reactors much safer than existing fission reactors, because there is no chain reaction. Unless we are constantly putting in energy to overcome the electromagnetic force, the reaction won’t occur. This means there can’t be any Chernobyl or Fukushima situations with fusion.
The safety of fusion is further bolstered by the fuel type: hydrogen and helium. Despite current fusion reactors running on the relatively reactive hydrogen isotope tritium, this doesn’t change the fact that hydrogen is less radioactive than uranium. Much less. The core of a fusion reactor will remain radioactive for 100 years, which is a much more surmountable challenge than fission’s 100,000 years as discussed in my last blog. The reaction does produce radioactive tritium, but this has a very short half-life and is immediately reused in the next cycle.
As you will know, hydrogen also happens to be more common than uranium. We can get it from water after all. In fact, with current technology fusing a glass of seawater would generate as much energy as combusting a barrel of oil. And instead of producing carbon dioxide and carbon monoxide, the only by-product of nuclear fusion is helium – something the world is running out of.
This all sounds great. But on the other hand, it really takes a lot of energy to overcome that electromagnetic force. How much energy? Well, the temperature must be about 50 million degrees Celsius. We have achieved fusion already, but we have never generated net positive energy because we consume so much energy just heating things up. What’s more, that’s just net positive energy, not net positive electricity. Converting the thermal energy from the reactor to electricity is going to work the same way as any power plant and is going to be an optimistic 60% efficient at best. From what I can tell, fusion isn’t going to be viable for a long time, if we’re lucky we’ll see a commercial nuclear fusion power plant in our lifetime. This means we can’t count on fusion to ride in and save us from the climate crisis – fusion happens in the next chapter.
How do you even make a fusion reactor that doesn’t melt? This where the engineering comes in. No material can withstand this kind of temperature; all substances will immediately turn to plasma. Therefore, our hydrogen plasma has to be suspended in the air, or rather, in a vacuum. We have two ways of doing this, but I’d like to focus on the more popular method: magnets. Magnetic confinement fusion does exactly what you’d think, it uses magnets to confine the fusion. This is important because the hydrogen elements must be packed closely together to increase the chances of collisions. Magnets work because plasma is electrically conductive, so responds to magnetic fields. These magnets are also used to impart a force onto the suspended plasma, causing it to accelerate to near-light speeds. This is how the fuel is heated to 50 million degrees. In order for this reactor to run continuously, the plasma needs to have an endless path to run along. This is how we arrive at the standard fusion reactor design: a donut.
This is the tokamak, by far the most popular fusion reactor type. It has the most research behind it and in four years from now will be the first nuclear fusion reactor type to generate net positive energy, at least if ITER keeps to its schedule. However, there is a growing faction of people that believe the tokamak isn’t the best way to do things. They believe there is a better magnetic confinement fusion reactor, the stellarator.
Why would this be the optimal design? First of all, it looks and sounds way cooler. Stellarator. Just reading that word conjures up images of spaceships and laser guns. There’s also a good reason for this shape, which requires me to do a tiny bit more explaining.
You see, tokamaks have a problem. By bending that tunnel into a circle you create an ununiform magnetic field, which makes the reactor less effective at confining the plasma to a linear path. Atoms closer to the inside will experience a stronger magnetic field than atoms towards the outside. This causes a phenomenon called atom drift, and it must be minimised. Modern tokamaks like ITER remedy this issue by placing a giant coil of wires called a solenoid in the middle of the donut. By running electricity through this coil, you can generate a huge magnetic field which imparts a twisting motion on the plasma. This twisting helps to centralise the plasma and suppress atom drift. However, this second magnetic field has huge energy demands. In ITER, the central solenoid requires 6.4GJ to generate a 13 tesla magnetic field, which in layman’s terms is a lot. It accounts for a significant fraction of the reactor’s energy requirements and makes achieving net positive energy much harder.
Enter the stellarator. We’re sticking with the donut shape, but this time we’re twisting the entire chamber to solve that atom drift issue. In doing so we can eliminate the energy hog that is that central solenoid, and dramatically lower the bar for entry into net positive energy.
This approach has recently been moving back into the forefront after a long hiatus. We initially thought they were a no-go due their extreme complexity and lacklustre performance compared to tokamaks, but new improvements in magnet technology and more powerful computers have enabled us to address these areas. It was only once we developed powerful enough computers that we were able to discover these crazy geometric solutions that make stellarators competitive again. Every wall panel and magnet must be uniquely formed to fit into this complex jigsaw puzzle.
This is where design finally comes in. All this complexity boils down to a matter of design and manufacture. I believe stellarators are poised to surpass tokamaks as the dominant approach, we just need to get better at making them.
Designers are now taking advantage of modern supercomputers and generative design to find the optimal arrangement of magnets. Considerations in this process include even magnet distribution, spacing wide enough to allow for maintenance access, and keeping magnet geometry smooth and continuous. Current algorithms also aim to reduce the number of unique magnet shapes, to ease manufacturing.
I believe there is also scope for modern manufacturing methods to have an impact. Large scale 3D printing is already being demonstrated to help make more efficient, custom designed fission reactors. Why not fusion too? This sounds like a textbook application of additive manufacturing – low volume, complex geometries that need to be made more cost effective.
There is a common sentiment that “nuclear fusion is only ten years away, and always will be”. I think this understates the incredible progress that has been made in the last 70 years. Regarding the goal of net positive energy, we have gone from being out by a factor of 10,000,000 to 10, with sights to exceed this landmark in 2025. Will fusion be with us in ten years? No, not commercially. Try 50 years. But we are travelling along the right path at breakneck speed. It won’t arrive in time to stop climate change, but it might play a role in eradicating ‘renewable energy’.
You’ll have to read my next blog for that one.
Sources
https://www.youtube.com/watch?v=FrUWoywZRt8&ab_channel=TechforLuddites
https://thebulletin.org/2018/02/iter-is-a-showcase-for-the-drawbacks-of-fusion-energy/
https://www.youtube.com/watch?v=vqmoFzbZYEM&ab_channel=QuaQuark
https://phys.org/news/2017-04-fast-robust-algorithm-stellarator-yields.html
https://www.youtube.com/watch?v=mZsaaturR6E&ab_channel=Kurzgesagt%E2%80%93InaNutshell
https://www.wired.com/story/coming-soon-a-nuclear-reactor-with-a-3d-printed-core/
McEwan Design 
One thought on “Nuclear Fusion and Generative Design”