Comment on Helion eyes Washington site for 2028 fusion reactor build
drwho@beehaw.org 21 hours ago
When last I heard, the record for a self-sustaining fusion reaction was 22 minutes. Either they’re expecting a very, very long time to get anywhere close to starting construction, or they know something (which I’d be interested to hear about).
Fermion@feddit.nl 19 hours ago
Helion is a completely different technology vs tokomaks which is what you’re thinking of. They pulse the plasma to create brief bursts of pressure/heating/fusion. They do already have their seventh prototype machine operational so while we can’t independently verify their claims, it’s probably not all bluster.
I have mixed feelings about their approach. They plan to use a deuterium and helium-3 fuel blend. That has a couple major advantages. Most of the reactions will be aneutronic and the energy is released in the form of highly energetic alpha particles and protons. The lack of a high energy neutron is a huge advantage for safety and longevity of a reactor. High energy neutrons are hard to shield from and they cause most materials to get brittle and weaken. Netrons are not good for personnel to be around and they can leave some materials radiactive making reactor maintenance/disposal costly. The other advantage is that since all the energy is released as kinetic energy in charged particles, they don’t have to try to absorb high energy photons or neutrons into a water blanket to drive a steam turbine. Instead, the kinetic energy results in an electromagnetic pulse that can be harvested by the same magnets that constrict the plasma to begin with.
Sounds amazaing, right? So why doesn’t everyone use this approach? Helium is rare, but Helium-3 is especially rare, making up only about 20 parts per million of helium found in geologic deposits. So simply put, it is currently infeasible to use Helium-3 at scale. Helium-3 can be collected as a byproduct of breeding tritium for use in nuclear warheads. Enough helium-3 is produced for some demonstration reactors, but any real amount of demand will quickly outpace what the DOE produces.
Helion plans on breeding their own Helium-3 in Deuterium-Deuterium reactors they will operate. However D-D reactions are not aneutronic. So all the materials lifespan/shielding/ maintenance nightmares that come with operating a nuclear reactor will still apply. That means operators will have to buy very expensive fuel from Helion indefinitely. Helion doesn’t exactly deny this drawback, but I really dislike how much they gloss over it in their public communications.
Here’s a video tour of their test facilities that explains the basics of their approach. youtu.be/_bDXXWQxK38
I’m inclined to think they’ve demonstrated enough results that they are likely to be able to build a working unit quickly, however, that would still be a long way off from creating any sort of sustainable supply chain that would be a viable option for anyone beside datacenters.
theneverfox@pawb.social 18 hours ago
I mean, true, although I thought they were pretty upfront on wanting to produce the fuel in-house and specialize two reactor lines
Which, yeah, isn’t ideal… But if it works, they can’t patent an element. They won’t be able to monopolize the tech forever
It seems like a very elegant solution… If they can overcome the engineering challenges. There’s probably far better designs that could be made from the core idea