Hacker News new | ask | show | jobs
by cctt23 2997 days ago
Unsolved problems for fusion as powerplant:

-Sputtering of shielding and every other part of the reactor from fast neutrons. If your reactor is becoming brittle as it’s in operation, and requires constant maintenance, it won’t be operating enough to be cost effective.

-Breeding blankets. If we’re not breeding tritium in the blanket (and no one has been able to sustain a reaction that way yet) then we’re just using fission a lot. Expense becomes an issue, as does radiological issues.

-The plasma diverter is very much an unsolved problem. I can get into more detail here, but in short this is the part of the reactor that “skims” some of the hot plasma off to do the work. The dynamics of very hot, magnetically constrained plasmas still escapes us, and when you throw a rock into that stream, the complexity increases. Current divertes wouldn’t last a day in an operating plant. Disassembling your whole plant every day and reassembling it is a non-starter.

-Containment of plasma at sufficient energies is still something measured in seconds, or fractions of seconds. The usual metaphor is trying to uniformly squeeze a balloon; it will just “squirt” out. For s research reactor a second or two of fusion is an achievement. For power generation it’s nothing.

-Neutron activation of otherwise inert materials means you’re going to have serious radioactive waste. It’s unclear just how dirty D-T fusion would be from soup to nuts, but “pretty dirty” seems like a good bet.

-Tritium penetration.

-Most of the energy produced is in the form of neutrons, and we don’t know how to use that as a source of power. Those neutrons, in addition to destroying the reactor itself and activating materials, represent a loss.

-What we really need is aneutronic fusion through alternative cycles to D-T, like p-p, but that’s a much hotter plasma and no one has a clue how to make it work yet.

-Coolant for a constantly running reactor is a boring, but unsolved problem.

There’s more, but these are the ones most poeple on HN probably are aware of when they dismiss this article.

Some further reading https://thebulletin.org/fusion-reactors-not-what-they’re-cra...

2 comments

I spent several hours watching MIT's presentations about this reactor design.

- The inner wall of the reactor is 3D-printed and replaced annually. This is easy because they've found they can make joints in the superconducting tape that add very little resistance, allowing them to include hinges letting them open the reactor.

- Surrounding the inner wall is FLiBe molten salt. It's heated by the neutron radiation, acts as coolant for the thermal cycle, and as the breeding blanket (each high-energy neutron releasing two neutrons from impact with beryllium, providing plenty of neutrons for breeding tritium from lithium). Having a liquid blanket makes tritium harvesting easier.

- Stronger magnetic fields damp down plasma instabilities, making containment easier. For years MIT has been running the Alcator C-Mod, which has more powerful fields than any other tokamak in the world, so they have some direct experience here.

- The neutron-activated wastes would only need containment for several decades.

(I don't know anything about the diverter, and I'm interested if you want to get into more detail.)

The MIT folks argue that we understand tokamak plasmas much better than any other configuration, and have gotten far closer to breakeven than alternative designs, so that's where we should focus.

However, there are some projects working on aneutronic fusion, mainly with p-B11. The biggest project is Tri Alpha, with $500M invested. They've achieved stable plasma at 10M degrees, and are about to start testing a new reactor which should reach 100M degrees. If the plasma continues working as the expect, they think it's a straightforward path to a production reactor; of course there could be surprises.

Another approach is laser fusion with petawatt picosecond lasers. We're not far off from having a laser with the specs to attempt this, and these lasers improve by a factor of ten every three years.

Helion, funded by YCombinator, is attempting a hybrid D-D/D-He3 reaction. (The output of D-D is half He3, and half tritium which decays to He3 with a 12-year half-life.) They say the hybrid reaction would release only 6% of its energy as neutron radiation. I don't know how it's going though.

The waste problem from neutron activation is beginning to get the publicity it deserves in both articles and in comments like the parent's.

Fusion is advertised as a clean power-generation technology, but it will, in general, have its own radioactive-waste disposal problems.

It’s still a couple of orders of magnitude or more better (in terms of half life of the byproducts). So if anyone knowingly compares fission waste and fusion waste putting them on the same plane is simply lying.
Or they recognize that for a human and generations of their descendants, 30,000 or 300 years doesn’t change anything for them. Either way it’s longer than the US has been a country for one metric. Both cases require similar solutions, long-term containment, and so both run into the same political inertia fielded by NIMBYism.

To claim otherwise is, as you would have it, lying.

More importantly, both are solved issues from a purely technical standpoint, it’s just that everyone wants someone else to deal with it. As a bonus, we can actually use fission to produce energy, right now. There’s no issue of, “it will be great when...” grid storage is solved along with intermittency for renewables. There’s no, “it will be great when...” fusion is producing energy rather than heavily parasitizing from the grid. There’s no waiting until we’re completely screwed by climate change effects.

And if we’re willing to deal with that waste, then we already have a viable source of atomic energy from fission! The future is whenever we decide to deal with the problem instead of leaving it pools or casks to solve itself.

@csallen: Be willing to actually dispose of it. We need to commit the money and political will to set up a single disposal site. Right now it’s a NIMBY nightmare so we get the worst outcome.

Curious what you mean by "deal with" it. Somehow make it non-radioactive, or improve our disposal techniques, or what?
I think the tradition is to dump it off the coast of Somalia.
My thinking is more around using thorium cycle nuclear reactors which consume almost all their input material [1].

1T of thorium produces the same amount of energy as 35T of uranium or 4166000T of coal. 83% of the waste products are inert, and 17% require only 300 years to reach background levels of radiation. Much safer than uranium cycle too.

This technology is very promising IMO, though given the general attitude people have towards nuclear, it's not surprising it is underdeveloped, more research is needed.

[1] https://en.wikipedia.org/wiki/Liquid_fluoride_thorium_reacto...