Hacker News Reader

Comment by collaborative | Hacker News Reader

> There are two broad approaches toward achieving terrestrial fusion. In magnetic confinement fusion (MCF), magnetic fields are used to confine the hot fusion fuel in the form of a fully ionized gas or plasma that persists for seconds or longer. In inertial confinement fusion (ICF), laser or particle beams are used to compress and heat a tiny capsule of fusion fuel to generate a micro-explosion of a nanosecond duration.

I thought there were now 2 more: projectile method via gas gun and piston containment of molten lead

View on HN

There's an even better one that for some reason gets ignored all the time. Fusion-fission hybrid. Use a fusion reactor that is well below break-even as a neutron source for a fission reactor.

Currently all fission reactors have a common safety problem: their fuel has to be very nearly critical. More precisely, there are 2 types of criticality, delayed criticality and prompt criticality. Prompt needs a higher concentration of fissile material than delayed, or delayed needs a higher amount of moderator (like graphite rods) than prompt. In any case, you absolutely don't want your reactor to become prompt critical. It did happen at Chernobyl, but I think you don't want to be in that company.

All reactors need to be just ever so slightly above delayed criticality some of the time. You can think of it as the R0 for Covid. When R0 is above 1, the disease spreads exponentially. If it's below, it dies down. You want it at a constant level. In this case "disease" is fission. You need R0 to be slightly above 1 a bit of the time, and slightly below 1 some of the time, so the fission keeps going, but does not go out of hand. The thing is this narrow band around 1 needs to be really, really narrow. We are talking 0.999 to 1.001 or so. Why? If for Covid a generation is about 3 days, for nuclear fission, a generation is about 1 millisecond (delayed fission, not prompt one; for that one, a generation is about 10 microseconds, which the funny guys at Los Alamos called "one shake"). If you have an R0 of 1.01, then after 1 second (1000 generations) you get 22000 more fission, and after 2 seconds you get half a billion more fission events than at time 0. That looks an awful lot like an explosion.

Keeping the R0 in a very narrow band is not that easy. It is for sure doable, and that's how all reactors work. But the specter of R0 going to 1.01 is never that far away.

But a fusion-fission reactor can achieve just that. The fission part of the reactor can be kept well below R0=1. You can keep it at 0.98 for example. And then you need to provide an extra 2% of neutrons coming from the fusion part of the reactor. If anything goes out of hand, you just shut down the fusion reaction, and the fission reaction dies down right away.

[1] https://en.wikipedia.org/wiki/Nuclear_fusion%E2%80%93fission...

It gets ignored because it combines the bad features of fission with the bad features of fusion.

Why are you saying that?

Because it's true?

Bad features of fission: generation of large amounts of radioactivity, proliferation, waste disposal, afterheat/meltdown concerns.

Bad features of fusion: complexity, cost, reliability.

http://web.mit.edu/fusion-fission/WorkshopTalks/skepticsvg.p...

That's a good link. But a fairly shallow dismissal.

Both yours (you are quite an active participant on HN, so you are very likely aware of the site's guidelines) and theirs (the authors of the presentation).

The wikipedia article on fusion-fission presents the downsides of the idea much better I think.

The link you provided appears to be some type of high-school debate level of argumentation.

The final report issued by MIT [1] at that 2009 fusion-fission forum is much more balanced (and more informative).

But overall, the negative tone of the report rests on the assumption that fusion is "just around the corner". It is not.

Their argument is roughly: we need research both for a pure fusion reactor and for a fusion-fission hybrid. Let's not get distracted with the fusion-fission, and invest fully in fusion. Which they call "the grand challenge" and "transformative".

The thing is, we don't need "grand challenges" and "transformative" things for their own sake. We need stuff that works. Fusion-fission reactor may be complex, but is clearly achievable. Fusion by itself, not in the foreseeable future.

[1] http://web.mit.edu/fusion-fission/Hybrid_Report_Final.pdf

That link was on slides presented at a professional workshop. If you judge slides by whether they are flashy ppt files, well bless your heart.

I agree that the slides maybe slanted too much in the direction of fusion being just around the corner. But that hardly helps hybrid reactors! If fusion is more problematic that promised, that tilts the table even more toward just using fission, with no fusion component whatsoever.

The major problem of fission is not safety, or waste disposal, or (near term) fuel availability. The major problem is cost. It's difficult to see how hybrid reactors would improve cost. As the slides say, the focus on hybrids is more of matter of "how can we make our fusion work appear more relevant", not "what would customers actually want?"

Maybe a little analogy would help: well before thermonuclear bombs were developed, people started using fusion in the so called "boosted bombs" [1]. In that design (which dominates to this day), the implosion of the Plutonium pit triggers a fission chain reaction, but it also triggers the fusion of a small amount of D+T mix. That fusion adds a tiny bit of energy to the whole explosion, but it also produces a huge amount of fast neutrons, which increases the number of Pu nuclei that split by about 30%.

Of course, we eventually had above-breakeven fusion in nuclear weapons too. But the below-breakeven was and continues to be useful (although some people debate the entire idea of usefulness in the context of nuclear weapons).

> The major problem of fission is not safety, or waste disposal, or (near term) fuel availability. The major problem is cost.

You sound like one of those opponents of nuclear reactors who doesn't understand that cost is driven by regulations, which in turn are driven by safety.

Increasing safety will decrease costs.

But that's not all. A fusion-fission reactor can burn much more of the Uranium/Plutonium/Thorium core. While the fuel cost is not that important in the economics of a nuclear reactor, the costs associated with refueling are. If you don't need to refuel every 18 months, but rather every 15 years, then this changes drastically the economics.

Of course, producing less waste is great too.

Now, you can say (and you said) that fast fission reactors can achieve the same result. The Department of Energy recently let the startup Oklo operate an experimental liquid metal fast reactor [2]. That's great, and I really hope Oklo will succeed getting their NRC approval the second time around.

But pure fission reactors have this problem with criticality. The presentation you linked to dismissed this

  Criticality control, while important, does not dominate fission reactor safety.

This is a very strange thing to say. The safety of fission reactors is dominated by what is in people's mind. And two things are in people's mind: Chernobyl and Fukushima. The first got critical, and the second one may have. The world just does not have the appetite for another one of these. Consequently, the regulators are very concerned with criticality.

Take a look at NRC's assessment of NuScale's design [3], and check how many times they talk about the control rods and their role in criticality.

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

[2] https://www.energy.gov/ne/articles/argonne-adds-new-testing-...

[3] https://www.nrc.gov/docs/ML2022/ML20224A508.pdf

The linked wiki article doesnt go into the reasons why so many hybrid fission-fusion projects have been abandoned, this technology seems so promising that there must be good reasons why.

I think after Fukushima the R&D budget for nuclear fission research has gone down.

For example, the latest Department of Energy budget proposal [1] asks for $1.7 BN for civilian fission-related projects (page 58) and for $0.7 BN for fusion science (page 36). The fission budget is still higher, but fission is something concrete, the US has 100 nuclear power plants still running.

In any case, anything related to nuclear research is expensive. So, even if fusion-fission is promising, it won't get done without quite a few billion being invested in it. This is not something that can happen without government support.

Looking at the Department of Energy portfolio of projects on the fission side, I can't say I don't agree with it. They are ranging from low hanging fruit (like the SMR where NuScale already got the NRC approval) to somewhat higher risk-reward profile. Still the risk-reward profile for fusion-fission appears to be even higher. So, I don't blame them for picking their current priorities.

I could blame them for continuing to pump money in ITER, but I think that's more of an international relations project, so the money needs to go that way.

[1] https://www.energy.gov/sites/default/files/2022-04/doe-fy202...

Reply on news.ycombinator.com