Hacker News Reader

So as someone who follows fusion somewhat I'm embarrassed to admit I'd never heard of this. For reference, pB here means proton-Boron [1].

To expound on what the commenter here has said: neutrons are a big problem for hydrogen-fusion reactors for two reasons:

1. A high-energy neutron represents a loss of a lot of energy, reducing the overall efficiency of the process since that neutron is basically "lost". It obviously has no charge so can't be captured.

2. Additionally, high-energy free neutrons tend to destroy their containers. I don't know if this will make the reactor itself radioactive or not. The commenter seems to think so.

Hence the idea of aneutronic fusion, which is fusion that produces no neutrons or very few. This is part of the attraction of Helium-3 fusion [2].

The neutron problem seemed so fundamental (in that neutrons are required to trigger fusion reactions, which is why you use deuterium and tritium rather than protium (Hydrogen-1) but those neutrons are the very problem that (so far) has made this method impractical. I remain skeptical if this particular problem will ever be solved. Neutrons are fine for stars. Stars are big so that neutrons don't go that far and they're big enough that gravity solves the containment issue.

It's also worth noting that there's no such thing as "free" energy. There is energy where the fuel is free (solar, wind and even fusion) but that still doesn't mean the energy is free. The plant that produces it costs money and takes up space. It must be maintained and it has a shelf-life. Look at the total cost of the plant over its lifetime and divide that into the power it produces and there's your baseline energy cost. So a $50B plant that produces 5MW for 5 years for no fuel cost isn't much use to anyone.

[1] https://en.wikipedia.org/wiki/Aneutronic_fusion#Proton-boron

[2] https://en.wikipedia.org/wiki/Helium-3#Fusion_reactions

I think some proposals on the design include a barrier material around the reactor chamber that would capture the neutrons and thusly convert them to heat, which can be used to drive a turbine.

IIRC the aim was to find a material that would convert to a shortlived radioactive isotope instead of a long lived one, so you can simply remove the barrier and have it sit in a quiet place for a few years, then recycle it.

Of course. Such a blanket is absolutely needed in a DT reactor, to regenerate the tritium that the reactor needs to keep running. But the materials from which that blanket is made, and the first wall separating the blanket from the vacuum vessel, are among the things that would be degraded by neutron exposure.

The "blanket" (pumped molten salt is frequently cited, or more exotic material, many feet thick) is needed to turn fast neutrons into heat, thence to boil water to drive a turbine.

But the equipment to contain the plasma, many, many tons of expensive stuff, has to be inside the blanket, and is destroyed in short order by neutron flux.

The "envelope" that keeps your "blanket" from spilling out all over the floor, a monstrous apparatus of pipes, is also being irradiated, but there are materials for the pipes that can stand up to it for a while.

This envelope is made in hundred-ton sections bolted together so it can be drained periodically and sections replaced. The used ones are radioactive and need to be kept somewhere safe.

This is all hugely expensive, finicky work that is very dangerous to be around, so you need even bigger remote-operated machines to do it.

MIT's ARC design solves neutron flux by making the inner core of the reactor easily replaceable. The superconducting coils are hinged, so once a year they can just open up the reactor and lift out the core, which is 3D-printed. The core is surrounded by molten FLiBe salt, which functions as coolant and tritium breeder. The new superconductors allow the reactor to be much smaller than ITER for the same power output.

MIT's project has been spun off into the startup Commonwealth Fusion Systems, which has investment from an Italian oil company and Breakthrough Energy Ventures. Neither party has an interest in physicist job creation or weapons work.

There are at several pB efforts and one D-D/D-He3 (Helion, which says only 6% of energy would be released as neutrons) but it's harder to get net power from aneutronic fusion and we don't understand the plasma physics as well for the designs they need, so it's more of a wildcard at the moment.

Thank you, this is informative.

I will say here that I have no knowledge that pB aneutronic fusion can be made to work, but whether it can has no bearing on whether tokamak generators will ever go on line.

MIT's ARC design has an estimated cost of $29/W(e), an order of magnitude higher than PV.

Each 170 MW(e) ARC reactor will use 40% of the world's annual production of beryllium.

The "easily replaceable" in your description is quite a stretch. The replacement will be much more demanding than the replacement of fuel elements in a fission reactor. In the latter, the radioactivity is almost entirely contained inside the fuel elements, and they are simply transfered, as fuel bundles, to a cooling pool.

But the 86 tonne reactor vessel in ARC will have been permeated with tritium, and the tungsten will have been loaded with activation products by the intense neutron bombardment. It will be much larger than the rather compact arrangement of fuel rods that makes up a PWR core (about 165 tonnes of fuel, including structural material, in a typical 1000 MW(e) power reactor, which has nearly six times the power output of the ARC design.) The reactor vessel will take more space to remove. And then it will have to be crushed. The workspace in which this happens will become inaccessible to people, as it will become contaminated with tritium and radioactive tungsten dust. The overall volume inside the reactor building where all this happens will be very large, making the building expensive.

"There was, and is, no intention ever to actually use tokamaks as an electrical-energy source."

I thought that's what DEMO which is planned to come after ITER is for.

I have listened to quite a few people working on fusion (ITER, also the people working on Wendelstein-X) and none of them were as negative as you are. Where do all the things you saying come from?

DEMO can be postponed indefinitely.

Have you heard of any significant investment in maintaining structural integrity of the tokamak itself?

Fission reactors need to maintain the integrity of the coolant pipes and the physical supports for the fuel, all just passive structure. I.e., valves, pumps, etc. are behind shielding.

It's all just basic physics. And basic politics.

Are you saying that all the people working on this this stuff are either idiots or liars? My impression is that there are a ton of problems but they believe they can be solved.

The people working on tokamak are engaged in fascinating technical problems, and generally leave politics and practical engineering problems to others.

They spent many years getting their particle physics degrees, and have no real job prospects besides this, research accelerators, or direct weapons work. You kind of have to feel sorry for them.

There are beginnings of private investment in what might turn out to be practical (i.e. aneutronic) fusion power systems, but they do not seem to need very many physicists.

How do you know all that stuff?

Friend of mine used to work on it. I think, from one of the pub chats we often had, the average number of times each atom in the reactor wall was expected to be knocked out of place by neutron collisions was in the order of 100.

This is why I am down on fusion compared to most advocates. On the other hand, I’m also up compared to most pessimists because the everlasting delay looks like it’s caused by an asymptotic-to-zero funding graph.

I still have silly ideas about how to improve the efficiency of Farnsworth-type designs — one I’ve never had time to simulate was “can star mode be enhanced with a simple magnetic field, and does it happen in the first place because of the magnetic field created by current flowing between inner and outer grids?”. I had that idea 10 years ago, I hope someone here has more time to sim it than I do.

Another thing that's rarely mentioned is that the power density of tokamaks will be very bad.

ITER's power density is 0.05 MW/m^3 (gross fusion power divided by volume of the reactor machinery; plasma volume is smaller). The power density of the MIT's ARC reactor is somewhat better, at 0.5 MW/m^3.

But the power density of a PWR reactor vessel is 20 MW/m^3, a factor of 40x better than ARC and 400x better than ITER.

Given that fusion reactors are much more intricate and use much more sophisticated materials and systems than a PWR reactor vessel, and are also much larger, the fusion reactors cannot help but be much more expensive per unit of thermal power output than fission reactors.

BTW, the main reason p11B fusion gets less investment is that for a given magnetic field the peak fusion rate of a p11B plasma is three orders of magnitude worse than for DT.

Of course one would not burn pB in a tokamak. So, that number is not relevant, although it is always trotted out to justify not diverting money from tokamak.

Practical pB designs rely on colliding beams as in the LHC, or other plasma-dynamic schemes that engage the materials in their own confinement.

Colliding beam schemes for p-11B will not work, as was shown by Rider and Nevins decades ago. This didn't stop TAE from selling such a scheme to investors, even though they were told it didn't work.

> There is a reason: the main purpose of the tokamak is a jobs creation program for high-neutron-flux physicists, to maintain a population to draw upon for weapons work.

No, that is completely misleading. The reason is that pB fusion requires 10 times higher plasma temperatures which are a much bigger problem then some neutron irradiation.

Neutron irradiation problem can be managed by using materials that don't generate long-lived isotopes on neutron capture and occasionally replacing the embrittled materials. Which are orthogonal solutions to solutions required for the already-difficult plasma containment for D-T fusion.

Translating activation energy, via required relative particle collision speeds, to crazy plasma temperature is a favorite rhetorical device of tokamak promoters. Those who know what they are talking about know that their brethren at LHC routinely achieve collision "temperatures" approaching 10^17 K. (That would be 100 million-billion degrees, or 10+ billion times as "hot" as hoped for a tokamak, if it actually mattered.)

Another way to say "embrittled materials" is "the reactor". And "orthogonal" here means that completely solving plasma confinement would not help at all with the problem that turning it on destroys your power station in a few short months.

But none of this addresses the statement quoted at all.

So what's your opinion of the NIF? Even worse than tokamaks? Or is it actually more realistic as a future energy source?

NIF didn't work. It failed to achieve ignition. The simulation codes were bad.

A project called LIFE, to use NIF results to design an actual fusion reactor, was quietly terminated not too long after NIF failed.

https://en.wikipedia.org/wiki/Laser_Inertial_Fusion_Energy

Even if NIF had achieved ignition, it would likely, IMO, never have led to workable reactors. The targets are simply too expensive, and there's likely no way to protect the final optics (which have to be in line of sight of the targets) from hundreds of millions of explosions.

https://wci.llnl.gov/facilities/nif I don't think NIF was ever about energy research, just a way for government to spin weapons research as having some energy benefit.

Where can I read more about this?

Terms to search for:

1 Neutron Embrittlement

2 Helium Embrittlement

https://www.aps.org/units/fps/newsletters/201610/fusion.cfm

https://en.wikipedia.org/wiki/Neutron_embrittlement

https://arxiv.org/abs/1311.5079

https://www.sciencedirect.com/science/article/pii/0022311594...

Might as well add proton embrittlement to that list. Bane of hydrogen economies everywhere.

https://en.m.wikipedia.org/wiki/Hydrogen_embrittlement

The world economy manufactures and uses 50 million tons of hydrogen a year, so I think that problem has been well-solved.