Hacker News Reader

I'm a little dismayed that your first link is, as you note, from the 80s, and that your second link is about ITER, another old design that is orders of magnitude larger than the scheme in the article, a size it must have due to the very same technical limitation that the subjects of the article are working to remove.

But that is hardly the only disappointing thing about your post; It's just fear-mongering. Of course a development reactor isn't going to be "manufactured," but "built." What prototype has ever been manufactured before it was proven? One of the very points of fusion is its low proliferation risk. I don't know why you burden fusion with "extreme safety," either. Bicycles don't have "extreme safety," and we seem to do just fine. The impact of a fusion "disaster" is probably much less than a fall off of a bicycle.

He is pointing out a very important reality though. First you prove physical viability. Then engineering viability. Finally, after operating a full commercial unit for a while, you may prove economic viability. It's easy to assume physical viability leads to economic viability but this really is not the case.

If you can show by general arguments that engineering viability is unlikely, demonstrating physics viability is pointless.

it's a she.

Oh crap. I'm sorry. Thanks for the correction. Too late for it to let me edit.

Let's just toss out all physics pre 2000 because it's old. lol. Truth doesn't age. Just using ITER as an example. I can update that link with a Commonwealth one in a few years I suppose.

Manufacturability is a function of a lot of things like parts, systems, interactions, size, mass, supply chain, transportability, etc. Generally speaking, more complex things are less manufacturable and they become manufacturable by simplification and reduction. As a kind of measure of manufacturability, look at the cost. Think about the most expensive piece of technology that is mass manufactured today. It's probably the 777 or 737 - so price tag on order <$100M. Big ships are on order $10Ms. Maybe military procurement goes into the $500M, but those are low quantity and generally failed programs.

So if it's more than $100M and much more complicated than a wide body aircraft, today's economy and technology level is probably not going to be able to mass manufacture it.

The size and complexity of these fusion systems is physically constrained by various factors many of which are physics limited. And even with the innovations they propose but have yet to prove out, the size and complexity is still too large to manufacture.

Regarding fusion's proliferation risk: "One of the best ways to produce material for atomic weapons would be to put common, natural uranium or thorium in the blanket of a D-T reactor, where the fusion neu- trons would soon transform it to weapons- grade material. And tritium, an unavoidable product of the reactor; is used in some hydro- gen bombs. In the early years, research on D-T fusion was classified precisely because it would provide a ready source of material for weapons."

The big point Lidsky was making was that, for general reasons, DT fusion reactors will have lousy volumetric power density, compared to fission reactors.

So in addition to the points you make there, let's look at ITER's power density. Dividing the gross fusion power of ITER by the reactor (not plasma) volume, it will be about 0.05 MW/m^3.

In contrast, the power density of a PWR (gross thermal power divided by the volume of the reactor vessel) is about 20 MW/m^3.

ITER is worse by a factor of 400. Lidsky was being generous to fusion reactors, compared to this. MIT's ARC concept is about 0.5 MW/m^3, also much worse.

Lidsky's point is still valid, 35 years later, because it was based on generic arguments. Pfirsch and Schmitter were making basically the same argument in Europe at the time. Had they all been listened to!

MIT us using 8MW/m^3 for DT. Apparently pretty basic physics and it applies over a wide temperature range.

https://www.youtube.com/watch?v=rY6U4wB-oYM&t=335

I believe that is using the volume of the plasma, not of the reactor. If you wish to go there, the power density of a PWR core is 100 MW/m^3.

Yes. The only number I could find for a Uranium PWR core was 69MW/m^3. Lidsky said in 1983 that "the power density is only one-tenth as large" (fusion vs. fission) so I assume this is what he was talking about. I only wanted to compare apples to apples.

Obviously, additional equipment surrounding the core or plasma adds to the volume, but the major portion of the construction cost is the steam equipment and power generators which is the same cost and volume per MW for either technology (or for fossil fuel power plants for that matter.)

The major part of the cost of a fission plant is the non-nuclear part, but I don't think that would be true of a fusion plant.

In any case, if some parts are common, but the non-common parts are cheaper in a fission plant, fusion will be more expensive than fission -- and since fission has already lost in the market due to its cost, fusion would do likewise.

>I've heard that the magnets are underperforming by about a factor of 2 - and that's 4 years into the project.

Hi - could you pls point me to the source for this? Would be great to see actual data on where we are on these projects. thx

There are no "magnets" at this point, so not sure what he is talking about. Basically they are building up HTS Twisted Stacked-Tape Cables (TSTC) and they plan on building their first full-size prototype coil in 2020.

Current tests include subjecting conductors to a 13T field in a 700mm chamber at the National Institute for Fusion Science (NIFS) in Japan. One of the key properties of HTS is that they maintain superconductivity at high magnetic fields.

Some slides from a recent talk discussing this phase of research: https://indico.cern.ch/event/775529/contributions/3309887/at...

Overheard a wailing researcher at a bar in Boston. Sorry.

gotcha thank you.

yep, at least until we get aneutronic fusion going with direct schema of electricity generation bypassing the heat engines, the fusion on Earth has no chances of competing with the most of the existing and especially solar/wind energy sources. Other than that the 2 most interesting potential applications are weapons and space propulsion. Those applications are better served by inertial confinement (laser, Z-machine, fusor, DPF) which historically been showing better results and usually at much lower costs (and even the high cost NIF can be redone today using modern solid state which would lower the cost/size dramatically) yet historically the resources have been poured into the huge systems like Tokamak.

> yep, at least until we get aneutronic fusion going with direct schema of electricity generation bypassing the heat engines, the fusion on Earth has no chances of competing with the most of the existing and especially solar/wind energy sources.

This sounds like unfounded opinion to me. While I agree that ITER derived designs are unlikely to compete economically, that doesn't prevent alternative designs, especially hybrid fusion/fission sub-critical reactor's, from competing.

For those not familiar, here's the wikipedia page on hybrid fusion-fission [1]. The idea is to use a sub-break-even fusion reactor as a source of very high energy neutrons that would trigger fission in uranium or thorium isotopes that are otherwise non-fissile. For example U-238, which constitutes 99% of all uranium on earth, and which is currently the dead weight as far as uranium is concerned (worse, you spend a lot, a lot of money to separate U-235 from U-238). U-238 can undergo fission when targeted by fast neutrons, but can't sustain a chain-reaction, so fusion-fission hybrid reactors can't lead to a Chernobyl event (but a Fukushima is possible).

Not clear why we don't hear more about fusion-fission hybrids on HN. It sounds like such a promising technology.

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

Hybrid fission-fusion reactors combine the worst features of both. Why would anyone want one?

If Apollo 11 had found oil on the moon, we would have a few thousand people living there already. There is just nothing profitable in going to the moon.

With fusion the opposite is true: we know energy is precious, and we know its market price. Just prove it can be done and commercialization will happen very quickly.

> If Apollo 11 had found oil on the moon, we would have a few thousand people living there already.

You're forgetting the most important piece of information in the economy: price.

Since oil produced on Earth is cheap, compared with hauling it back from the Moon, we don't get any from the Moon.

Now, if we had found some exotic energy source that didn't exist on Earth at all, that would be a different story.

Hauling it back from the moon wouldn't be overly expensive. It'd be getting there in the first place that would.

> Now, if we had found some exotic energy source that didn't exist on Earth at all, that would be a different story.

Which does exist, btw: helium-3 which is deposited by the solar wind and captured in cold traps on the lunar poles. It doesn't exist on Earth in any quantity, except artificially produced in nuclear laboratories. He3 fusion is easier to achieve and less dangerous to operate.

Sounds like an excellent reason to go to the Moon :-)

and build a fusion reactor =D

Which is relatively trivial to do with He3, which is rather the point :)

He3-He3 fusion produces no neutrons, so it requires no mass shielding. The energetic particles that result are electrically charged, so their kinetic energy can be directly turned into electricity, which further adds to efficiencies and avoids all the real engineering problems that come from coupling power generators to the fusion plasma in some way. The ignition temperature for He3 fusion is higher than the traditional D-T fusion, but it's a heck of a lot easier to achieve high temperature containment when you don't have to worry about things like shielding, and that shielding becoming radioactive from neutron bombardment. A great deal of the engineering challenges that plague D-T fusion simply don't show up in He3 fusion.

However the way that you get He3 in large quantities, on Earth, is through D-T fusion. So He3 fusion being easier doesn't help much because to get the fuel you have to do the harder D-T fusion first... unless you go to cold traps on the Moon, or mine the atmosphere of Jupiter/Saturn.

3He-3He fusion is probably impossible. The 3He fusion reactors being talked about would use D-3He, and would still produce neutrons.

The two big issues with neutrons are damage to the reactor structure, and induced radioactivity preventing hands-on maintenance. D-3He could help with the first, but not the second. A D-3He reactor would still have to be maintained remotely, with robots. At best, it would reduce the radiation load on the robots.

The need to maintain fusion reactors with robots reminds me of what they did at the hot cells at Hanford to ensure they could be maintained after being used for reprocessing. They required that the operators install all the equipment there using the remote handling equipment (mechanical waldoes, not robots). I'll believe the fusion people can maintain their reactors when they do the same thing.

Unlikely:

"The helium-3 incantation" - http://www.thespacereview.com/article/2834/1

tl;dr the popularity of mine-the-moon-for-he3 meme is due to wishful thinking about sustainable moon colonies, and there's no analysis that suggests it's even remotely practical to do.

That article is bogus. Let's take the outline points:

"There are no fusion reactors."

Well, duh. Fusion power isn't that far off though. Even Lockheed has a self-funded fusion power program being run out of Skunkworks, which isn't known for ivory-tower boondoggles. Even so, part of the point is that He3 fusion is different from D-T fusion, which brings us to...

"Helium-3 fusion is even more difficult than regular fusion."

No, He3 fusion ignition temperatures are higher than D-T fusion. But in every other way aneutronic fusion reactors are easier to build and maintain than neutron-generating D-T. They can be smaller, lighter, generate less radioactive waste, permit more efficient direct electrical current generation, and require simpler electrostatic containment than Tokamak-like designs. You can literally build one in your garage.

"Helium-3 may be very difficult to locate and mine on the Moon"

This was invalidated by LRO's data on the cold traps in permanently shadowed craters on the Moon (and MESSENGER's data on Mercury), which were then validated by simulations of the Moon's exosphere and interaction with the solar wind. The regolith in these craters is as much 40% volatiles by weight, of which a economically extractable fraction is He3 from the solar wind.

The huge point against lunar 3He, even ignoring that it's two orders of magnitude less reactive than DT, is the difficulty of extracting it on the moon.

3He occurs in regolith in concentrations measured in the ppb. Even with beneficiation of fine fractions and efficient recycling of heat from the thermal extraction step, the energy requires are large, a significant fraction of the energy the 3He would produce when fused.

So to power the terrestrial economy with 3He, you need to put a significant fraction of the terrestrial energy output on the moon to get the 3He. Power plants on the moon will be much more expensive than on Earth (because labor, materials, and supply chains will be more expensive there, even with cheap space transport), so this is unlikely to pencil out.

Your information about the composition of the moon is 30 years out of date. He3 would be mined from the cold traps in permanently shadowed craters at the north and south poles of the moon. There it exists in extremely high concentrations, alongside other useful volatiles that together make up 40% of the surface material by weight.

“Refining” is a process as simple as shoveling it into a pressure container, heating it, and then letting dissipatively cool back to ambient temperature, collecting the other volatiles as they condense out. What’s left is a mix of inert gasses, including He3.

There is no data on 3He at the poles. The area there is small anyway.

The concentration of 3He on the moon in general is such that simply heating regolith without heat recycling and separation of fines would use more energy than the 3He would yield if fused.

Oil on the Moon wouldn't be hauled back to Earth, it would be used for an independent space based industry that produces things that are harder to do on Earth.

Not for energy, it wouldn't - good luck burning anything on the moon!

I was thinking of fancy plastics. Perhaps bring-your-own oxidizer rocket fuel.

You're missing the point. We know there are asteroids with price changing amounts of precious metals - we aren't mining them.

Commercialization involves so much more than proving it can be done physically. And it also involves more than meeting a price in a padded environment. There are hurdles to jump that are totally disconnected from the price.

> We know there are asteroids with price changing amounts of precious metals - we aren't mining them.

Are there?

As I understand it, the mass of the asteroid belt is primarily confined to just a handful (about 12) of objects, and everything else is so diffuse that it would never be profitable.

I think they are going for the smaller ones anyway, given they are looking to fetch and park them closer for the actual mining. If you are patient and know some orbital mechanics, it takes surprisingly little energy to move stuff around up there. If you can automate it and be sending missions out regularly, sure it takes a while to get your first asteroid back, but after that you can have a steady supply.

Yes, there's a dumb amount of precious metals up there. If you were to bring back the contents of a single asteroid you'd crash the market. Here's a master's thesis on the topic: http://digitool.library.mcgill.ca/R/-?func=dbin-jump-full&am...

Yeah, but if you look at that paper, platinum is extracted at roughly $20,000,000 per tonne--roughly the price of a single rocket launch right now. And we extract only 200 tons per year. That's a $20 billion market, but it's so sensitive to supply that nobody is rushing to open more mines.

Presumably, any public or insider information about a credible funded technologically-sound attempt to mine the asteroid would crash the market first. So, the first sign of such an attempt would be huge shorts in the paper commodity markets by the people funding the mission. They would have to make their money on the financial hedging, not on the actual sale of mined PMs.

Perhaps it also means that Bitcoin has a more credible claim to a limited future supply than traditional PMs with a "proven 5,000 year track record".

> Just prove it can be done and commercialization will happen very quickly.

You seem to be saying "just show Q high enough and we're home free to commercially successful fusion power".

But this is totally wrong. Fusion faces profoundly difficult obstacles to success even after the physics problems have been solved. These may be "mere engineering" problems, but engineering problems are perfectly capable of sinking a technology.

I figure this sci-fi punchline is relevant:

> Rod: Your kind began as self-repairing distributed storage systems for our supercomputers. When the cities fell, some of you survived, and changed just enough to continue to survive and evolve. You are now at least a million generations removed from anything we created.

> Kevyn: You created? You don't mean you personally, do you?

> Rod: Me personally? What do you think I look like? Some kind of an engineer?

> Kevyn: Actually, What I thought was that you looked really, really old.

> Rod: Heh... You youngsters, are always so cute with your little accidental insults. No, I didn't create you. Engineers did. I created something far more powerful... The marketing campaign that made you profitable.

Source: https://www.schlockmercenary.com/2001-12-16