[cletus]

The article mentions one big issue: plasma containment. It doesn't even mention what I think of as being the biggest hurdle to commercial fusion power generation: neutron containment.

Fusion hydrogen requires heavy isotopes, namely tritium, to generate sufficient reactions. This generates s lot of free neutrons however, enough that they will tend to destroy what container they're in. This is a significant, possibly commercially insurmountable, engineering problem.

Helium-3 is one alternative but is super rare on Earth ( even the heavier Helium-4 escapes the Earth's gravity once it reaches the atmosphere (so party balloons are consuming an irreplaceable resource thanks to an effective subsidy from Congress who narrowmindedly decided to offload the Strategic Helium Reserve at submarket rates).

People like to bandy about phrases like "free energy" when it comes to fusion. Well, free fuel and free energy aren't the same thing. A plant has a capex and running costs, a finite lifetime and a power output. Put those numbers together and you have a base energy cost even with free and essentially limitless fuel.

The article talks about producing tritium from lithium. Great. The demand for batteries is already going to stretch the worlds lithium supply so that's another advance we need.

[WorkLobster]

> The demand for batteries is already going to stretch the worlds lithium supply

The usual comparison here is that one laptop battery's worth of Lithium is enough to provide an individual's energy need for life. So at current consumption rates, there is enough Li available through traditional mining to supply the world with one thousand years of energy, and closer to one million years' worth if we recover it from seawater.

[kobeya]

> so party balloons are consuming an irreplaceable resource thanks to an effective subsidy from Congress who narrowmindedly decided to offload the Strategic Helium Reserve at submarket rates

I prefer to think of this as "providing incentives necessary to exploit the resources of Jupiter and Saturn."

Maybe when the history of space settlement is written it will have a chapter on the contribution party balloons and political squandering of strategic reserve had on resource prices...

[graphitezepp]

That is one way to look on the bright side. Fairly certain I read a write up on HN about why the helium situation isn't actually a big problems balloons aside with the crux of the argument being most of Helium's practical applications it can be effectively recycled in. Wish I remembered it better, chemistry is not my strong suit.

[kobeya]

The main industrial use of helium is in selling cold -- things that need to be done cold are done at liquid nitrogen temperatures. Things that need to be _really_ cold are done at liquid helium temperatures. Think superconducting magnets and MRI scanners. This usage can, should, and usually is recycled. Being a noble gas it doesn't really have much chemical applications per se. However according to wikipedia it is also used in "pressurizing and purging systems, welding, maintenance of controlled atmospheres, and leak detection." These usages are lossy by their very nature.

However helium leaks like almost nothing else. Only hydrogen is worse. It is so small that it is able to leak (slowly) through the crystal lattice structure of metals, ceramics, and other materials used to contain it. So recycling systems can't be made perfect.

And of course, even if recycling systems were perfect, a limited supply means no room for growth.

[CamperBob2]

Actually I'm not sure hydrogen is worse, because it's normally found in diatomic form. What's bigger, a single helium atom or a pair of conjoined hydrogen atoms with their own nuclei?

Hydrogen will leak through porous metals like palladium, but as you say, helium will leak through almost everything.

[dest]

If I'm not mistaken, the amount of lithium required for tritium production is quite small compared with the needs for lithium batteries

[CamperBob2]

The article talks about producing tritium from lithium. Great. The demand for batteries is already going to stretch the worlds lithium supply so that's another advance we need.

Can't imagine that this sort of process will use anywhere near as much lithium as the battery industry. It's fusion, after all.

[DennisP]

Helium-3 is the waste product of pure deuterium fusion. The YCombinator-funded fusion startup Helion is working on a hybrid D-D/D-He3 reactor, which it says will produce only 6% of its energy as neutron radiation. Deuterium of course is absurdly abundant on Earth.

Another aneutronic reaction is proton-boron. Boron isn't as abundant as deuterium but there's still enough on Earth to last tens of thousands of years. (The reaction use B11, which is 80% of natural boron.) pB11 fusion is especially difficult but several startups are trying it; the biggest is Tri Alpha, with about $500M invested. They attained stable plasma a year ago and just completed a larger reactor.

With D-T, the easiest reaction, there may be ways to engineer around the neutron issue. General Fusion does fusion pulses in the middle of a vat of molten lead and lithium. MIT's ARC design is more conservative, with a compact tokamak and modular construction. The inner wall is 3D-printed and replaced annually; they say after a couple decades the radioactivity will have decayed enough for cheap disposal. They surround the core with a blanket of molten FLiBe salt as coolant and breeding blanket.

[kirrent]

I listened to an interview that discussed ITER's blanket in detail recently and while it seems like a very impressive bit of materials science, they seemed to have no doubt it'd work.

Plus, won't nearly all neutrons be incident on lithium to breed tritium in a final reactor anyway? And you don't need very much lithium anyway so there shouldn't be a concern over lithium availability.

[shaqbert]

On the neutron absorption challenge: There are big advances in material science, especially with a FLiBe salt blanket fill. The challenge is more on the engineering side, i.e. how can you build it so that you can do maintenance and replace the blanket on a regular interval in a fully automated manner without interrupting the service of the fusion plant for too long.

[nikdaheratik]

Fission also generates a large amount of neutrons, but we still managed to build reactors it just required a large amount of shielding. This is a much smaller problem than finding a way to contain high energy plasma long enough to allow us to produce a fusion reaction.

I also don't understand why you're complaining about the loss of super-rare helium isotopes, but think that we're going to run out of Lithium due to this process, which is orders of magnitude more common.

The bottom line is that even a rudimentary fusion power system can capture 10x whatever energy you put in place. This blows away any other form of power generation no matter what measure you go by. Which is why it's still getting research money even if the result may be decades away.

[oakwhiz]

Is it possible to recycle old lithium batteries and extract tritium from that?

[ph0rque]

In fact, it's possible to recycle old lithium batteries into... new lithium batteries.

[rbanffy]

> neutron containment

Well... If you run a fusion reactor for long enough, you can dismantle it and use the scrap as fuel for much easier to build fission ones. ;-)

[raattgift]

You need a big enough reactor that a particular type of extreme dismantling produces reasonable amounts of fission-fertile radioisotopes; the bright side however is that the bigger such reactors are the less time "long enough" tends to be. And of course there's the time one might have to wait before beginning to mine them from the scrap...