Hacker News Reader

Comment by hinkley | Hacker News Reader

What are you going to ignite? We need four more planets to get enough tritium to power a single first world country. It’s bullshit all the way down.

View on HN

Your pessimistic take got me researching. Here’s a great article that describes the issue and some potential workarounds: https://www.science.org/content/article/fusion-power-may-run.... Your assumption rests on the fact that the “winning” tech will use tritium. And yes, maybe that’s the most viable path, leading you to your take. But it seems there are viable paths forward with and without tritium.

Tritium is the basket most of the eggs are currently in. So if you’re using “other fuel sources” as a counter argument for “fusion is 30 years away and always will be” then I’m sorry but you’re just as much of a bullshitter as they are.

Switching fuels is at least one full international collaboration and two generations of reactor away, and if you think that’s going to happen in fifteen years then boy are you gonna be disappointed by geopolitics.

Who said anything about a full fuel switch? All I’m talking about is your claim that this technology will never happen. I shared an article with potential workarounds. And sure, it still might not work long term. But that’s impossible for you to say with certainty. I have no horse in this race aside from being a human that would benefit from low cost energy. I know nothing about fusion beyond what I read in that article, and a few others before it. You can call me names all you want, but that doesn’t change how innovation works. Something is impossible until it isn’t. Many have come before you with similar attitudes with other seemingly impossible technologies, and have been proven wrong in unexpected ways. I’m not waving my hands and saying “this will definitely happen!” I’m saying it could happen, vs you saying it’s a certainty NOT to happen. And again, I’m talking about the technology itself, not the geopolitical, infrastructure, economic, etc. issues around it.

Fusion reactors generate tritium generally as I understand it. A more real risk is we run out of tritium running experiments before we figure it out OR before we use something else. However, even if we ran out, it just means the cost of tritium is higher (very energy intensive) vs totally out of reach.

They can generate tritium, by capturing escaping neutrons in lithium contained in the blanket wall. I believe the second law of thermodynamics means that we can never breed tritium faster than we consume it. (someone with more insight correct me on this)

That’s apparently what the beryllium is for. It splits and makes more neutrons.

The overabundance of neutrons cause the beryllium to become radioactive waste, thus eliminating a supposed major benefit of fusion.

Well the cladding and the lithium for sure. The wikipedia page on beryllium seems to think it breaks down into 2 alpha particles and some neutrons.

But then I'm finding that the naysayers are pointing out that nuclear physics is a big statistics game and you don't get to cherry-pick which reactions happen. Any one that can happen does and you have to be able to mop up/separate/repair the consequences of each.

Is Wikipedia right about beryllium, or do you know of other reactions that render it nastier than some ionized helium?

Pure beryllium (few enough trace contaminants to not activate) is very very very very expensive. And you need a lot. And it's pretty nasty on its own even compared to irradiated steel or similar. Highly experienced nuclear engineers need regular safe handling training because it's so toxic.

As best I can understand, yes, it's a neutron multiplier, and some of those additional neutrons will end up interacting with the lithium and produce a net gain. In exchange you get a couple of alpha particles.

Where did you get that belief from? Tritium isn't found in nature, the half life is way too short.

We can make it, one of the ways is D-D fusion.

You got a D-D fusion plant in your pocket, or are you just happy to see me?

The way ITER is planning to get tritium is by bombarding a lithium lining with neutrons. But the neutrons tend to miss lithium and so they’re using a beryllium lining on top of that to multiply neutrons. So you’re making a tritium gas that’s on the wrong side of a very expensive barrier, how are you going to harvest that from a power plant that’s supposed to run 24 hours a day?

Slowing neutrons is one thing. You can do that with plentiful stuff. Creating new ones means donor material. You’re consuming the beryllium. Almost nobody produces beryllium, and I don’t think 400k tons of the ore - not beryllium, ore - is going to be sufficient for a consumable. A highly toxic, easily oxidized one at that.

ETA: the reaction for neutron multiplication using beryllium apparently splits the beryllium into a neutron and two helium. The lithium reaction produces tritium and more helium.

> You got a D-D fusion plant in your pocket, or are you just happy to see me?

No, but it's (genuinely) been on my to-do list for while.

But bigger than a pocket, more like very large rucksack at best, probably more like CRT monitor sized.

> So you’re making a tritium gas that’s on the wrong side of a very expensive barrier, how are you going to harvest that from a power plant that’s supposed to run 24 hours a day?

No idea, but I used to live in Cambridge, and IIRC one of locals did his PhD on neutron induced atom dislocation within metallic crystals specifically with regard to the engineering of fusion reactors, so I can say with a certain level of confidence both that (a) I wouldn't understand the answer and (b) that it wouldn't fit into this comment box.

I had a friend in college who worked with a physicist on femtosecond lasers. What’s a femtosecond, I said. It was about fifteen to twenty years before they were commercialized.

If your friend did his PhD on something then the clock might have started ticking on that tech but many, many things that work in a lab never get used to make products or medicines. For everything else it takes around twenty years, which is a huge fraction of 30.

I moved there in 2007, and moved away in 2017. This would've been the earlier part of that period, given the geek pub didn't really recover that reputation after the fire in 2010.

Reply on news.ycombinator.com