Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

MIT-designed project achieves major advance toward fusion energy

news.mit.edu

Post reply on HN
Print view View original

171–180 of 438 posts

Re: MIT-designed project achieves major advance toward fusion energy

#171

Post by ncmncm »

mLuby wrote: post #92
cletus wrote: post #87

I've long been skeptical of ITER making any sense given its insane cost. I mean even it succeeds, then what? Here's the truth: there's no such thing as free energy. Even if the fuel is so abundant it's actually or effectively free (eg deuterium), the energy isn't. Say it takes $50B to build a plant that produces 1GW of power, which I'll estimate at about 7TWh/year based on [1]. Let's also say it has a lifespan of 40…

It seems you're suffering from "neutron embitterment." ;P Space-based solar power generation (itself "fusion power" in the loosest sense) would be great in the inner planets. Though to open up the outer planets, Kuiper belt, Oort Cloud, and any other stars, we'll need non-solar* power: hopefully fusion, at least fission. *Unless we want to go the stellaser route, but I'd bet we'll crack fusion before getting near K2.

It won't be Tokamak fusion in any case. FRC (burning D+H-3) might work, but there is no money for it. Neutron-emission fusion eats all the fusion money.

H-3 is not nearly so scarce as cletus suggests. It is uncommon, but you don't need much.

Re: MIT-designed project achieves major advance toward fusion energy

#172

Post by snek_case »

worldvoyageur wrote: post #94
anonuser123456 wrote: post #17

Earlier quoted context omitted.

> What's the catch this time? This is D-T fusion. Which means you have to have T. Which currently comes from fission reactor and has a half life of 15 years. So the plan is to use a molten salt blanket with Be to breed T. But Be isn’t scalable for consumption, so maybe lead eventually. That’s probably do-able, it just slows down the rate new reactors can come online since Pb is not as good a neutron multiplier. Once…

Tritium is a natural byproduct of CANDU fusion reactors, of which there are some 25 or so in operation globally, mostly in Canada. CANDUs use heavy water as a neutron moderator (D20 instead of H2O), making T2O a natural byproduct. Though most of the reactors do not harvest the tritium, a small number do. CANDU operators have long been ready to make the capital investments in tritium harvesting, once demand materializ…

I'm wondering, fusion reactors themselves produce neutron radiation as a byproduct. Once you have a fusion reactor running, could you use the fusion reactor itself to breed tritium?

Also thinking, we target deuterium + tritium fusion because it's the least energy intensive. However, once we have working proof of concept reactors, could we just make them slightly bigger and fuse more abundant molecules/isotopes instead?

Re: MIT-designed project achieves major advance toward fusion energy

#173

Post by ncmncm »

gibolt wrote: post #110
cletus wrote: post #87

I've long been skeptical of ITER making any sense given its insane cost. I mean even it succeeds, then what? Here's the truth: there's no such thing as free energy. Even if the fuel is so abundant it's actually or effectively free (eg deuterium), the energy isn't. Say it takes $50B to build a plant that produces 1GW of power, which I'll estimate at about 7TWh/year based on [1]. Let's also say it has a lifespan of 40…

Your numbers sound like generation 1 numbers, after ITER. ITER is only a test facility to prove hopefully that it can be net positive. However, those maintenance costs (your estimates) would be the first thing to drop. Any company producing/operating these will be competing with wind and solar, and thus highly incentivized to improve. There should be plenty of low hanging fruit, since it hasn't happened once yet.

Maintenance cost is not a place to expect major cost reductions. Those tend upward.

Re: MIT-designed project achieves major advance toward fusion energy

#174

Post by snek_case »

mcswell wrote: post #95
anonuser123456 wrote: post #17

Earlier quoted context omitted.

> What's the catch this time? This is D-T fusion. Which means you have to have T. Which currently comes from fission reactor and has a half life of 15 years. So the plan is to use a molten salt blanket with Be to breed T. But Be isn’t scalable for consumption, so maybe lead eventually. That’s probably do-able, it just slows down the rate new reactors can come online since Pb is not as good a neutron multiplier. Once…

"Hydrogen is very corrosive and hard to work with" Corrosive compared to what? You can put it in a rubber balloon and hand it to a kid. "T is radioactive hydrogen": True, it emits low energy beta radiation, which is an electron, and is stopped by a sheet of paper. I used to have a wrist watch with a tritium dial; I haven't died of cancer yet.

Corrosive is the wrong word, but hydrogen is such a small molecule, it can leak through metals and weaken them, as I understand it. It's hard to contain.

https://www.imetllc.com/hydrogen-embrittlement-steel/

Re: MIT-designed project achieves major advance toward fusion energy

#176

Post by baking »

phicoh wrote: post #25
ChrisMarshallNY wrote: post #6

Earlier quoted context omitted.

The running joke has always been that “Fusion is 20 years away,” and has been, for the last 50 years. I really want this to work. I am a bit concerned, with how “the old guard” will react, once we have successful, productive, fusion. I foresee an astroturf NIMBY campaign against construction of fusion plants.

At the moment it is very far for clear that fusion will be cost effective. The article says this about the fuel of fusion: "The fuel used to create fusion energy comes from water, and “the Earth is full of water — it’s a nearly unlimited resource. [...]" They forgot to say that it is not the H2O that comes out of your tap. The earth is especially not full of tritium.

The fuel is deuterium, lithium, and a smaller amount of beryllium or lead used as a neutron multiplier to assure a net positive tritium production. Yes, tritium is needed to start the reactor, but it can be replenished with lithium tritium breeding in the blanket.

Deuterium is plentiful in tap water.

Re: MIT-designed project achieves major advance toward fusion energy

#177

Post by apendleton »

pfdietz wrote: post #86
apendleton wrote: post #37

Earlier quoted context omitted.

Is that the right metric? You wouldn't need to build a huge containment structure around it like you would for a PWR, so I'd imagine the power density of the plant as a whole wouldn't be anywhere near 40x worse. Why focus just on the primary reactor vessel?

Because the rest of the power plant will be similar (or also worse for fusion; consider the tritium handling facility or the robotic equipment for maintaining the fusion reactor). You would need a confinement structure, just to keep the tritium in (which will leak all over even in normal operation). So if you swap out a cheap PWR reactor for a much larger, and hence much more expensive, fusion reactor, you get a powe…

I guess I still just don't really see it... like, coal plants are also much less energy-dense than nuclear fission. So is solar, so is wind. We build all of those anyway. There are lots of things other than power density that contribute to whether or not a particular generation technology is economical.

As to why massive fission reactors aren't built: there are plenty of already-available passively-safe/meltdown-proof fission designs (many gen-IV designs qualify), and from what I can tell, the reasons they're not built are as much political as anything -- people don't like them, and the consequent regulatory regime has made any fission projects prohibitively expensive regardless of their size. None of this need be the case with fusion.

As to tritium: I think you're overstating the tritium risk. They're only dealing with grams at a time, and even if it all leaked out, it would rapidly diffuse such that risk to the public would be infinitesimal as compared to normal background radiation (plus its half-life is only something like 12 years). ITER has a safety page: https://www.iter.org/mach/safety that essentially says as much.

Re: MIT-designed project achieves major advance toward fusion energy

#178

Post by danans »

UnFleshedOne wrote: post #75
danans wrote: post #55

Earlier quoted context omitted.

> I wouldn't call it a "meh", even if it comes off as much more expensive than fission. It's not competing with fission, though. It's competing with renewables + storage + load shifting + efficiency. Compared to those, it might indeed be "meh".

In the grid this would take role of coal or gas plants for base load, no?

Storage at sufficient scale would supply some of what we refer to as baseload today, much as hydro provides baseload power in many places today.

Re: MIT-designed project achieves major advance toward fusion energy

#179

Post by loufe »

gfodor wrote: post #140
JDDunn9 wrote: post #131

Even if we can get fusion to work, it will never be economical. Just because the fuel (water) is free, that doesn't make the energy free. The fuel rods for fission power plants are already a rounding error in the cost of energy. It's the capital costs that dominate the equation, and fusion plants will be at least as expensive as fission, which is more expensive per KWh than solar. https://thebulletin.org/2017/04/fusi…

You’re citing an article from 2017 talking about a reactor design from 1988 in response to an article about novel fusion technology from 2021.

I agree that the sourcing does seem off in JDDunn9's post but your comment doesn't invite further discussion much.

Re: MIT-designed project achieves major advance toward fusion energy

#180

Post by carbonguy »

snek_case wrote: post #172
worldvoyageur wrote: post #94

Earlier quoted context omitted.

Tritium is a natural byproduct of CANDU fusion reactors, of which there are some 25 or so in operation globally, mostly in Canada. CANDUs use heavy water as a neutron moderator (D20 instead of H2O), making T2O a natural byproduct. Though most of the reactors do not harvest the tritium, a small number do. CANDU operators have long been ready to make the capital investments in tritium harvesting, once demand materializ…

I'm wondering, fusion reactors themselves produce neutron radiation as a byproduct. Once you have a fusion reactor running, could you use the fusion reactor itself to breed tritium? Also thinking, we target deuterium + tritium fusion because it's the least energy intensive. However, once we have working proof of concept reactors, could we just make them slightly bigger and fuse more abundant molecules/isotopes instea…

> Once you have a fusion reactor running, could you use the fusion reactor itself to breed tritium?

I'll have to find the citation, but IIRC the answer is "theoretically, yes" - the concept is that molten lithium could be used in a tokamak to absorb neutrons and produce tritium at the same time.

EDIT: Here are two citations I was able to find quickly - it looks like one of the ITER experiments will be to validate the concept [1] and that this could also be the way that heat is removed from the reactor. [2]

[1] iter.org/mach/TritiumBreeding

[2] https://www.euro-fusion.org/faq/top-twenty-faq/what-is-a-lit...

Post reply on HN