Nuclear fusion is close enough to start dreaming | Hacker News Reader

Nuclear fusion is close enough to start dreaming

bloomberg.com

Recent and related:

Helion - https://news.ycombinator.com/item?id=29119180 - Nov 2021 (249 comments)

Just for general information, YCombinator is an investor of that particular "new" company.

Nuclear fusion isn't a new thing, many already established companies are working on it, making the headline and focus on this particular one is disingenuous, to say the least

https://www.helionenergy.com/who-we-are/

The OP specifically mentions Helion, so it made sense to link to a recent thread about Helion. Otherwise I wouldn't have bothered, because there are too many past threads about fusion in general: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que....

I had never heard of Helion and had no idea it was a YC startup until I saw Sam's article (i.e. the thread I just linked to) on the front page like any other user. YC has invested in thousands of startups. Many get discussed on HN. The community has a close relationship with YC and with prominent ex-YC figures like PG and Sam. Nothing about this is hidden or unusual; it's how HN has been ever since its origins 15 years ago.

If you'd like to post an interesting article about a different nuclear fusion startup, project, or development, that would be great. They're just as welcome.

Helion is the only one I know that's saying they'll "demonstrate net electricity production, as early as 2024." Maybe they will maybe they won't but it makes a change from the usual "maybe in thirty years" stuff.

https://archive.md/3b3Oy

> The nuclear fusion startup Helion, which announced last week that it has raised $500 million, says it has developed new technologies that may make nuclear fusion viable — practically, economically and environmentally.

This is the only actual news in the article. The rest is bog-standard speculation.

Note that seven years ago, Helion said they were three years away from energy positive fusion. Today, they say they're... Three years away from energy positive fusion.

They said they were three years away given investment, which they did not get. Now they have the investment they said, then, such progress would depend upon.

Tokamak eating all the fusion research capital has much to answer for. The only sense I can find in Tokamak exclusivity is that aneutronic fusion research does not maintain employment of a population of hot-neutron physicists that weapons work would need to draw on.

The SPARC/ARC thing seems to make sense as a research project. I'm not sure ITER is a good use of funds though.

I am dead certain ITER is not a good use of funds.

At least, for any public goal. Certainly the people--military contractors, mainly, I expect--whose pockets the $billions are going into may be justified in considering it a good use. Not a single penny disappears; each goes into somebody's account.

Thanks, glad I didn't buy a subscription to Bloomberg for that.

To us older and more cynical old farts, the story has been: "Fusion power is twenty years away". It's been that for nearly 70 years. I don't see anything changing that elderly estimate for at least another 50 years. (And that's being optimistic.)

Helion is changing that story to fusion power always being three years away.

Thanks.

Note it’s an opinion piece published on bloomberg, not an actual bloomberg article.

I don't get the reason for saying this. They published it, without rebuttal, they must support it.

Some newspapers (I don't know if Bloomberg is included) will publish opinion pieces from multiple sides (not necessarily at the same time). If in a few months, Bloomberg publishes a "no fusion" opinion piece, this will not mean Bloomberg is "now opposed to fusion". Rather, Bloomberg publishes a wide range of opinion pieces as they should.

"May" is pretty specious. Is there any evidence of something likely to work? I would love to see practical fusion, but I see no reason to believe it will exist soon. Someone has to figure out how to actually make it practical first.

In magnetic confinement fusion, power density scales to the fourth power of the magnetic field strength. The only way to reasonably achieve the kind of magnetic fields needed without massive constant power draw is superconducting magnets. For a very long time, the best superconductors available limited the maximum field strength to ~ 5-6T. This limited the space of possible magnetic confinement reactors into basically ITER, or "big enough that you need a major multi-decade international collaboration to build it".

Last decade, major discoveries were made in practical high-temperature superconductors. By 2017, you could commercially purchase superconducting tape that can do >8T, the state of the art has improved since, and keeps improving. All the constraints and limits on fusion that were known since the field began are changing now. Most importantly, with high-field magnets, the size of a theoretically functional net-energy fusion reactor comes way down, to what is achievably by a reasonably well funded startup. The whole field is open now, and all the different magnetic confinement fusion technologies that were tested and discarded over the first 60 years of fusion research are now being re-evaluated in ligth of the new magnets that are available. That's why there are now like a dozen well-funded fusion startups.

Do you mean that ITER could be obsolete before even reaching a production phase ?

ITER is absolutely obsolete, in any sense requiring it to be a stop on the way to production. But it was never intended for that. ITER could, if ever completed, provide a wealth of research results on plasma confinement. But only that. And, it is likely to be too late to the party.

Absolutely, as is the case in any huge project, it takes long enough to build that at some point you have to freeze the design and technology keeps progressing. It's hard to predict which pieces of technology will mature on what timeline, and at some point you just have to start.

That ITER doesn't use superconducting magnets to me means that it will definitely be "obsolete" in the sense of using the most modern technologies, but it's still going to do the job it was designed for.

Iter uses superconducting magnets. They are the best superconducting magnets that were available when the design was finalized in 1998. Modern ones are much better.

Sorry, yes, you're right, I meant modern ones.

Yes and no. Yes in that the technology will be stale and there will be far superior materials science, electronics, etc, by the time it is online.

No, in that its principle purpose is fusion and plasma research. It'll still be able to do science.

It already is, they're still using Nb-Sn superconductors for ITER. The whole project is a manifestation of the sunk costs fallacy.

Sounds like the story of deep learning once computing power cheapened.

Their current prototype, no 6, reached 100 million C. The next, prototype 7 they say should produce net electricity in 2023 or so. We'll see. That's not commercial electricity production but any production would be a step forward.

I’m an interested layperson at best, but I gather that we know how to build safe fission reactors now, and that the safe storage of spent (or “slightly used” as I’ve heard it called) fuel pellets is a much harder political problem than technical problem.

Gorbachev once speculated that it was actually Chernobyl that brought down the USSR.

I personally wonder if it’s also why we’re going to have a climate catastrophe.

What to do with spent fuel rods is more of a political problem than a technical one. Read up on Yucca Mountain.

Current thinking is that the best approach to long term disposal is to find a cold hard-rock mountain where nothing has changed much in millions of years. Dig tunnels that are far above any possible flooding. Drill vertical holes in the tunnels. Package the nuclear waste in stainless steel thimbles, with the waste embedded in ceramic. Place thimbles in holes. After a few decades of nothing happening, fill in the tunnels.

Sweden and Finland, which have lots of stable, isolated hard rock mountains, are working on this. So is China, which now has two vitrification plants for nuclear waste.

Why don’t we just end Carter’s prohibition On reprocessing the spent fuel rods?

Why do the tunnels need to be above any possible flooding? With the waste contain the way you described it being submerged in a flood doesn’t seem like it would cause problems.

Stainless steel is not forever in the presence of salt water.

Ceramics, though, hold up well over millenia. The oldest ceramic pot known is 16,500 years old.

First question I have about it: wouldn't it be an attractive target for bombing/terrorist attacks or would it be rock solid enough to withstand those?

If you have weapons able to blow up tens/hundreds of meter thick rock, then you have much more attractive targets with more "spectacular" results.

The threat of terrorism is massively overblown across the board, not just with respect to nuclear waste. The means to make conventional explosives are widely available, yet outside of a handful of conflict zones it is rarely done. Americans can buy gunpowder at walmart, yet how many pipebomb attacks have you heard of? A handful, but not many. Not many considering the country has hundreds of millions of people with ready access to the means to make them.

The incidence rate of people inclined to terrorism is vanishingly low. And when you start talking about plots that require numerous people to cooperate, rather than a single lonewolf nut, there is even less to worry about. The more people a plot requires, the less likely it is to actually happen. Coordinated attacks against nuclear power plants or waste disposal facilities are fantasy.

>The threat of terrorism is massively overblown

There's some temporal sampling bias, here.

The threat of terrorism was huge and salient in 2002, and a lot of NRC regulations were updated at that time. It was less obvious in 2010. Here in 2021 it's hardly a threat at all. If we see another 9/11-scale attack this decade you can bet that will change.

This is true, particularly in developed nations. [0]

[0] https://ourworldindata.org/terrorism

I gather that we know how to build safe fission reactors now.

One can hope. "Gen III" fission reactor construction is coming along, mostly in China. Sanmen Nuclear Power Station has two Gen III reactors, designed by what's left of Westinghouse. Future versions are being entirely designed in China.

Most of the "small cheap reactor" concepts are arguments for why they don't need a containment vessel. That's probably not going to sell. Chernobyl wasn't supposed to need a containment vessel, and didn't have one. Fukushima didn't have a big enough one. Three Mile Island did have a good containment, which is why that meltdown was expensive, but didn't cause much trouble outside the plant.

I think Hiroshima also is a big factor.

Chernobyl, too, sure, but if Hiroshima hadn't already shown the terrible risks of "nuclear", I think it would have mattered a lot less.

Lftr, no stupid spent rods. Online reprocessing, thorium fuel, 99% fuel use. No meltdown risk. No huge containment domes. Scale in size.

Lftr can use old waste for fuel or breed yo usable fuel.

Transport of spent fuel rods is a legitimate problem.

Think of all the fusion powered direct air CO2 capture we could have <3

When you find yourself in a hole, stop digging.

The first landmark is to make coal plants unprofitable. Once all (most of) the coal and oil plants are closed, it would be interesting to use the spare electricity for carbon capture. Otherwise you are burning coal to power the carbon capture facility.

I don't know, we might need technology to save us from our seemingly bottomless stupidity. Maybe in 2050, Robo-Joe Manchin will secure a permanent government slush fund for Greater West Virginia coal mining and we'll have to stick a fusion-powered CO2 capture facility right next to where they burn it.

Don't disagree but we are already at the stage where we have the technology to solve climate change , what we have is a problem of status quo-ism

I just don't know if we should firmly say "we have enough technology so now the only problem is how to make humans amenable to radical change". It's frustrating but maybe "more technology" is a more practical path that can still get us to the same place.

>Think of all the fusion powered direct air CO2 capture we could have <3

Oh, trees? ;-)

This popped up on HN reasonably recently https://caseyhandmer.wordpress.com/2021/05/04/so-you-want-to... - basically "it's not necessarily as simple as that, where are you getting the water from?".

Is this still an issue when fusion power is available?

No, but the parent to my comment implied trees as being the source of that fusion power. Nobody's going to be able to use trees to power the desalination plants that allow us to irrigate the land so that we can grow those trees.

People have been growing coppice forests with salt tolerant trees such as willow for ages. No need to desalinate, the trees do it themselves. You can also use algae, which can grow in salt water.

The fusion energy that is powering the system is that of our sun. No perpetual motion machine required, our planet is being bathed in outside energy.

For a tree to be an effective long-term sequestered of carbon, you have to ensure the tree never burns or decays. Obviously this can be done; you can sink it in anaerobic water or bury it deep in a mineshaft... but just letting forests grow wild doesn't get the job done anymore. Not since the end of the Carboniferous era. New forests don't make new coal deposits, that's a thing of the past.

Trees are still great though, for innumerable other reasons.

Left to their own, trees will grow, die, fall over, and build up soil. The soil is carbon sink. You just have to keep it from eroding.

As usual, everything that is infinite meets another barrier in physics. Initial investment is the barrier for every energy plant. But I also imagine ourselves already using energy like it’s cheap, trying to illuminate the moon with a big ad for Coca-Cola, “since we now have enough energy for it”.

I wonder how many capture plants and energy would be required to bring it back to a normal level.

Let's run the numbers. I believe Climeworks' Orca is the largest DAC plant right now, and they claim it'll capture 4000 tons/year. We currently emit something like 40 gigatons/yr. (4*10^10)/(4*10^3) = 10^7 DAC plants. So let's say that DAC isn't a silver bullet :-) There are a whole portfolio of other carbon capture and sequestration technologies that'll be needed to get us carbon-negative enough to avoid going over +2 degrees C.

Of course, by the time we have spare fusion power, that'll also have replaced much of those 40 gigatons, and Orca is certainly not the upper limit in terms of scale.

And we've emitted 2 teratons of excess CO2 that needs to be dealt with as well.

Renewables are already there.

In their "World Energy Outlook 2020" report the International Energy Agency reported that the world's best solar power schemes now offer the cheapest electricity in history, and solar is still getting cheaper at around 15% per year. Wind is not far behind, and large scale battery prices are dropping at around 20% per year. These price curves have been pretty consistently dropping for nearly 40 years.

With solar already the cheapest electricity ever known, just project one more halving, 5 years, and think what you could do with that.

Solar is acknowledged to be intermittent, but that's the sort of problem that power engineers love. The cheapest electricity ever, but only during daytime, with wind in strength available intermittently throughout the day.

How would you use that? Batteries obviously, to load shift. Move heavy industry and heating and cooling loads to the daytime peak. Tune aluminium smelters to run flat out during the day, and on low load with insulation overnight so the pots don't freeze up. Water desalination during daytime, storing the water produced. Link electricity grids east-west to prolong the hours of sun light for the overall grid. Produce hydrogen during daytime.

None of that requires any technology breakthroughs, they're just another day's work.

Do all that and coal power consumption will drop drastically, while using the world's cheapest electricity. Making these changes can just be background tasks, and will get us to 80% replacement of fossil fuels without any heartache. Recent modelling shows the sweet spot for renewables is actually around 400% replacement of the current grid, with that solving a lot of problems that were expected with the old thinking that we needed to replace coal with 100% renewables.

And then, if fusion comes along, great. We can close off the remaining 20%.

It's been all dreaming, all along.

Tokamak dreams are just the fever sort: extracting energy from a Tokamak fusor will necessarily cost overwhelmingly more than from fission. But fission is already not competitive, and gets less so every day. Fission could, in principle, deliver not unreasonably cost-ineffective power, but not in the US where its construction mostly delivers tax money via political patronage, with a sideline in grid power, maybe, someday. Small nukes might sidestep that system, but they have a hard time drawing capital in competition with the now reliable graft conduits.

Helion is interesting for its direct "aneutronic" electromagnetic energy extraction route, but dependent on a supply of Helium-3 that it will have to breed up from side deuterium-deuterium fusion. That one, anyway, might work.

But most of the dreaming right now is of the venture capital sort. Billions of dollars will change hands, for years, without a solitary watt-hour pushed out to the grid, while solar is built out as best it can absent those $billions. The more solar and, soon, storage is built out, the lower the price target the fusors and small-scale nukes must compete against.

It is not all gloom for them: they don't need to match the price per watt of peak output from solar, but something like three to six times that, because they only need to match the price of power delivered from a storage system that was charged up during the third of each day when solar is producing well. The odds-on favorite storage systems (favored first for cheap build-out) are not especially efficient, delivering round-trip maybe half of what was put in.

As intermittent solar picks up an increasing share of power generation, some industries will increasingly adapt to maximizing power use when it is cheapest. Those industries won't buy much of more-expensive power, but will instead wait for it to be cheap again. Prime examples of this will include hydrogen, hydrocarbon, and ammonia synthesis from water and air feedstocks, but even steel and aluminum production. Warring against this tendency will be debt service for capital equipment not making money sitting idle. Economic analysis will choose at what prices to start and stop production.

I read then link to the other article explaining the new technologies developed by Helion that should permit commercially viable fusion reactors, there are no words about Q plasma or Q total:

https://www.youtube.com/watch?v=LJ4W1g-6JiY&t=5s

so the affirmation that "Nuclear Fusion Is Close Enough to Start Dreaming" could be other speculation of the press not supported by facts. Obviously I hope to be wrong.

No, it really isn't. We are nowhere near achieving over-unity fusion when considering total system input vs. output.

Actually, nuclear fission is close enough to start dreaming. NuScale got their Small Nuclear Reactor approved by the NRC last year [1], and a few days ago they announced a deal to build a set of 6 SMR's in Romania, for a total of almost 0.5 GW [2], as soon as 2028. Chances are this could actually happen. Their other current contract is for a power plant in Idaho, which may open in 2029-30 [3].

[1] https://www.nuscalepower.com/technology/licensing

[2] https://newsroom.nuscalepower.com/press-releases/news-detail...

[3]https://www.nuscalepower.com/Projects/Carbon-Free-Power-Proj...

This sounds like 1950s claims of Fission being so cheap and clean that electricity would be almost free. Unanticipated externalities appeared in practice, resulting in high cost and pollution. Fusions heel may stray neutrons turnin reactor wall material highly radioactive.

It is more like fission externalities ended up way over-anticipated, almost killing the industry.

Fission is the technology that is statistically (even taking into account all the incidents) safer than most other viable generation options, yet people have very hard time accepting it

Well, at least this physicist has a completely different opinion on how far nuclear fusion is away:

https://www.youtube.com/watch?v=LJ4W1g-6JiY

That video is about the ITER that is designed using the tokamak model. https://en.wikipedia.org/wiki/Tokamak

The device proposed by Helion is very different. I think they are overoptimistic, but anyway the analysis of the video is not applicable.

It's cool that cheap money means that there's investment in stuff like fusion. I just hope when they inevitably go bankrupt when the money dries up the knowledge does not just get lost.

By cheap money you mean the rich have become so obscenely wealthy they can afford to spunk their money on moonshot ideas like interfering in elections or building rockets.

"Too cheap to meter". How does this compare with 1950s futurism? It at least has a nod to climate change and carbon capture and fuel synthesis.

"Too cheap to meter" == the investment does not pan out to the private sector. That's been the historical problem. No one wants to build a big fancy thing and thing run it at cost.

The reason fission is going for prefab is not only volume lowering costs, but selling gizmos for a profit that the buyer runs at cost does make sense.

(There are some interesting vaguely Marxian dynamics with high tech with minicsule operating costs that should always be considered. That we all drive cars and not ride trains is similar. The car industry surely involves 3+ orders of magnitude more people per mass * distance transported)

Wishful thinking with no evidence to back up this click bait.

Nuclear Fusion has been 20 years away for 70 years.

Just like electric cars...

Oh, wow, now we're actually driving electric cars!

We had electric cars 130 years ago.

Exactly. Just like we have fusion for decades.

As fusion isn't practical, electric cars were not practical mass mean of transportation until very recently.

Electric cars have always been doable, they've just been more expensive than fossile fuel alternative which is ostensibly free energy. The situations are not at all comparable. Nuclear Fusion has still barely been demonstrated as a proof-of-concept.

I don't believe that is true about electric cars. Battery technology is now only good enough for a half decent range.

Nuclear fusion has been proven for decades, it's the goal of producing net-energy that is the problem.

It really makes you (not) think.

I was hopeful for a bit, then I saw some stuff like this: https://www.youtube.com/watch?v=JurplDfPi3U and realized nothing has changed. It's the same old shit: reporters being willfully ignorant, business people misleading the public, and on and on. Wake me up when you have plans for a full-scale, fully-functional fusion plant that can actually be built in an economically feasible way. Until then, I don't care.

The headline says "close enough to start dreaming"

I don't get it. If you're not into following the tech development, only interested in the final stage, why would you even open this and bother commenting?

Reply on news.ycombinator.com