Fusion power is getting closer
economist.com
economist.com
"All" we need is political action to build more clean power plants (nuclear, solar …) and continuation of research to electrify more sectors.
Fusion will not solve any of those important issues. It will only be the cherry on the cake if it become deployed globally anytime in the future. But we are far from it and we shouldn’t be that excited or hope for it to make a better world.
I very much see fusion as going to the moon : it’s not going to solve any short time issue but it can be a milestone in human advancement if we manage to survive until it becomes our primary source of energy.
We’re building solar about as quickly as we can, and without cannibalising EV battery supply chains are also maxed out on batteries. Fission power remains uncompetitive with natural gas; you’d have to convince Western consumers to bear higher prices for reduced emissions which nobody is willing to do.
Which is why I mentioned consumers. Voters have made this incredibly clear every time polled: we are not wiling to pay more to reduce emissions.
That's putting it mildly.
This approach to "polling" also appears to ignore the fact that statistically speaking, almost half "the population" made the opposite choice. This makes sweeping statements about "what people want" based on this sort of thing pretty suspect.
Of course, the claim could be true, nevertheless. This sort of "evidence" isn't really very salient, however.
2018 in America, "57 percent of Americans are willing to pay a $1 monthly fee; 23 percent are willing to pay a monthly fee of $40" [1]. 2018 in France, "carbon taxes on petrol and diesel" prompt les gilets jaunes [2]. 2024 in America, "more than half of Americans are unwilling to pay any amount of money to combat climate change" [3].
And then there is the global success of climate-skeptic or outright denialist citing the cost of Paris Agreement compliance.
[1] https://apnorc.org/projects/is-the-public-willing-to-pay-to-...
[2] https://edition.cnn.com/2018/12/31/opinions/macron-warning-c...
[3] https://epic.uchicago.edu/insights/2024-poll-americans-views...
I do recall one of the catchphrases that surfaced around les gilets jaunes: "these people are going on about the end of the world while we're just trying to make it to the end of the month".
It's too bad that doesn't prompt deeper reflection about and anger at just who benefits from such a bifurcation of interest.
By noting it’s irrelevant. Solar production isn’t capital constrained. Certainly not in the West.
I was mainly reacting to this. Can nuclear really not compete with gas even if the government subsidized both equally? (with all the caveats that come with comparing subsidies)
What I do know is fission has a horrible cost profile. First there’s a huge build cost, which can often skyrocket unpredictably. Then you’ve got a hugely profitable part where you’re basically printing money. Then there’s an enormous cost at the end.
Huge cleanup costs, whether it’s nuclear or coal, basically means either government has its eye on the ball to an excellent extent (Australia seems to be good here.) or it’s not going to be fixed except by the same government that didn’t tie down the guys who made the money enough.
I’d call fission a bad choice for countries with weak or chaotic government. You need a rock solid civil service to allow it.
(You’ll observe the same is true of almost every mineral extraction project.)
https://www.eesi.org/papers/view/fact-sheet-fossil-fuel-subs...
Here in France, most of our power is from nuclear. This gets us really cheap power production, compared to our neighbors in Europe.
As much as I am a proponent of green energy, this claim is outlandish. If we had started building more nuclear reactors 10 years ago, sure, but by themselves wind and solar aren’t quite sufficient enough for places like the US to fully replace fossil fuels. There are limitations for both, namely the need for space and acceptable weather conditions. Solar panel energy yield still has potential to unlock, as well. Last update I’d seen about panels said they still only hover around 25% energy yield.
All other problem will be solvable, I we archive this goal
As such, some skepticism in people's timelines is always warranted. I'm reminded of the voodoo fusion article. https://vixra.org/pdf/1812.0382v1.pdf
I'm pretty sure he was talking about sustainable electric grid usable power. He spent some time with a key in an underground silo though...
Fusion will be important for the future of humanity, especially in deep space, so we should keep researching and developing it. But it's not relevant to the climate crisis... the already existing technologies have that covered, we only need to deploy them faster.
Going all in on solar and batteries either requires nuclear energy, thereby raising power costs, or natural gas, which means we can’t make our 2050 goals. We simply can’t manufacture the panels and batteries fast enough, the latter especially if we also want to decarbonise transportation.
Fusion won’t come fast enough for the climate crisis. But it may be relevant for complete decarbonisation, e.g. replacing the natural gas plants we’re spinning up as quickly as we’re deploying solar.
…why? If you want massive lasers you use a nuke to power it [1][2].
> If you think "practical energy source" is what fusion research is really about
Outside the NIF, which is openly about stockpile stewardship, yeah. That’s what it’s about. Tokomaks aren’t militarily useful.
The biggest downside to solar is the amount of surface area required and cost per panel. Not to mention the amount of rare earth required to make just one.
I’d be interested in much power fusion will output in terms of square footage and cost to compare them directly.
Regarding the amount of rare earth required to make just one solar panel, i was surprised to find out that effectively all you need is something called 'silicon' as well as an initial investment in energy to bake the thing.
It just so happens to be that, according to google, this 'silicon' stuff is quite abundant and can for example be found in desert sand.
Since you can stack those fusion reactors vertically, i suppose they'd come out on top in the power/sqft category.
As underlined in the article, ITER is already 9 years behind schedule. Not because they're incompetent or underfunded, but because fusion is hard. I don't expect a private company with much less funding to do any better.
That said, I'd like to be wrong and see one of these companies crack commercial fusion in their basement.
The argument from SPARC wasn't that iter was incompetent but that being so vast and expensive requires multiple nation states cooperating which is inherently slow. Slow things being more risks as thing a change too (9 years is probably 3 governments here). Making it smaller makes it cheaper, a d cheaper means a smaller collaboration (or single company).
Now, don’t get me wrong, I’ll be ecstatic when / if it arrives, especially if it does so with sufficient lead time to be relevant within my lifetime, but no number of “we’ve halved the distance to the wall again” articles will get me excited (or us there).
I think it's better to think of it this way: are you a lay person not in the field? Then most of this work doesn't matter to you. But you should keep funding it.
And yes, absolutely, fund baby fund, keep making those Zeno steps… just don’t write another article promising me that fusion / Jetson cars / AGI / Unobtanium is nearly here.
Did we read the same article? Helion's entire design strategy is to sidestep sustaining fusion plasmas. It's unclear if it will work. But it's theoretically sound. Your wall is a chain-link fence.
Clean enough energy alternatives are being deployed everywhere.
And we might be already running out of time, the climate system could had surpassed a point of no return by now, we keep adding fossil carbon to the system and if, or by when, this became practical enough to start deploying power plants based on it will be too late for a meaningful change.
The actual problem is reaction vessel wall ablation.
c.f. coal. The bulk of the cost of a coal power plant is the steam turbine etc. And coal is not economical.
In the way a fire is a steam engine.
Its interesting to think about how our economies will change if the price of energy drops by one or two orders of magnitude.
Desalination could then be used for agriculture and industry. Coastlines could become more valuable.
We could make huge amounts of aluminum for almost nothing. We'd probably end up with aluminum packaging for everything as it's much easier to recycle than plastic if you don't have to worry about electricity.
(not trying to make any kind of point, just a fun fact / video)
You won't. Sorry for being cynical, but the true cost of energy is the capital invested in the equipment to harvest it. A fusion reactor would be at the minimum, as expensive as a fission one, which means all that capital will have to be amortised over the years. This doesn't even take into account the cost of extracting deuterium from sea water or producing helium 3 and tritium from lithium. Fusion doesn't even solve the problem of nuclear waste, since it will create radioactive isotopes by bombarding components of the reactor with neutrons.
You want cheaper, cleaner energy ? Solar and wind with battery storage is the best option, and it's already here, no need need to wait for fusion magic that may or may not happen in our lifetimes.
What are you basing that on? Many of the fusion designs being talked about are quite small.
Nuclear is mostly capital costs, not the fuel. Fusion doesn't look any cheaper to build. Like any other bit of complicated infrastructure, it'll have a limited lifetime.
So, power cost is ((capital_costs + running_costs) / amount_produced) * profit_margin
While I'm very much in favor of research, I don't understand the expectation of that it'll make power cheap. I'm pretty sure nothing will beat solar and wind because they're cheap, simple and very amenable to mass production in ways that fusion and fission can never be.
At this point IMO neither fusion nor fission have a future as a main source of grid power. They're still worth researching, and will still have specialized uses in places like submarines and aircraft carriers.
But let's suppose their stuff works as well as they imagine. IMO, not even then. Just look at their youtube videos. All sorts of big, complicated, and expensive looking magic. Huge capacitors, vacuum chambers, extremely high temperatures, etc.
Will that scale better than nuclear? Maybe. Will it scale better than a tech that can be pumped out by the million, shipped to normal people and put on any random roof? Highly doubt it.
Of course keeping it indefinitely on will require huge maintenance costs but if the infrastructure is modular enough to be able to replace every elements, that may be doable.
With fission the structure of the building is part of the infrastructure’s security so it’s harder to change when it’s too damaged. Maybe I’m wrong but it seems like fusion doesn’t require huge cement walls that would take a lot of constraint and which would effectively limit the maximum lifetime of the building.
This is likely going to happen anyway because of the rapidly falling costs of solar power. Although it certainly doesn't hurt to make energy cheaper yet again.
Are there other signals that you’d point to as influencing your shift to optimism?
Why would that be possible? Because they don't go on yet another fancy way to boil water to run a steam engine. They do direct energy conversion.
In a fusion reaction energy is mostly released as kinetic energy of particles. Their fusion reaction is aneutronic, ie with few neutrons. That means that the energy is released as movement of charged particles. By definition this is an electric current.
This current is directly captured by induction in the coils and sent to grid.
If their 2025 experiments go well, in 2028 they will start selling electricity to Microsoft which whom they have a power purchase agreement.
Thanks to the simplicity of their machine, they aim at a cost of 10$/MWh, competitive with solar. Their 50MW generator is shipping container size and be built in a factory by the dozens.
Ok, they may fail, they could also succeed though. Worth watching.
Charged plate conversion system require the plate spacing be lower than the Debye radius of the plasma. So it's out for most tech (especially pulsed fusion)
And a lead-lithium blancket (the only mature tech to capture neutrons rn) is unwieldly, only capture neutrons (so already between 20 and 25% losses, let's say 20), and will loose more energy via reaction (creating deuterium-tritium, so not useless, but still, lost energy, 25% losses). Then between the heat exchangers and the steam engine, let's say 65% losses (and i think i'm generous). So you'll get at best 21% of the generated energy as electricity (and i'm being _Very_ generous. imho if you have more than 15% it's a success).
With that, you'll have to contain and breed the plasma. Good luck.
Far cheaper and more practical.
No reason for pointless cynicism, this is also against the guidelines. Fusion power is like childbirth - yelling at 9 months that it has been 9 months, and is therefore never arriving, would be delusional.
This is low effort. Are you arguing we haven’t made material progress in the last twenty years?
Material science advances in magnets alone should raise your curiosity to not have to wait for a packaged consumer product to notice advancement.
Arguably there has been a time when fusion got further away (compared to predictions) because of reduced interest and funding. But I'd agree that for the last two decades you could print the headline "fusion is getting closer" at any time, and it would always be true.
Maybe that's just me ranting about how titles have shifted from trying to inform you and distill the essence of a news article to just trying to get you to read the article. It's really a comment on journalism rather than fusion.
This is a terrible assumption when it comes to basic research.
> that for the last two decades you could print the headline "fusion is getting closer" at any time, and it would always be true
What are you basing this on?There have been a few discrete steps forward. In between, ideas and experiments.
> really a comment on journalism rather than fusion
Can you quote the line from the article that renders the title misleading?
Even trying out things that don't work is progress, since it narrows the space of possible solutions. The only ways to really set back progress would be by getting others to believe something that's untrue, for example by faking research.
You could manage to not advance the field by not publishing a finding and leaving the field. You can slow progress for example by popularizing a path that doesn't lead anywhere, discouraging people from going down the right path, or just reducing popularity or funding of the field. But slowing progress doesn't erase progress. If you do genuine work and at least talk about it with colleagues on a conference that's advancing the field.
> What are you basing this on?
The classic graph everyone uses would be 1, which shows consistent progress over ~40 years. It's a bit out of date now, but I assume the trend has held since then.
> Can you quote the line from the article that renders the title misleading?
I didn't call it misleading, I called it lazy and uninformative. A (mild) form of clickbait, because it merely hints at information without providing any. My English teacher would have deducted points for that title.
1: https://www.researchgate.net/figure/Progress-to-Date-in-Achi...
Maybe, maybe not. You may be searching an infinite domain, or the wrong domain entirely.
> If you do genuine work and at least talk about it with colleagues on a conference that's advancing the field
Sure. But it’s not advancing a technology.
> classic graph everyone uses would be 1, which shows consistent progress over ~40 years
Fair enough. I’d argue that’s constant progress in one of the problems of fusion power, namely, ignition. The article’s point is the other problems are either being solved or showing similar progress. That wasn’t true until recently.
> it merely hints at information without providing any
It's debunking the assumption at the top of this thread. Especially with the "no really," it's arguing that the ground truth has changed.
Close enough to get the energy, but far enough away to allow isolation of the reaction.
Maybe the energy could just beam down on us from space?
That'd be awesome...
Now, wouldn't be awesome if there was a global power grid that could provide power anywhere on the planet? (still cheaper than fusion)
No need to cover all of Michigan, the area available on flat roofed buildings everywhere is enough.
Of course, a distributed power source, locally generated and consumed, is anathema to the capital motive of concentrating the most wealth in the fewest hands...
One, neutron. Two, doesn't glow. Three, aneutronic fusion is real.
Even if good net-energy happens, even if they solve neutron flux irradiating the reactor vessel, even if they can sustain power generation for commercial periods, if they can safely operate the superhot plasma
... it comes down to, ECONOMIST, the cost. As someone that hammers fission power for being currently infeasibly expensive, fusion has a long road to economic viability. I suppose the public will at least not NIMBY it to the same degree as a fission reactor
If you want anybody to be able to make calls for free, that’s one thing and i think it could be feasible (probably already feasible by now).
If on the other hand you want everyone to be able to make 4k60fps group video calls that’s unlikely, at least for now.
As long as the communicator is free (I don't have to pay to have it or replace it for the newer version) and private (I can be sure no one will listen or record).
If we slow it down, I think humanity can manage without 4k, with a 50% cut in digital entertainment, and without buying new expensive useless communicators each couple of years.
Could sure shake up some of the US’ geopolitical relationships, in which the US spends an awful lot to maintain influence over the way oil gets made and sold. That pot of oil money sure creates its share of headaches, and probably means a lot more to smaller nations than it does to the US.
Building fusion plants would be a massive money maker initially as everyone races to adopt it, but is it a long term money maker when no doubt other nations will be trying to cash in on it themselves.
It absolutely does, just in the opposite direction from OP’s hypothesis. Washington likes cheap oil. It wants high production levels from its producers to keep prices low; producers want high and stable prices.