Microsoft Bets That Fusion Power Is Closer Than Many Think (wsj.com) 94
Many experts believe fusion power remains decades away. Microsoft thinks it could be just around the corner. From a report: In a deal that is believed to be the first commercial agreement for fusion power, the tech giant has agreed to purchase electricity from startup Helion Energy within about five years. Helion, which is backed by OpenAI founder Sam Altman, committed to start producing electricity through fusion by 2028 and target power generation for Microsoft of at least 50 megawatts after a year or pay financial penalties. The commitment is a bold one given that neither Helion nor anyone else in the world has yet produced electricity from fusion. "We wouldn't enter into this agreement if we were not optimistic that engineering advances are gaining momentum," said Microsoft President Brad Smith.
Fusion powers the sun and stars, and has the potential to provide nearly limitless amounts of carbon-free power if someone can harness it on earth. The International Atomic Energy Agency expects electricity from fusion to be produced in the second half of the century. Helion is building a prototype that it says will demonstrate the ability to produce electricity through fusion next year. "The goal is not to make the world's coolest technology demo," Mr. Altman said in an interview. "The goal is to power the world and to do it extremely cheaply." Mr. Altman, the chief executive officer of OpenAI -- the artificial-intelligence startup behind the viral chatbot ChatGPT -- said having a first customer is critical for keeping Helion grounded in the realities of business, including working with clients, utilities and electric-grid operators.
Fusion powers the sun and stars, and has the potential to provide nearly limitless amounts of carbon-free power if someone can harness it on earth. The International Atomic Energy Agency expects electricity from fusion to be produced in the second half of the century. Helion is building a prototype that it says will demonstrate the ability to produce electricity through fusion next year. "The goal is not to make the world's coolest technology demo," Mr. Altman said in an interview. "The goal is to power the world and to do it extremely cheaply." Mr. Altman, the chief executive officer of OpenAI -- the artificial-intelligence startup behind the viral chatbot ChatGPT -- said having a first customer is critical for keeping Helion grounded in the realities of business, including working with clients, utilities and electric-grid operators.
Is Microsoft betting for or against? (Score:4, Interesting)
Since Microsoft stands to receive a penalty from Helion if they fail, we cannot judge without seeing the numbers whether Microsoft is betting for, or against, Helion succeeding. It might be a more or less neutral (profitable either way) bet on Microsoft's part, whose purpose is to give Helion some upfront capital and associate their brand with a big famous one.
Re:Is Microsoft betting for or against? (Score:5, Interesting)
Since Microsoft stands to receive a penalty from Helion if they fail, we cannot judge without seeing the numbers whether Microsoft is betting for, or against, Helion succeeding. It might be a more or less neutral (profitable either way) bet on Microsoft's part,
Only if Helion has enough money left to pay the penalty after failing.
I'm a bit skeptical about the odds of that.
Re: (Score:2)
Re: (Score:2)
Things that promise magic technology often are.
Re: (Score:2)
Re: (Score:2)
Since Microsoft stands to receive a penalty from Helion if they fail, we cannot judge without seeing the numbers ...
Based on TFA, Microsoft either buys electricity at market rates, or gets paid a penalty.
Unless there's some upfront money TFA didn't mention, the only way Microsoft can lose is if the market goes crazy high AND commercially viable fusion power actually exists.
Re: (Score:2)
no prototype (Score:2, Interesting)
Re: (Score:3)
in USA fusion plants will be under much less stringent rules than fission plants. They're more like a particle accelerator than big bad nuclear device with waste needing centuries of isolation. "Approval queue" not a big deal compared to the atom splitters.
https://www.cnbc.com/2023/04/2... [cnbc.com]
Re: (Score:2)
Given that no prototype exists ...
Maybe they've got a plant with a *bunch* of these [wikipedia.org] strapped down running generators?
Re: (Score:2)
"I AM AN IDIOT"
You scream into the void while not even bothering to Google a statement.
Re: (Score:2)
Re: (Score:1)
Of course a prototype exist.
Dumbass.
nor is anything in the approval queue for nuclear regulators
a) you don't know that
b) you can't possible know that
c) a fusion plant does not fall under any "nuclear regulations"
Dumbass.
Re: (Score:2)
Of course a prototype exist.
No prototype of fusion reactor producing usable power exists as of today.
We are still barely scratching the surface, with the longest sustained fusion reaction being at 17 minutes [smithsonianmag.com]. Then we will need to find ways to harness that energy, and create electricity from it
The Tokomak design plans to find ways to generate steam from the heat [iter.org], but there are engineering "details" to solve in order to be able to do it. Those are still being researched, and there are no prototypes yet for that part of the process.
Re: (Score:2)
If it's not a prototype that outputs more energy than it takes to run the prototype then it's effectively not a prototype and is not particularly any better than some quacks perpetual motion device.
Fusion has still got a very very long way to go, current fusion machines do not appear to be built to output energy, just just create it but don't harness it.
the problem with Helion's method (Score:5, Interesting)
Helion claims it will run a a mostly aneutronic He3-D reaction, with the He3 produced from a side D-D reaction. The rest of the world is struggling with D-T fusion at a hundred million C, and Helion is claiming they're going to have a side show of a reaction that requires 450 million degree C temperature! Oh, and that lovely He3-D with (mostly) no neutrons? that's 600 million degree C
Yeah, hard to see how they're not just full of shit and investor money sinkhole. Be great if it worked eh?
Re:the problem with Helion's method (Score:5, Interesting)
Their machine is pulsed and uses a magnetic pinch which makes high temperatures not particularly unusual. Sandia's z pinch machine gets up to a few billion kelvin and has performed D-D fusion.
The challenges are more along the lines of being able to control the plasma to get enough compression for long enough to create net energy, and to repeat shots quickly enough to produce a reasonable amount of it.
There have been major machine learning based breakthroughs in fluid dynamics recently, and there's a bunch of new interest in pulsed fusion with various types of magnetic pinches. It's not impossible Helion actually has something. They're apparently willing to commit to a hard timeline.
Also, your reactions are off. The D-He3 is most reactive (more than D-D) between a few hundred million to about 10 billion K. At 10 billion K, D-D, D-He3 and D-T are about the same.
Re: (Score:2)
I haven't looked into them that much. Are their plasmas Maxwellian or not? Because the temperature / density requirements non-Maxwellian plasmas can be a LOT lower than for Maxwellian plasmas.
Re: (Score:3)
I'm not sure Helion shares information to that level, but their machine generates two blobs of plasma and then accelerates them at each other, so you wouldn't expect it to be entirely Maxwellian. It would depend on what proportion of the energy is thermal versus from the velocity of the plasma packet.
IIRC there are some details about how Helion has described the plasma acceleration that have raised some skepticism.
Their website is kind of cool. You scroll through the animation: https://www.helionenergy.com/ [helionenergy.com]
Re: (Score:2)
The colliding torus idea is nice because it gives you hot ions, and cold electrons (since they are moving at the same velocity) but I believe, and some studies have shown that the plasma thermalizes quickly enough that the advantage doesn't last.
Re: (Score:2)
The difference, if there is one, is that now we have fast fluid dynamics solvers. Magnet technology has also advanced quite a bit.
I don't think I'd buy in with the little I know about what they're doing, but they also haven't asked me.
Re: (Score:2)
> You might note that that is almost exactly what TAE (Tri Alpha Energy) tried to do for a decade or two
And still claim so today while building ever-larger machines that also fail to post interesting results.
However, TAE does not compress their FRC, which is the basis of the Helion design. TAE injects fuel radially into the FRC in an attempt to produce a colliding-beam-like process so they can operate on the pB11 peak.
Unfortunately, some very basic math has repeatedly demonstrated (see IEEE, Nature, othe
Re: (Score:2)
I haven't seen any details on Helion but the colliding plasmas look a lot like the early TAE systems before they found out (as calculations indicated) that they thermalized very quickly so they switched to neutral beam injection.
Fusion is hard. Far harder than most people think. It
Re: (Score:2)
> Trying for p-B when they haven't gotten D-T working is just silly - or worse
Indeed. One *might* accept the argument that introducing T involves all sorts of regulatory issues and such, but then one can run in on D-D which avoids all of those and is still orders of magnitude "better" than p-B in demonstrating the physics works.
The fact that so many of these companies propose using fuels that require many more years of development seems... well, let's just say "fishy".
>... so they switched to neutral
Re: (Score:2)
Aside from the plasma physics, I find it very attractive that Helion's approach directly extracts electrical power from the expanding fusion plasma, rather than collecting heat to spin a turbine (and the accompanying thermodynamic limitations thereof).
Re: (Score:2)
> Helion's approach directly extracts electrical power from the expanding fusion plasma
Well, they **claim** they will do so, but this has not been demonstrated. They *have* claimed to extract 90% of the original energy in their last device, but there was no fusion involved, they simply injected plasma and let it relax. No actual details have been released that I know of, just the press release (there was a poster presented, but I have not seen it and don't know if it covered this bit).
I do need to remind
Re: (Score:2)
There's not much fiddling around with the way they're harvesting the energy. Hot plasma pushes on its confining magnetic field. Hotter plasma pushes harder. The thermodynamics is pretty unequivocal on that point.
Re: (Score:2)
> Hotter plasma pushes harder
You are assuming the energy is trapped in the plasma. That is a bad assumption. From the very earliest days of the fusion effort we have seen that energy escapes from plasmas in a rather wide variety of methods that do not result in the energy of the bulk being increased.
For instance, let's say that the FRC becomes unstable during compression - which is a good bet given it's been the case in literally 100% of previous experiments and there's nothing different in their design
Re: (Score:2)
Hotter plasma pushes harder. As I said, the laws of thermodynamics are pretty unequivocal on that point.
Your ability to confine the plasma may not be adequate to harvest that energy, but it doesn't mean the hot plasma isn't pushing harder.
You're trying to argue something by analogy to an unrelated process. That's not a good argument. Your point about difficulties confining pinched plasma is a much better one, and if you look waaaay back to the beginning of the thread you'll see it's one I mentioned. It's al
Re: (Score:2)
Using the compression magnets to also harvest the energy is a neat trick all right.
Re: (Score:2)
You're hilarious. The z-pinch machine did that 2 billion degrees for 10 nanoseconds, which was enough (17 years ago) to do nothing much. A few neutrons isn't reason to party. Z pinching makes unstable plasmas, and that's why there are no real breakthroughs with them.
Helion has demonstrated nothing of substance with producing either the 500 million C for the D-D nor the 600 million C for the He3-D, which is what will be needed. A mere smattering of those reactions would show they're failing for the same
Re: (Score:2)
Idiot
What is funny is an idiot like you calling someone else an idiot.
Re: (Score:1)
That is easy.
I have an high accurate idiot detector.
Someone who claims other people liars without even havinb the slightest clue what they are doing: is an idiot. In some countries that is even a civil offence.
Re: (Score:2, Insightful)
I have an high accurate idiot detector.
Someone who claims other people liars without even havinb the slightest clue what they are doing: is an idiot. In some countries that is even a civil offence.
Does your high accurate idiot detector keeps beeping all the time when you are around?
You might be on to something there...
Re: (Score:2)
I have an high accurate idiot detector.
I remember awhile back we started modding you down when you did these ad hominem attacks against people that disagreed with you. Do we need to start doing that again?
Re: (Score:2)
I was replying to a person who invoked Z-pinch as some kind of silly proof as to why Helion's claims could have some merit.
Why don't you pay attention the the flow of conversation in a thread before going off half-cocked?
Re: (Score:2)
You should read more carefully.
And don't be such a dick.
Re: (Score:2)
The Helion approach is not based on compressing plasma.
They basically have a very simple acceleration machine which shoots ions from two sides into the reaction area, captures everything escaping to the sides and brings that back to the starting position.
It's not impossible Helion actually has something
They have something. They have continuous fusion as long as they want.
They are working on scaling it up and extracting electricity.
Re: (Score:2)
From Helion's web page:
Re: (Score:2)
The other thing is, if you measure the power output of the sun in Watts per Cubic Metre, it is actually very tiny, way smaller than the energy output of a human body at rest. It only produces usable amounts of energy because it is so huge.
Re: (Score:2)
This is another talking point that nobody really seems to think much about. First, what are you using for the volume of the sun?
If you do a Google search for the sun's volume you get wildly varying answers because the sun doesn't have a surface. Going with a power output of 3.8e26 W and volume of 1.4e27 m^3 you get 0.27 W / m^3, which is indeed a lot less than the human body's 100 W / 6.2e-2 m^3 = 1.6 kW / m^3. Incidentally, the sun's density is 1.41 g/cm^3 (again, wildly varying depending on what you consi
Re: (Score:2)
And why do you not just "google Helion"?
Or go on youtube and watch their science videos?
Or why exactly do you want to make an idiot out of yourself?
If you say "claim" you imply they are lying. Well: if you had 3 brain cells, you would know they are not lying.
Hm, I used "dumbass" already 2 times today. So you are an mere idiot.
Re: (Score:1)
Re: (Score:2)
What they say in the video is that they can produce the H3 in a tokamak-like lab which gets bombarded with neutron radiation and has to be replaced every five years, harvesting the power via steam turbines or not. But the H3 can be stored for 12 years and then direct drive coils to harvest much more power in a reaction without neutron radiation. Whether this works out cheaper, they don't know.
a customer? (Score:1, Insightful)
How can you need "a customer" for fusion power ?
if it works you'll have more customers than you can possibly handle.
also, you don't need a customer to keep you grounded, you need a CEO who's not an idiot.
the end goal of "power generation, and cheaply" is the motivation in and of itself.
doesn't even need to be "cheap" just competitive.
this all sounds very weird. why is that i think there's a bunch of c-level types hiding in the corners who are making money on this deal in some barely legal way.
Re: (Score:2)
> How can you need "a customer" for fusion power ?
Launch customer.
This is how most new power sources get built. Consider the NuScale build in Utah or the BWRX-300 in Ontario. Launch customers.
Re: (Score:2)
It may be a gimmick but Helion is generating massive amounts of power already, more than even they expected -- not net gains but still. The next prototype is dealing with that extra power whilst trying to harvest it. It will likely produce massive spikes of electricity -- what anyone can do with that, I don't know.
Board Members = NOPE (Score:5, Insightful)
Helion's board is two investors and Sam Altman, who we know we can trust because he runs that open(*), non-profit(*) company OpenAI.
Nope.
https://www.crunchbase.com/organization/helion-energy/people
MS is smart (Score:3)
Both this and natrium are just up the road from me. Ppl are talking about these. Both are coming along nicely.
And both will scale via SMR approach.
For those that say that we must stop fission, nothing could be further from the truth. We have 'spent fuel' that was ran through thermal reactors. They used a whopping 2-3% of the available energy. However, if we run this 'spent fuel' through fast reactors (like Natrium, arc, etc), then we can finish the nuclear fuel cycle (i.e. burn it up), and leave us true spent fuel.
However, we NEED both.
Re: (Score:2)
Don't we largely make sure that reactors only extract 2-3% of its power because its the higher percentage of usage that leads to weapons-grade "Highly Enriched Uranium"? Can fast reactors use it up without making it also weapons grade?
Re:MS is smart (Score:4, Informative)
Enrichment of uranium is something you generally do on the regular uranium ore. Spent fuel processing usually involves separating the different elements in the fuel. One of those is plutonium, and it can be used to make bombs.
More modern fast neutron reactors aren't as picky about their fuel, so they can burn less processed "spent" fuel. The ARC the GP mentioned is one of these, although it can also start with fresh fuel without requring it to be removed and reprocessed at all.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: MS is smart (Score:3)
Never buy a fusion reactor that uses shingled magnetic recording!
Need is Temprary (Score:2)
For those that say that we must stop fission, nothing could be further from the truth.
The only reason we need fission reactors now is because we don't yet have fusion and fission is the current best source of "on-demand" power generation that is carbon-free. When we eventually get fusion there will be no need for fission since literally its only advantage over fusion is that it's achievable with current technology.
Re: (Score:2)
Re: (Score:2)
As to building in undeveloped nations, probably better to use NuScale, ap300, etc.
Re: (Score:2)
> and fission is the current best source of "on-demand" power generation that is carbon-free
Ring ring... hydro calling.
And it needs to be noted that most fission reactors are actually quite craptastic for on-demand. That's because, depending on the design, have strict limits on the throttling based on neutron economy and the production of short-lived neutron poisons. While it is possible to ramp them down fairly rapidly, ramping them back up generally takes hours. Normally it's expressed as a percentage
Re: Need is Temprary (Score:2)
Re: Need is Temprary (Score:2)
stepping in it (Score:2)
Microsoft has a history of stepping in shit. This is the company that produced Windows ME, Clippy, and Internet Explorer among other wonders of the world.
Re: (Score:1)
No no no, flying cars come first, then YLD, then commercial fusion. If you do it out of order, you'll break my DeLorean's Flux Capacitor, and the timeline.
Helion will buy solar power, resell to MS and call (Score:5, Interesting)
Re: (Score:2)
and call it fusion power, which would be the AI thing to do. Depending on how the contract is drawn up, that is totally plausible.
Also, they could use fusion power to heat up bits of the earth's atmosphere, which will generate air currents which could be used to generate electricity.
Or use fusion power to evaporate water, and then generate electricity as it is caught on high ground as it comes back to sea level.
Or they could use fusion power to power growth of plants, and burn that to generate electricity.
Obviously the plants could have been grown by fusion power millions of years ago.
They appear to be promising not to use f
All power is indirectly fusion (Score:2)
My house is already fusion powered... (Score:2)
Using 26kw of solar panels.
It's only 20 years away (Score:2)
As it has been for over half a century now.
Re: It's only 20 years away (Score:2)
Re: (Score:2)
Well, a good MS product is not (Score:2)
Closer than I think it is, that is. Not going to happen, ever.
That said, from a recent podcast with the Wendelstein X7 people, steady progress is being made, but there is a ton of things still to do and 50...100 years is realistic.
Re: (Score:2)
> That said, from a recent podcast with the Wendelstein X7
Stellarators are an order of magnitude behind tokamaks in terms of performance:
https://pubs.aip.org/aip/pop/article/29/6/062103/2847827/Progress-toward-fusion-energy-breakeven-and-gain
W7-X's best performance from 2017 is 1/10th that of JET in 1997.
They have some advantages, but also some *really serious* disadvantages in terms of practical designs. It seems unlikely any of the current concepts can be made into a practical device, they are simply t
Re: (Score:2)
What "performance"? There is no performance in any of these systems. They are all doing materials research and plasma research at this time. That is why 5 or 10 years to practical energy generation is complete and utter nonsense.
Re: (Score:2)
> What "performance"? There is no performance in any of these systems.
Of course there is, it's call the triple product.
I linked to an article explaining the concept and detailing up-to-date real-world results.
Duh.
Fusion scams right around the corner (Score:2)
Fusion power remains right around the 50 year corner but the scams are already here.
Re: (Score:2)
Most of the companies are quite open about their progress and plans. Helion is not one of them.
Re: (Score:2)
You are an idiot.
Helion is open.
Everything they do is published and on youtube.
Re: (Score:2)
No scientific articles, fancy youtube CGI doesn't count.
Re: (Score:2)
Then watch the movies made in their reactor ... no idea about CGI.
I never saw anything made by CGI from Helion.
Re: (Score:2)
Helion has published an article in a scientific journal:
https://www.researchgate.net/p... [researchgate.net]
Hmmm... (Score:2)
The biggest baddest most powerful fusion reactor in the world can't do what these guys claim they'll do in a few years.... and are probably several decades away MINIMUM.
I smell bullshit.
Easy win for him (Score:2)
The clue is in the summary. Set up a 50 MW solar farm and call it won.
Average overrun (Score:2)
does the... (Score:2)
"...or pay penalties if they can't" clause mean anything of the company has vanished into receivership and the board is insulated by LLC provisions?
Not even close (Score:2)
Warning: long post.
Helion's system "works" by merging two FRCs created by theta pinch and accelerating them together into a reaction chamber where they mix to product a larger FRC. This process also provides much of the initial heating of the plasma. The new FRC is then compressed with magnets, thereby increasing both its density and its temperature (ideal gas law) to fusion relevant conditions. In order for all of this to work, I think they need to demonstrate the following:
1) successful merging of FRCs in
Re: (Score:2)
"I cannot find any published results that demonstrate their confinement time scales according to their predictions. In fact, they have simply stopped publishing results entirely, and now only put out press releases announcing some number or another with none of the others needed to understand its meaning."
So Helion are being deliberately disingenuous with their published results.
Not a good look.
Avoid.
Closer ie less than 20 years from now (Score:1)
Not surprised (Score:1)