Microsoft's big bet on helium-3 fusion explained Microsoft is making a big bet to cut its carbon emissions with a staple of science fiction: helium-3 fusion power plants. Of course, that's assuming nuclear energy startup Helion can stick the landing and that's a rather big if. Despite decades of research and development and billions in investment, fusion power has thus far …
That's why the smart money is on the Wendelstein 7-X - that's a proper Bond villain gadget name
After watching Real Engineering's YouTube video on Helion, I was convinced they had a pretty good chance of pulling it off. Then I watched Improbable Matter's response to the Real Engineering video, and I was convinced that it's not far off of a scam, with no realistic chance of succeeding. The trouble is, I'm nowhere close to being able to evaluate the physics myself, so I depend on those smarter than wiser than myself to judge. But as we saw with the likes of Theranos or Magic Leap, big companies with deep pockets that should be able to afford the very best advice can still end up pouring money into companies that promise things they just can't deliver. So IDK, I certainly wouldn't invest my own money in Helion, but I hope Microsoft has picked a winner here, we could certainly use commercial fusion power sooner rather than later.
It would be a challenge to get a power station using proven well established technology up and running in 5 years, and impossible to do for a nuclear fission plant.
For an experimental technology like this you might get another prototype up and running in 10 years, but even if it meets the objectives of the experiment, it is not going to be anywhere near production ready.
" I certainly wouldn't invest my own money in Helion"
But what if in 10 years time you're sat there thinking, "fuck, I wish I'd put a couple of grand into that at the beginning!!"
I wonder how many people "turned down The Beatles" and chose not to invest in MS, Apple, Google etc and were in an actual position to have done so?
I suppose it's a bit like long odds at the horsies. How much can you afford to lose? Even with long odds, you still need to know enough about the game to pick winners once in a while. Most of us would be picking the losers every time and end up broke unless the luck struck at the right moment. The Warren Buffets of the world got rich by investing in what they understand. A few got rich by taking a punt and hoping, but mainly investing in what they understand so they never lose everything. After all, who of us would have invested in lasers when it was little more than a novelty with no obvious real world commercial uses? :-)
"I wonder how many people "turned down The Beatles" and chose not to invest in MS, Apple, Google etc and were in an actual position to have done so?"
The people that lost out are completely overshadowed by all of the people that lost vast sums on things that didn't go anywhere and companies that completely failed to make a go of it. Enron, Theranos, Solyndra, etc, etc. There's also the people that bought Tesla stock at the peak. If they would have bought the stock 10 years ago and sat on it making next to nothing, they'd still be laughing all the way to the bank, but that's not proper investing. You expect to see a return much sooner than 40 quarters.
Back in the 1990s, when it was the right time to invest in Apple, how would you have picked it in favour of companies like Amiga and Atari who were equally big in the 1980s, had lost their way in the 1990s, and are now only found in museums and retro-computing YouTube channels?
As for Google, I don't think there were many people choosing not to invest in it, obviously there were some, but again, there were plenty of other similar-looking companies at the time that people threw their money at, who are now dead.
"but I hope Microsoft has picked a winner here, we could certainly use commercial fusion power sooner rather than later."
I'd be much happier seeing M$ investing in projects that aren't ongoing science projects. Many LFTR business plans could use some investment and also some horsepower behind them to get their designs past the Atomic Energy Agency so they can be built with real fissiony stuff. Many of those fossil agencies won't approve anything that hasn't been done before so are a major impediment to any advancements. If the kids of the agency higher ups are banned from their xBox accounts, maybe some proper debate will be forthcoming. The whine from a denied gaming addict is worse than nails on a chalkboard.
There are lots of energy projects that can be done now. They aren't exciting, they aren't glamorous, but they work.
So they need to produce a fusion reactor to produce fuel for the fusion reactor. If they have the first one working why would they need the second?
I think I'd want to see a proof of concept running continuously producing excess energy before signing up as a customer.
Nevertheless I'd genuinely like to see them succeed.
That struck me as well. The Helium-3 stuff does sound an awful lot like a magic bean. Still, I'm glad some of Sillycon Valley's money is going into physics for a change and not another (advertising-based) "disruption".
ITER should have some news over the next year or so. And, for all its problems, it's the closest to getting fusion at scale of any of the approaches.
"Still, I'm glad some of Sillycon Valley's money is going into physics for a change"
I love to see money being invested in science, but I hate to see it going into junk science where there's a high probability of it being wasted. It means that several much more viable programs don't have a shot at those funds and the biggest, flashiest projects collapse into black holes when they fail sucking in a whole bunch faith in the rest of science.
I'm much more the tortoise. Slow and steady is my mantra.
Don't hold you breath waiting for ITER.
First plasma was scheduled for 2025 but it's accepted it will slip by a couple of years, though some sources are predicting 2031. Full power deuterium tritium operation was not scheduled to start till 2035. Whilst the current issues do not necessarily push D-T out by the same amount, 2035 does look very optimistic.
"So they need to produce a fusion reactor to produce fuel for the fusion reactor. If they have the first one working why would they need the second?"
Sounds like the first reaction is a net consumer of energy, and the second (the helium-3 one) is a net producer of energy. We can do fusion today, that's not the hard part. The hard part* is generating more energy than you consume.
*relative to all the other difficulties of working with insanely high pressures and temperatures and exotic physics, that is.
Personally, I think that will be relatively easy. Building it economically, then making it run continuously without breaking down, so that you can make power for $0.05/kWh over 30 years will be much harder. I don't rate the chances of the ITER approach, I doubt the engineering problems can ever be solved at an economic price.
"So they need to produce a fusion reactor to produce fuel for the fusion reactor. If they have the first one working why would they need the second?"
Efficiency. If you can use the less efficient process to create a fuel component that lets you have a reaction that so much more efficient it swamps out the added costs, Bob's your uncle.
Everybody likes donuts - cos the maths is simpler and it was all worked out 50 years ago
They have a problem that you can't run them continually so tricky for power plants and they work better as they get bigger - in fact small ones don't work at all until we got the new super-duper conductor (*) magnets
* not the official name, but I wish it was
I see. The company responsible for the overwhelming majority of the world's buggy shitware is going to diversify into the production of nuclear reactors. What could possibly go wrong? The attached icon might be a hint.
Maybe Clippy will make a return to help out with the design and operation of Redmon's fusion reactors?
So Helion is on their seventh generation design and it still doesn't work? What other technology gets to seven iterations with no success and still attracts massive investment? I'm sure their are some, but I remember reading about fusion reactors in grade school in the 80's and it was just 'ten years away' from working.
Short answer is that there are always gullible people with more money than they know what to do with, and fall for the line it is different this time. In govt research, there are always 'other applications' that really fund the research, and the fusion energy app is just an add on to show they 'care'.
Or, sometimes the research produces tech that can then be repurposed into something else entirely. I was just watching an episode of Aussie Inventions That Changed the world. Some obscure astronomer wanted to detect the explosions of mini black holes. He and his team had to develop signal processing far beyond what the best supercomputers of the day could do. They never did detect the black hole explosions, but the digital signal processor they created was too good to waste. So they went off and invented WiFi with it{**} :-)
(Then the Yankee Pig Dog[*] corporations tried to rip them off and they got rich from the patent infringements fines and licensing fees)
* sorry, couldn't resist :-)
** grossly simplified explanation, obviously.
Hmm, Megajoule sounds big.. But it gets very humble, when one remembers that 1 kWh is 1000 * 3600 J (W= J/s, or J = W * s) = 3,6 MJ.
And: for nearly 50n years I've time to time discussions with colleagues who work in the field of fusion: and it is alway 20 years to a running fusion reactor
"If they don't deliver they'll probably go bankrupt so there's nothing for Microsoft to get reimbursements from."
I'd expect that if M$ took the money they will invest in this and used it to add solar panels to all of the roof tops and car park canopies at all of their facilities, they'd see a much bigger return. They'd also be able to provide free EV charging for their employees who would then be able to own an EV even if they live in an apartment and don't have off-street parking where they can charge at home.
Author of this article did a good job of due diligence. Dan Jassby is the goto guy if you want to get the unvarnished truth about these fusion startups. He has an axiom of alternate concepts in fusion that goes something like the hyped desirability of any concept is inversely proportional to the number of fusion neutrons generated. The helion concept uses the DHe3 fuel cycle which is considerably more difficult to get net power production than the more common DT fuel cycle, not even counting the rarity of He3.
However, most of our successes around fusion have been from the standpoint of harnessing it to blow up cities, wipe humanity off the face of the planet, and other similarly gruesome ends.
Fortunately those successes have been purely theoretical and we haven't actually razed any cities down to bedrock using fusion weapons. Probably more luck than judgement, but those of us who grew up in the Cold War with endless talk of four minute warnings to complete annihilation are grateful to have avoided that fate.
"Magnets, again with the magnets"
Magnets fascinate people to no end. Since most people don't read much science, they don't know the difference between energy and force. With a magnet, you get a magnetic force, but in and of itself, there's nothing to be had from it. Any free energy scams I see that show the 'inventor' using magnets I know are a scam from the first couple of seconds. It's just as funny when a non-engineer doesn't know the difference between voltage, energy and power. You can assemble a giant high voltage battery from almost flat 9v's and measure that very high voltage, but it doesn't mean you can get any work out of it.
First you have to be able to get a sustain reaction that's nicely contained so it continues working for more than milliseconds. Second, this design needs all of the bits to breed a different fuel component. Once all of that is worked out (and gets past the regulator's automatic "no") there still remains getting the power conversion part working. If all of the science was well defined, the engineering would take another 20 years. Who knows home much time calculating the bribes and battling all of the lawsuits would add.
A number of us could have access to an existing very large fusion reactor (2 x 10^30 kg). I am fortunate that I live where I have a moderately efficient energy collection device on my roof (17 m^2) that generates 5.2 MWh of electricity a year, which is similar to my consumption. It cost AU$4,000, but as it only works about 34% of the time, back-up and storage to cover the down time is currently a bit expensive.
One problem that I can see is "How will the energy oligarchs make money off a system that is relatively inexpensive, and will last 25 years?"
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
