The chase for fusion energy
nature.com
nature.com
https://media.nature.com/original/magazine-assets/d41586-021...
Getting to a Fusion Pilot Plant
Still looking around you can find many seemingly reasonable designs. Using fusion bombs to heat up rock and then extracting energy is completely buildable today, but not currently economically viable. The history of inertial confinement fusion had more to do with validating models of fusion bombs than a realistic energy source, but things have ramped up enough that it’s starting to seem more possible though still unlikely to work any time soon. Outside of that things get odd. Simply firing a fusion bullet scales really well, but runs into serious issues designing the gun. A coil gun in high vacuum might in theory work, but trying to actually build one is another story
Piston fusion is quite similar to inertial confinement in theory. Of course it’s a long shot that’s closer to snake oil than a practical design, but high risk reward long shots aren’t scams as long as the team is upfront about what’s going on. Sometimes physics just doesn’t work the way you think it would. Sonoluminescence is one of my go to examples that seems stranger than fiction.
PS: A liquid vortex is actually one method for building a blanket for steady state magnetic fusion setup. It solves several real problems, but is mostly about economic viability not achieving break even.
1. Lazard has published a (favourable) preliminary levelised cost of energy assessment for it.
(2. Which implies that there have been convincing demonstrations.)
3. Fusion companies are spending their time on working with regulators and explaining their systems to investors (manufacturing, construction, operation, decommissioning and disposal), and their money on bringing up core systems and assuring the public that it is safe.
https://www.youtube.com/watch?v=g18xpYXamAs
IEC (Inertia Electrostatic Confinement) fusion can be done, with a high-vacuum system, a high-voltage DC supply, and a vacuum-chamber with a high-voltage feed-thru. It is dangerous, but doable, and I believe I had the first working, amateur-built fusion reactor in Canada. One can confirm that fusion is occuring, by using sensitive chemical neutron detectors that workers in nuclear fission plants typically carry. I used BTI (Bubble Technology Inc.) detectors, which look like little test-tubes, and show small bubbles in a gel-like liquid, if exposed to a neutron flux.
My reactor was never able to produce more than 10^6 to 10^7 neutrons per second, which is a very weak neutron field. I used a small "lecture bottle" of non-radioactive deuterium (heavy hydrogen gas), and it worked well. I had to construct a deuterium injection system, using some high-vacuum valves, and a gas-feed-thru on one of the vacuum-chamber ports. The vacuum-chamber itself, and the high-vacuum diffusion pump (which used a mechanical vacuum pump, connected to a device which uses hot-oil vapour) and the high-voltage DC power supply, were all sourced from surplus equipment found on eBay, and imported from USA into Canada. The importation of each item was an adventure in itself.
One needs to put about 15,000 to 25,000 DC volts negative, on the confinement grid, and the vacuum chamber needs to be evacuated to about 10^-4 torr or better. I estimated the device needed mean-free path length of almost a 1 cm, for the D2 ion circulation, which is a very rarified environment.
What is really interesting about fusion, is how difficult it is to make happen. A palladium-wire grid would be better for the spherical confinement grid, but I used stainless steel. (Palladium is expensive, I had no source for it, and I am not Tony Stark...)
But the reactor did produce nuclear fusion events, using the D2 gas, and the high-voltage charge on the confinement grid. This technology was first invented and patented by Philo Farnsworth, an American researcher and inventor, who in the 1930's, invented electronic television. He invented the first electronic TV camera - the iconoscope - and fought with Sarnoff at RCA for years over the patent rights for electronic TV. Farnsworth's IEC fusion reactors also worked, but attempts to scale up the technology were not successful.
The various attempts at creating commercially viable fusion reactors are interesting. We keep getting close - but nothing that produces a useful level of net energy output, has yet to demonstrated, as far as I know. I sincerely hope some sort of breakthru is possible, at some point. My device - and most of the IEC devices - are able to produce a significant amount of neutrons, and research efforts at the University of Wisconsin, Madison, were able to demonstrate a powerful IEC reactor, which produced an intense neutron flux, sufficient to ablate the walls of the vacuum chamber (the chamber wall looked like swiss-cheese, under electron-microscope examination), so we know this technology is able to produce a serious amount of fusion events.
But taking this approach and using it to generate a net-energy output, has not been possible. My device was roughly 1500 watts in, and 1500 watts of heat out, plus maybe one or two extra milliwatts (or maybe a few hundred nanowatts). Not even enough to really measure.
I have followed all the various attempts to get some net useful energy out of fusion reactor technology, and I truly hope something becomes possible. We seriously will need this technology, and we are so close to having it. I wish all these researchers - public and private - the very best of luck.
LFTR scales down, has working examples (fit in a closet in ORNL)
fuel availability: lots of thorium, ignition fuel is a problem though
vessel lifetime: lots of caustic stuff in LFTR, but fusion is high temperature and will suffer vessel degradation from neutrons, won't it?
Safety: LFTR is meltdown proof due to liquid fuel and plugs, and seems to consume all its nuclear fuel (compared to solid rod). Fusion is "clean" but makes the reactor eventually radioactive?
Cost: Fusion is a ???, LFTR is undeveloped, and both will be a decade+ down the road during which solar/wind/battery will get substantially cheaper, moving the goalposts for cost/investment.
Whoever figures out the hypermagnet will be just one step away from producing net fusion that is useful. I'd say divert all attention to that instead of new tokamaks. We know tokamak will work (economically too), if it only had the secret ingredient of unbelievably powerful magnets.
Because here's a presentation by the head of MIT's fusion program, who introduces those scaling laws and concludes that the latest commercially-available superconductors are indeed sufficient for net power, from a reactor of reasonable size and structural strength:
Its pretty simple really. A charged particle when it slams into another will gyrate in a magnetic field. If that magnetic field is twice as strong, the gyro-radius is half as wide. And the containment vessel would require half the volume. It takes many many strikes to fuse two nuclei. That is why it is necessary to have ultra powerful magnets. Because it might take dozens or hundreds of strikes, each one potentially enlarging the radius of the circulating plasma. Maybe stellerators have an edge but I don't see it with my limited knowledge. But simple brute force ought to get 'the job done'.
The relevant scaling law is that tokamak energy output goes up with the fourth power of magnetic field strength. According to the presentation I linked, that allows ITER-level output from a reactor a tenth the volume of ITER.
Assuming that works out, a reactor somewhere between those two sizes should get us to commercial levels.
The magnetic field is the foundation of fusion energy at this time sans xray compression.
The reply, and article, reference CFS, a public/private initiative by MIT, where the gentlemen whose laws you are referencing had a major breakthrough via a materials science breakthrough that allows the ITER magent you're discussing to be produced a couple orders of magnitude smaller, changing all of the problems.
One could argue that these funds are being poorly invested, but that begs the question: why now? Why didn't we see this money being 'thrown away' 10, 15, 20 years ago? What's changed that those with the means to make such investments believe that now is the right time to be putting money into long shots?
And so I'm left to have one simple hope: maybe the very rich people have hired people smarter than me to assess if there's a chance in hell this will work. It sure seems like a lot of people suddenly think they can do it.
I’ve heard that before. Are we also going to have room temperature superconductors and usable quantum computers?
Elsewhere in the world, people have to scramble to get money to do groundbreaking work and in the US you get thrown after you for knowing the right people.
I'm not saying the other places in the world are good places to be. Caring about every penny you spend is definitely not how you want to run a company doing groundbreaking work, but there's definitely a middle ground.
This one wasn't quite as egregious as the other flamewar tangent in this thread (now collapsed) but it's still a change of subject into something less interesting, more generic, and more inflammatory—and that's what the site guidelines ask everyone to try to avoid here. It makes discussion more tedious, more repetitive, and usually nastier.
There's lots of interesting material in the OP and clearly that's what a thread like this should be engaging with.
Yes there’s a lot of dumb venture capital going around, but there’s also a lot of money going towards trying things beyond searching for the next unicorn tech fad.
There are few other places just now to train up plasma fluid physicists. The main beneficiaries, though, are the contractors. This is a gravy train that just doesn't quit. The reasons so much money easily goes to it have far more to do with corporate welfare for otherwise mainly-military contractors, and maintaining a population of hot-neutron physicists ready to draw on for weapons work, than any conceivable expectation of competitive power generation.
But the main product of the Tokamak-adjacent projects is lies. They cultivate the confusion between "Q>1" meaning more kinetic energy of neutrons emitted than microwave photons injected and magnets energized, vs "Q>1" meaning electrical power to the grid exceeding grid power injected. There are at least two orders of magnitude between the two Qs. Achieving the former leaves you very, very far from the latter. ITER itself makes no pretense of ever producing so much as one watt-hour of grid power.
Thus, any system resembling ITER (particularly ITER itself) is a technological dead end. A fusion power plant that relies on hot neutrons would necessarily cost many, many times as much to build and operate as a comparable fission plant. Big fission is already not competitive (perhaps mainly from deeply entrenched official corruption, but so what?) and gets less so every day. We have yet to see whether small-scale fission can work, economically. It has failed before.
The Helion and TAE systems would be "aneutronic", generating power electromagnetically, without a side trip through neutrons, heat, and a turbine. They have a chance to be useful, because they would be cheap to operate, even at a small enough scale to leave little scope for official corruption. TAE is taking the harder road, chasing hydrogen/boron fusion, which reactants are both extremely abundant. Helion is chasing hydrogen-2/helium-3, which is more plausibly achievable, but helium-3 is very scarce, so they would need to synthesize it themselves by also fusing hydrogen-2.
The Princeton FRC reactor project, not mentioned in TFA, is working on a shoestring NASA budget, hoping to loft a 2 MW space-probe propulsion test in 2035. That might work, and they would not need much helium-3 for that. They could probably get something done much earlier if ITER were not pissing away all the money, but moving money from public to private purses (as with NASA's SLS and DoD's F-35) is ITER's true purpose.
As for aneutronic fusion, there are reasons to be skeptical it can even in principle be made to work (see e.g. PhD thesis by Todd Rider). I'd be happy to be proven wrong, though.
Any impossibility found for aneutronic fusion does nothing to improve the commercial viability of hot-neutron fusion. If aneutronic fusion can't be made to work, the outer solar system will remain a lonely place.
There is a real risk that by focusing on the wrong Q factor you sink all your budget in to plants which can never work as a real power plant.
Still I am hopeful that ITER will be sufficiently valuable as a research system that we do move closer to commercial fusion power.
Don't really understand your point.
This lesson also applies to ITER.
when a scientist tells you 2050+ it means they don't know and everything is being made up as they go along. It means 'outside the realm of any realistic estimate, but close enough to feel tempting so you don't immediately remove funding'.
One project started in 2007 planning activation in 2025 and planning reactions in 2035. The other started in 2015, activated 2017 and steady state reactions have been delayed from the original Sept 2021 plan to 2022. It's not even close the results so far.
> Caring about every penny you spend is definitely not how you want to run a company doing groundbreaking work, but there's definitely a middle ground.
Private business allocates funds toward profit. Apparently, investors don't see much potential return on additional funding compared to, e.g., electric cars, phones, online retail, etc. An investment's value can be calculated, in one sense, as risk x potential profit. Given the latter is very high, investors arguably must perceive the former as very poor, though possibly they perceive that the companies don't need more funding at their current stages (i.e., it's marginal return on additional funding that matters for investors, not total return on all funding).
The people of a country, through government, can allocate funds toward work based on importance and regardless of profit. Such work, such as basic science, is where the market fails. (Probably much of the basic science behind fusion was already funded this way.) Per the OP, we've allocated $22 billion toward ITER. So fusion is pretty well funded.
It will be just like deep learning. There wasn't really much investment in neural network research and hardware until the field started having some big breakthroughs, at which point cash started flooding in and the community got extremely excited. I think we'll see fusion breakthroughs in the next decade or two, and it's going to be interesting.
Maybe radiation-resistant robots will be the next investment bubble after that.
From my point of view, most of the fusion startups have zero chance of success at delivering practical fusion energy. Some might have spinoff technology that's worthwhile. Just a few might still succeed at practical fusion (Helion is the one I'm least down on).
ITER is a shining example of how to do this maximally wrong. It is a path only to ongoing expenditure, not useful energy.
There are some smaller startups doing interesting things such as using superconducting coils with much higher field strengths. They might be able to leapfrog ITER. If any of them can demonstrate net positive energy, then and only then the human race should throw tens of billions at that.
ITER is an example of what happens when you dither and drag things out for decades. If member states had made a major commitment to not only fund ITER, but to have it built within 5-10 years and actually done so, it would have been money well spent.
If you wait long enough all designs for anything will become outdated. An operational ITER would have provided, and still can provide, crucial experience and data. But the cost-benefit ratio is now much poorer for the delays.
This is the lesson of SpaceX--not a lesson about financial frugality or simplicity of engineering. SpaceX has spent far more money than will be spent on ITER, and its designs are not known for their elegance. The world is awash in both cash and talent. The lesson is about process and pace and iteration--all that cash and talent is wasted if you don't apply it. Optimize for time. Emphasize process over product, as the former makes for the latter, not the other way around. When you have confidence something is worthwhile, as with ITER at the outset, don't hold back.
On the other hand, ITER is breaking entirely new ground on the engineering side, and it even requires fundamental research. It is not even known if what ITER is setting out to do is fundamentally possible given realistic engineering constraints instead of idealized models.
The Raptor family is a major, major innovation.
Yeah we should hold out before spending tens of billions on fusion research... Meanwhile, 1.5 trillion gets spent on a failed fighter jet program...
The amount of money spent on other projects is so, so much greater than that spent on fusion.
The point GP is making isn't that fusion is a guaranteed win -- GP gave examples of other failures. It's probably their belief that there's more money in fusion than Magic Leap.
Considering the potential payoff of fusion I believe this money is spent way better than than on things like returning to the moon or a lot of defense projects.
ITER's goal is fine, except that there's no way to extend what ITER is doing to a power plant that could be economically competitive (not even competitive with fission, never mind renewables). A fusion reactor that could be competitive must take a different approach, like one involving advanced fuels. This would make ITER's goal much harder (and would likely rule out tokamaks entirely), but the long term goal easier.
Show me examples of startups and private ventures that did completely fundamental research required for their engineering efforts on a product. At least one that wasn't mostly based on technology and research created by the public sector.
I really don't understand how neoliberalism got so powerful as an ideology to force people to think that smaller startups are really better at fundamental, unprofitable research than tax-subsidised government research programs.
There is no place in the neoliberal "free market" ideology for private companies (even less smaller startups) to pour money into expensive research for 10-20 years with the unknown if it's even feasible and possible, much less profitable.
In almost every thread here on fusion someone eventually says something along the lines of "what we need for fusion is a Manhattan project", but ITER by itself is costing more than that in inflation adjusted terms. Fusion has soaked up massive investment and continues to do so.
Fusion is undoubtedly critical to our future, but I think I can see why the investors don't think it's a profit machine in the same way as a new commercial gadget.
I dont see why you couldn't patent a fusion reactor. You can patent the underlying science, but the physical realization is patentable. If you manage to pull it off, you are probably super rich. However the risk of failing is much higher, and the time horizons are much longer than a silicon valley startup.
The journo always has to point this out, despite SpaceX simply being the cheapest, best option even for government launches.
I don't think that's a good comparison. SpaceX didn't develop ground-breaking firsts like Sputnik and the moon landings; they took something we've done for 50 years and do it more efficiently. Government agencies are the ones with helicopters flying on Mars, who visited Pluto, developed and operate the space telescopes, etc. Fusion energy generation is like the Manhatten Project.
Ingenuity was build with multiple non Government contractors.
The Lunar Module was developed and built by Grumman under Government contract. Rocketdyne developed the J2
Boeing/North American/Douglas designed and built the Saturn 5. Why do people attribute all of these vehicles to the Government when they are were contracted to private companies to be built?
Its very strange that people deny that SpaceX created modern reusable boosters and they fall back on the damn DC-X project when a) McDonnel Douglas was the manufacturer of it b) it was a prototype that never made it to orbit, c) is widely considered a failure
Meanwhile SpaceX is absolutely trouncing every other company with Launching ability while Beoing, given every single possible kickback from the Government can't launch their Starliner for another year after already a year of delays.
They have developed the only full flow staged combustion cycle rocket that has ever gotten off the ground, something even the Soviets couldn't manage.
might want to actually read the wiki before downvoting
Because the alternative doesn't fit their ideological world view.
- First full flow staged combustion engine that actually flies
- horizontal flight, vertical landing (belly flop)
- booster reuse on an orbital class rocket
Indeed, and I'll observe that the significance of this one is often lost on more casual observers.
> took something we've done for 50 years and do it more efficiently.
One can argue that this is "all" that happened in AI in recent years. We took something that was arguably done before, in some similar way or shape, and made it practical. But this is HUGE. It's not a minor accomplishment! It may very well be harder than doing some prototype for the first time.
SpaceX is landing rockets at sea on comically small platforms.
Efficiency gains and ground-breaking firsts aren't mutually exclusive. For example, I've personally invested in several aerospace startups, and the general public is not remotely aware — even now — of the extent to which SpaceX's efficiency gains have utterly revolutionized space access. It's no exaggeration to say they've driven a step function change in the sorts of companies it's possible to build in this industry.
Very often, ground-breaking firsts are only achieved once efficiency along some axis passes a critical threshold; I'd argue it's fair to compare fusion to SpaceX in this regard.
This includes 25 Dragon cargo launches (24 successful, and include two COTS demo launches) and 5 crew dragon launches, (two demo, one in-flight abort test, and 2 operational launches). There have been a few unmanned NASA launches.
So, it's not really correct to say NASA is what's holding them up. They have launched a lot of comsats for other companies, defense payloads, smallsats, foreign government launches, and 29 Starlink launches.
And I'm not sure if you consider launching starlink to be "commercial", if you considered it to be an internal customer of the launch service then it seems like it should count.
I would say the ground breaking things were done by amateurs like Mr Von Braun as children playing with rockets.
What governments(Hitler first, then US and Russia) did was escalating something that was already proven and worked, and that was proven my amateurs.
The flights of Otto Lilienthal or the Wright brothers were extremely short, but they were the groundbreaking things that governments later escalated very fast.
I certainly believe that break even will be proven by private companies, not by Governments bureaucracies that continuously say that fusion in 30 years is reasonable.
After breakeven is proven, there will be a tsunami of investments from Governments just like with aviation or rockets.