Fusion startup plans reactor with small but powerful superconducting magnets | Hacker News Reader

Fusion startup plans reactor with small but powerful superconducting magnets

sciencemag.org

311 pointsrbanffy332 comments

The MIT Club of Northern California posted a video a few years ago on this exact topic and reactor:

https://www.youtube.com/watch?v=KkpqA8yG9T4

This thing came out of MIT, at least according to the video, and was really the collective efforts of a bunch of MIT grad students who made the breakthrough partially by taking a very Silicon Valley startup approach of using off-the-shelf parts, experimenting with new ideas, and starting small. I don't know if Professor Whyte framed it that way to appeal to the crowd or not.

I haven't seen that one, but here's another talk on the same topic in the context of fusion energy broadly, by some people involved https://www.youtube.com/watch?v=L0KuAx1COEk

Here's a more recent (1 year old) talk that gives a bit of a progress update: https://www.youtube.com/watch?v=rY6U4wB-oYM

Here is an even more recent (last month) update https://www.youtube.com/watch?v=h8uYNhevRtk

tl;dw: SPARC is on track to Q=9, and there will be a magnet demonstrator in June, this year

How in January 2021 after a full year of pandemic lockdowns and video conferences, has someone hosting something like this not invested in a microphone?

They are a post-doc? They probably don’t see it is a priority in their budget.

Most laptops and phones already have a microphone or two.

You missed "rather terrible". The other day someone on our sales team did a presentation to the company, and they used the microphone on a $3 headset - I practically couldn't hear half the things they said because the audio was so muffled. This is someone who has calls with people all day!

If you record any audio or make calls with people outside your organisation, for the love of God, please invest in some sort of microphone upgrade. Even the $20 no-name Chinese microphones you get on Amazon are miles better than what is built into laptops.

While I agree with you get a better microphone. Buying Chinese crap is a lottery I bought my self a Chinese condenser microphone 60usd that thing is worse than my iPhone headset. My lesson when it comes to buying quality is to only buy name brand things.

> If you record any audio or make calls with people outside your organisation,

... please ask your employer to invest in some sort of microphone upgrade. Unless you're self-employed, paying for work equipment is not your responsibility.

Sound dampening as well. Fabric lined partitions, just sitting angled to walls, etc. $3 headset isn’t the end of the world by itself but capping at $3 total invested in setup is surely going to ruin the sound.

And if you're on desktop make sure to plug it in the right jack port. I spent 3 years wondering why all the mics I bought were garbage until I realized it was my front panel that was garbage and was messing up the signal.

Because it sounds perfectly fine for them. Nerds tend to obsess over a narrow field of interest and their field is not audio quality and gear, unlike say, yours.

It's not just about geeky interests. I am not a native English speaker, and I was in the artillery (so I have mild hearing loss). I can follow presentations with good audio quality without effort, but at times this video is damn near incomprehensible to me. At least the subtitles are good.

Maybe researchers and professionals is more appropriate than nerds.

Take a look at streamers and Bill Gates collaborating, then compare that to Gates appearing in remote news interviews. Quality difference is significant with TV being always worse.

As far as I can tell that framing is probably for crowd appeal. The real breakthrough here is the availability of high-magnetic-field superconductors, which push the size requirements for break-even down a lot and so put construction costs in reach of private investment.

(This is why MIT is at the center of the work - they have a really good materials science program that's good at working with these finicky ReBCO tapes.)

An important part of the article mentions bringing down the cost of high-temperature ReBCO tape conductors, where SPARC itself needed more than entire than what existing companies would expect to produce in an entire year. Existing technologies using superconductors like MRI machines, large transformers/motors, and synchronous condensers for power generation have each benefitted from bringing down these costs as the use cases mature.

ITER's estimated timeline having a working reactor by 2025 is ambitious, but is also supply constrained in that they're projecting the need for more Nb-Ti and Nb-Tn exceeding current yearly production amounts as well. For reaching the end goal of affordable hyperscale energy production, it's promising to see demand increase in order for new competitors to invest in related research projects.

What is Tn (in Nb-Tn)?

I'm assuming it is a typo/thinko for Nb-Sn.

From Wikipedia:

For higher magnetic fields, higher-performance, but more expensive and less easily fabricated superconductors, such as niobium-tin, are commonly employed.

I have some hope for this approach, but as the article points out making reliable magnets is really the key. When a fusion reactor quenches, the plasma will basically eat the reactor through "hole" in the field. I keep hoping an effort to manufacture ReBCo coils directly will be successful, it would both make them less expensive and likely more reliable. However I expect it would require something like a 5-axis 3D printer capability.

Lastly, fusion power is one of the possible 'good' future events (unlike climate change, or nuclear war) that give me hope for the future of the planet.

Controlled fusion seems one of the few awesome future things that's unambiguously good.

Particularly if there's some way to capture carbon that's otherwise impractical due to energy requirements.

Fusion power will likely not be significantly cheaper than solar and wind for this purpose. Fusion plants are probably quite expensive and won't have an infinite lifetime. The big benefit is that you don't need batteries, but that doesn't matter if you want to use the power for carbon capture.

Why is is unambiguously good? There's great reason to doubt it could be anywhere close to economically competitive. That makes its value pretty ambiguous, in my view.

I think that an argument can be made that it will always be economically competitive. That reasoning includes the 'fuel cycle' is non-waste producing, the economics of other energy sources continues to rise thus creating a wider window for economic recovery, the liability associated with fusion will always be less than the liability associated with fission, and as the carbon externalizations of fossil fuel are priced into its production it will become in-economic as well.

Well designed fusion power should come in at or below hydro-electric power without the environmental impacts or risks associated with dams.

This is a bad argument. The fuel cycle is a small fraction of the cost of fission power. The main cost is the cost of building the plant. Fusion reactors will be much larger than and much more complex than fission reactors, per unit of thermal power output, so they should be more expensive. Fission reactors today are uncompetitive, even with liability artificially set to a small value, so that argument fails too.

The reason nuclear fission is uncompetitive is its high operating costs. The cost to build the plant is a sunk cost: if the reactor is expected to be reasonably profitable to operate, you can fund building it. If the reactor is not expected to be reasonably profitable to operate... Good luck.

That's what fusion aims to solve. The fuel is plentiful and you can easily buy it; the same can't be said of uranium. It's also much safer to run than fission, and produces vastly less dangerous waste.

Perhaps your argument is that solar and wind are sufficient to power humanity's needs, without fission or fusion. That's debatable. But compared to fission, fusion is theoretically better on some pretty critical metrics — if we knew how to build a fusion reactor, which we don't yet. If you assume solar and wind won't be sufficient, fusion seems worth research.

> The reason nuclear fission is uncompetitive is its high operating costs. The cost to build the plant is a sunk cost

This is nonsense. You can't ignore the cost of building plants, when trying to determine if a technology is competitive. Nuclear plants have to be paid for; they're not given to us free by the Nuclear Fairy. If you include capital and financing costs, you will find they contribute more to the cost of energy from the plants than do operating costs.

Economics of a nuclear reactor https://youtu.be/cbeJIwF1pVY?t=58

Yep (although he's understating the capital cost of nuclear plants.)

> The cost to build the plant is a sunk cost: if the reactor is expected to be reasonably profitable to operate, you can fund building it.

You're missing one important part: the reactor must be profitable enough so that it can recoup the build cost for its lifespan. The lifespan of fusion reactors, even disregarding failure modes, will be severely limited by the high energy neutrons which it generates, and which turn all materials brittle.

And maintenance for the brittle material will mean stopping the reactor, and sending in robots to dismantle and carry away the brittle radioactive walls, and build new walls in place. It may well be about as cheap to scrap it for parts (for those parts that haven't been turned to radioactive Swiss cheese) and build a new one.

The various links to the talks by the head of the ARC/SPARC project in this comment thread talk about the size of the fusion reactor, as does wikipedia:

> ARC is a 270 MWe tokamak reactor with a major radius of 3.3 m, a minor radius of 1.1 m, and an on-axis magnetic field of 9.2 T.[2]

> The design point has a fusion energy gain factor Qp ≈ 13.6

So the reactor is about 25-30 feet in diameter, plus the steam plant / FLibE processing.

Have you seen the size of a nuclear cooling tower?

That's not the reactor itself. I'm talking about the cost of the reactor. The reactor part itself of a fission NPP is quite cheap. Replacing it with a much larger fusion reactor means your containment building (and one will be needed) is larger and more expensive, and you still need the same things (turbines, generators, cooling towers, switchyard) outside the nuclear island.

Next generation fission does not need cooling towers. As soon as you go to a high temperature design, they can run perfectly fine in the desert.

Not familiar with that design. Unless you dump heat, any generator will melt down. How does the 'next generation' dump heat? A cooling tower does it very quickly, and can be scaled arbitrarily large.

Wind and solar have zero fuel cost, but they cost money to build. A fusion plant is always going to be a rather high-tech, expensive thing.

Wind and solar are non-continuous and impossible to use in space, underground, underwater. Fusion potentially could be used everywhere as a relatively high density energy source.

Solar is impossible to use in space? Please explain this claim.

The energy density of fusion is inferior to fission, when you take the mass of the reactor into account (as you must, if you are worrying about the mass of your system.)

Yes, anything beyond Mars is impractical for anything but very small missions.

Is the way from here to Mars not "space"?

What if fusion was combined with renewables? Meaning fusion served similar role as batteries. Storage costs are really high so fusion might be competitive in that space. What I am thinking of is using output from fusion in molten salt storage.

Generators are coupled to this thermal storage. When wind and sun is good, there is a continuous build up of heat stored in molten salt by fusion reactor. When it is low, net heat gets drained from molten salt storage to power steam generators.

At this level of planning, it would likely be easier to just lay the superconducting undersea cables and start building massive solar on opposite sides of the planet.

The reality of energy today is we basically have the technology to cleanly power the human race for the rest of it's existence - politics and tribalism is why we can't.

> What if fusion was combined with renewables? Meaning fusion served similar role as batteries.

Fusion, like fission, is a high fixed cost, low variable cost power source. The cost of energy from such a source escalates rapidly as capacity factor goes down. They are very poorly suited as backup sources for intermittent use. Hydrogen (made by electrolysis from renewables and stored underground, then burned in $400/kW turbine power plants) would be much cheaper as a backup source.

You are right, in short term, fusion has not much value. But in (very) long term, unlike fission, which no matter how incredible it is, is still only 'planetary-level tech', fusion is the first 'stellar-level tech' we've unlocked so far. If we ever plan to go somewhere else than Solar system, fusion is the key.

Beamed power will likely be much more practical for interstellar missions, or even large activities in the outer solar system. In particular, beamed power minimizes entropy generation at the target (and can even actively cool the target), which is vital for high thrust at high Isp.

Molten salt reactors and thorium fission is pretty exciting too.

Making reliable magnets is A key, but it's by no means the only one. Fusion faces many grave obstacles even if the magnets were totally reliable and cost nothing.

What are some of the others ? I’ve heard even if everything worked the generated neutrons would eventually destroy the reactor.

That's one. Another showstopper is that all the energy has to be radiated through the wall of the reactor. By limits on this areal power density and the square cube law, the volumetric power density of fusion reactors will suck. Compare ITER (0.05 MW/m^3) or ARC (0.5 MW/m^3) vs. a commercial PWR fission reactor primary reactor vessel (20 MW/m^3). Stronger magnets don't save a reactor from this.

The large size and complexity of a fusion reactor also means their reliability is a huge problem. There are many parts and joins there, and the machine will be so radioactive hands on access will be impossible. A single leak of coolant into the vacuum chamber renders a fusion reactor inoperable (while a fission reactor can keep operating even with multiple fuel rod leaks.)

I’ve heard the sun has a very low power density it’s just that it’s massive. Maybe it’s just easier to create a star and live in orbit around it

The key point is the Sun is already paid for. :)

Are you certain? ;) Sounds like the plot of a sci-fi in the vein of Hitchhiker's guide where galactic aliens show up with an overdue bill for the sun.

Or find an existing unsettled one and move in!

Tricky bit will be getting to that nearby star...

If only we had an engine with high enough Isp to accelerate to a decent velocity! Maybe something based on nuclear fusion...

That's not that hard actually. We already know how to do fusion, we just have a hard time confining it into a reactor. In space you don't have that problem, bombs work fine.

It's true, but at least partially deceptive because the fusion only happens in the core, which is <2% by volume

The rest is needed to keep the pressure and the magnetic field.

Well .. we do already live in the orbit of a star. (intended joke?)

And we have lots of desert land, that could be covered with solarthermal plants.

https://en.m.wikipedia.org/wiki/Desertec

Also not easy, but sounds more predictable than a possible fusion future.

A good comparison would be, what's the capital cost of a hypothetical fusion plant, compared to equivalent renewables (e.g. battery-backed solar/wind)

https://arxiv.org/abs/1409.3540, Table 11 and Section 6.6.

Yikes. For a 250MWe reactor:

> *Assuming the higher cost estimate for the REBCO tape, the materials costs for ARC total $428M and the total fabricated component cost estimates total $5.56B.

Given it's a novel nuclear project - it's probably safe to double/triple that cost estimate. Still wildly uncompetitive. That's actually really disappointing.

Their cost estimation methodology in that paper seems absurdly simplistic, though.

The fabricated cost estimates were extrapolated from other novel nuclear fusion projects, so doubling it would be double-counting.

> Compare ITER (0.05 MW/m^3) or ARC (0.5 MW/m^3) vs. a commercial PWR fission reactor primary reactor vessel (20 MW/m^3). Stronger magnets don't save a reactor from this.

Stronger magnets absolutely would save the reactor from this, because for a fusion reactor of any given size the power density goes to at least the fourth power of the magnetic field. If we could have ITER, but with 40T in the on-axis toroidal field, it would have a power density of ~150 MW/m^3. This is of course ludicrous because the thermal output of 1.5TW would not be possible to contain, not even considering how the coils are supposed to be kept together. But this clearly show the path that also allows beating the square-cube scaling: You scale magnetic field up, and reactor size down, until you have reasonable power outputs from reasonably priced reactors.

There is potential future in fusion, but it specifically requires better superconducting magnets. ARC at least is research in the right direction of high-field superconductors, even if the REBCO magnets they are using are not quite up to the task of making a reactor that would be economically viable.

> Stronger magnets absolutely would save the reactor from this, because for a fusion reactor of any given size the power density goes to at least the fourth power of the magnetic field.

... right up until you reach the limit on what your first wall can handle. At that point, making the magnetic field stronger gains you little or nothing, because your reactor does not survive operating at the fusion power density it would enable. The claim I am making is that this power/area limit causes DT fusion to be inferior in volumetric power density to fission. The factor of inferiority is roughly (minor radius of fusion reactor)/(radius of fission reactor fuel rod), and the numerator there at least about a meter due to the need to stop neutrons. The factor is independent of magnetic field.

What strong magnetic fields might do is allow you to go to advanced fuels. The hope there is avoiding a thermal power cycle entirely, saving on the non-nuclear side of the power plant. Tokamaks would be hopeless for this, though; their beta is too low.

I'm also thinking of the highly diffusive radioactive tritium gas which is a PITA to contain and work with.

https://en.wikipedia.org/wiki/Tritium

They're hoping to use a lithium "blanket" that would absorb and use the emitted neutrons to breed tritium (that's needed for D+T fusion anyway).

That doesn't save the confinement walls from becoming more and more brittle, it just limits the damage to those areas, and it helps recover some of the tritium it leaks.

It's not that they hope to breed tritium; tritium breeding is absolutely required for DT fusion to make any sense at all. Making the tritium in fission reactors for a 1 GW(e) DT fusion reactor would cost $20B/year.

For anyone interested in small fusion reactors there is another startup called Tokamak Energy with great Youtube channel: https://youtube.com/channel/UCuSlFJbBUIj1zfJLRnGXSow

For anyone interested in even smaller reactors, I'm working on a device that believe might be able to scale down to the size of a button cell battery.

http://www.ddprofusion.com

Do you have any published papers on this? I’m fascinated!

They link a patent on their page: http://www.ddprofusion.com/US10354761.pdf

I haven't yet found anyone to help me to describe and analyze it with rigor.

Some related discussion by fusor practitioners about the concept. It's a tantalizingly complex field where many have attempted and failed to find the pot of gold at the end of the rainbow: https://fusor.net/board/viewtopic.php?p=78286#p78286

Yeah, that's me there as well. I've been working on this for a few years.

Is this somehow related to cold fusion?

Not at all. I believe a Penning trap could have an oscillation mode which causes a single species to reach extremely high densities for short periods of time.

I'm hoping to try for a density record with my prototype. It may not happen, but I think I can get higher densities than NIF.

There is a list of active fusion organizations at :

https://www.fusionenergybase.com/organizations

Also the fusion subreddit is reasonably active :

https://old.reddit.com/r/fusion/

For anyone after an up to date book on fusion The Future of Fusion Energy is good:

https://www.goodreads.com/book/show/43700662-the-future-of-f...

I can't say anything about that website since I am not familiar.

The subreddit is godawful and no one should waste their time.

The book is top notch and everyone interested should give it a read (assuming you care enough and have enough disposable income to afford it, it is not cheap).

Where do you get fusion news?

There is news on that subreddit of all the fusion startups, even the lesser known ones like HB11.

The Kindle version of The Future of Fusion Energy is $US 12.

Well, I work on a fusion research machine in an academic setting, so weekly seminars. I hear about startups from coworkers, but I’m not too interested in them because I am jaded about their prospects. I understand that they are considered moonshots by VC, but I still think their chance of success is 0%. I’d love to be proven wrong, but a lot of smart people have been working on this for a long time. I don’t think we’ll be shaking out any magical revelations anytime soon. Advancements in turbulent transport is where the important work lies imo.

Anything funded by DoE (which is a lot) necessarily has to keep all publications public. Just because you don’t have a subscription to Nuclear Fusion or IEEE Transactions on Plasma doesn’t mean you can’t learn about what’s going on for free. It’s unfortunately just a pain to find. Look for papers hosted on websites of projects at PPPL, ORNL, UWisc, IPP, etc.

What about degradation of superconducting materials under high neutron flux which will be generated by a "commercial" load? IIRC modern superconducting materials rely on relatively fragile meta-structures, which can be easily damaged by a sufficiently strong radiation. Changing magnets one-two times per year does not sound good for economic viability of such reactors.

Neutron flux at the coils is one of the primary things looked at in reactor studies. It is a factor in choosing the thickness of the lithium blanket and boron coating.

The lithium beryllium molten salt they’ll use as a coolant and useful heat extractor should be enough to capture free neutrons.

It’s the same mixture that is used in molten salt fission reactors so it’s neutron absorption profile of it is well understood.

As a bonus you should be able to extract tritium from the molten salt which means it will produce some of its fuel too so it’s a partial breeder reactor too.

I think if I understand correctly, there'll be a fluid (FLiBe) between the plasma and the magnet coils that absorbs most of the neutrons. The FLiBe heats up and is used to boil water to power a steam turbine.

SPARC isn't particularly designed for durability, but for the ARC reactor which is meant to be the commercially-useful iteration they're looking at having solder joints on the superconducting magnet film so the whole top of the reactor can be removed so they can pull out the inner lining in one piece and replace it. (Apparently they figured out that regular non-conducting solder joints don't actually introduce very much resistance.) I don't think there's any plan to replace the ribbon.

regular non-superconducting solder joints. They conduct, just not nearly as good as the material itself at that temperature.

Still, interesting that that would not result in enough power to boil away the solder.

Yeah, that's what I meant. Thanks. I guess if the solder joints are super-thin, there isn't much resistance. But still, it defies intuition.

One of the other weird things the MIT group working on SPARC has been experimenting with that sounds like it totally wouldn't work but apparently it does is that they don't bother to insulate between the Rebco tape windings. The superconductor is just a thin layer on top of stainless steel, and the stainless steel is a sufficiently mediocre conductor that the vast bulk of the power takes the long way around following the superconductive layer rather than taking a shortcut through the stainless steel. Apparently the insulator is less durable than the tape itself, so not having to rely on it makes for a more durable device.

Huh. I would have assumed kapton tape wouldn't have been too hard to add, but if you can do just fine without that, it's even fewer manufacturing steps.

The best part is no part!

>IIRC modern superconducting materials rely on relatively fragile meta-structures

The HTS 'tape' they use is very robust. A lot of the work they are doing is qualifying the coils and magnets under various scenarios.

So their main advancement is stronger magnets with "rare-earth barium copper oxide (ReBCO) on metal tape" being wound into extremely tight loops (and is presumably extremely thin).

Will we enter into a tipping point of materials science that allows magnets strong enough and suddenly we get fusion and it becomes ever better as we make better superconducting magnets?

Quite possibly. The magnet as a key tool for harnessing the immense power of nuclear reactions is only just now coming to light, most recently when Dr. Indiana Jones survived a nuclear blast by hiding in a fridge covered by a variety of small magnets.

Fusion seems like it will advance the fastest through (a) materials science unlocking better magnets and (b) simulation and physical experimentation with reactor designs, including highly experimental designs.

Computational modeling seems to be helping as well:

https://ai.googleblog.com/2017/07/so-there-i-was-firing-mega...

Disclaimer: I am not a plasma physicist.

I'm not sure how much potential improvement you can expect there in the near future. There's a pretty large jump between the conventional superconductors and the new high-temperature superconductors used here, I'd suspect that it'll take a long time of incremental improvement to build stronger magnets with these new materials.

There's a lot of practical problems with building very high-field superconducting magnets. I'm also not sure how much you can gain from the thinnness of the material, conventional superconducting magnets have a lot of non-superconducting material in there as well to conduct heat so that the magnet isn't immediately destroyed on a quench.

"Suddenly", no. They're targeting 2025 for the first tests if everything goes perfectly, 2027 for a first demo reactor, and we don't know how expensive a real reactor would be so we have no idea if this is the design that will one day unlock UNLIMITED POWER!

but on the scale of civilizations, yeah this could be it.

4 years in the fusion field is pretty "suddenly".

Why is that fusion reminds me of the James Webb space telescope .... https://xkcd.com/2014/

Fusion at this point is closer to general intelligence AI than to JWST.

JWST is a project level failure not a “field level” failure.

Fusion has been 10 years off for the past nearly 100 years.

Tokamaks date to the 50’s and the first patents for a fusion reactor were issued in the 1940’s https://worldwide.espacenet.com/publicationDetails/biblio?CC...

Uh, no. We are fairly sure that we can build a working fusion plant and we know with good accuracy what remaining engineering problems are. We have literally no idea what the path to AI even looks like.

Energy positive fusion? I’m not so sure that we do...

We already have working AI, not prototypes. Look at big corporations for example.

@dogma1138 said general AI, what we have in big corporations is specialist AI.

A general AI must, by definition and at minimum, be capable of doing any intellectual task currently done by humans; right now, they’re good at what they do, but are very limited in what they do, and slow to learn. For example, self driving cars can still only do limited environments and conditions, despite Tesla having over 18,000 human-professional-equivalent years of driving experience as of April last year, which they achieved shortly after they added “Traffic Light and Stop Sign Control” to their feature list — and that item is still listed as “(Beta)”.

It’s the same everywhere: GPT-3 is fantastic… but despite having “read” more than any human could in a lifetime, it still struggles with moderate arithmetic; Voice assistants do amazing things… but I’ve literally had easier times attempting to converse with pet dogs; Google Translate has made moving and travelling abroad much less stressful, and it knows more languages than I can name to a higher standard than I know a second language… but I’ve seen it hallucinate words in lawns, and it does make translation mistakes that even I can spot.

And even if they were perfected within their domains, none of these are generalists for all domains.

Iter and other big or old projects are going the safe route with magnets that need to be at 4 or so Kelvin. All these ‘hotter’ superconductors are relatively unproven (it’s a oversimplification for a broad category), especially their strength when made into the shapes that make the magnetic field and form part of the load bearing structure of the container. The winners would enter that tipping point.

It is worth pointing out that Iter started in 1985, I think before REBCO was discovered (and it was a little bit after that they acheived superconductivity at 77K, liquid nitrogen temp). Sturdy and flexible tape form from ion beam-assisted deposition I think was later but I can't find the date.

(edit: maybe in the lab in 1992? I'm not sure when it scaled production: https://en.wikipedia.org/wiki/Superconducting_wire#cite_note... )

Practical use of the new high-temperature superconductors in high field magnets is even more recent. NMR is probably the closest commercial application for this kind of magnet, and the first spectrometers using the new superconductors were sold last year.

The sandwich structure described in the article makes sense: separate the superconducting and the load-bearing functions into different layers.

the magnet singularity is coming

> Will we enter into a tipping point of materials science that allows magnets strong enough and suddenly we get fusion and it becomes ever better as we make better superconducting magnets?

I think that's the idea. Iter is about as small as it could possibly be and still work given the magnet field strength they had designed around. With stronger magnets, we can make smaller reactors, which are cheaper to make and (if I understand correctly) have better power density. At some point it stops being practical to make it any smaller as the limits become "how thin can we make this shielding material?" or "how much heat energy can we remove by pumping fluids around?" And then once we've proven the concept and we've settled into an optimal size the engineering focus turns to "how cheaply can we manufacture this?" and "how can we reduce the total operating cost per megawatt hour?".

That's pretty much what's happening here, fusion output increases with the fourth power of the magnetic field. Double the field, 16X the output. But if we take it much further, we'll reach a point where the limit is the structural strength of the reactor.

A few years ago I got to tour MIT's Alcator C-Mod, which had the most powerful field of any tokamak to date. A grad student showed us a metal tie rod, about a meter long, and said they'd calculated that two of them could hold down the Space Shuttle while it was trying to launch. To hold the reactor together while it was operating took 38 of those.

The peak on-coil magnetic field in the ARC design (from the 2014 paper) is 23 Tesla. The pressure of a 23 T magnetic field is 2100 bar, nearly twice the pressure at the bottom of the Marianas trench, and comparable to the chamber pressure of a handgun. 60% of the mass of their design was the stainless steel structure needed to support the magnets against the outward JxB forces.

Fusion's prospects would likely be helped some by plasma configurations with much higher beta than tokamaks. At least the experiments would be cheaper.

I remember being in high school and visiting the large magnet facility in Tallahassee at FSU (you know, where they levitate frogs). They explained to us that the superconducting coils were made of niobium-titanium alloy, and I remember asking: "aren't there better superconductors?". The answer was yes, there are, but they're insanely difficult to make or something, so we don't consider it practical, "maybe someday".

It looks like "someday" finally got here -- the cuprates are being used in practice.

Are there any startups working on the material science side of making these magnets? It seems like a very tight engineering team might do well here.

In one of the SPARC talk videos I think someone asked who makes the Rebco film, and the speaker said there were a handful of small companies around the world. I don't remember if he was any more specific than that, but anyways it seems it's kind of a novelty low-volume product at this point. I assume with demand picking up that manufacturing capacity is going to grow along with it.

Low temperature superconductors in general is, of course, an active area of research. There may be better alternatives to Rebco just waiting to be discovered.

Fusion is one of those topics where I'm always going to click on the comments before I click through to the article. There's a lot more to learn from you than from the link. Thanks for being here HN.

How does a software engineer with a cs/ml background get into cool and impactful industries like energy?

Same question here. I have been a bit disappointed in the portability of my similar background to problems like energy, climate, and biochem.

Therein lies the problem with software. It’s common enough now that the lab scientists can write their own code. They may need a few dedicated programmers but most of the work is not writing the code. So instead we all keep optimizing news feeds.

... I don’t buy this.

Software launches and lands rockets. You think JPL just had a couple of rocket scientists writing some python scripts on the side for the perseverance mission?

But I agree Facebook has a ridiculous amount of engineering potential wasted on a pretty useless problem (serving ads even better!)

It's easy to get an entry level job as a programmer that pays very well, and you will be taught how to be an experienced programmer on the job.

Once you are an experienced programmer in a software company, you are earning a lot, and moving out of the software industry, where the prima-donna employees are physicists or engineers, you generally take a pay cut. A fusion company isn't going to hire an entry level programmer who hasn't proven himself.

So, to answer the question more explicitly, you need to be willing to follow your interests and not maximize the bottom line. This is my 28th year as a software engineer, and I've seen this pattern countless times. I've done the follow my interests, and also follow the money jobs, and prefer the respective good aspect of each approach over the other.

In my experience, there is quite a bit of untapped potential for applying engineering/cs/ml lessons in less traditional domains - not despite, but precisely because the other scientific disciplines are embracing and leveraging computation more.

I'm happy to elaborate on a more private channel, email/twitter is in the profile.

Source: Cofounded a startup in comp.bio space ~3.5 years ago, been busy supercharging our scientists and increasing pace of innovation and haven't really ran out of ideas yet.

There is plenty of software that is needed or could use improvement. Anything from plasma simulation software to embedded firmware used in the solar industry.

Improving Julia language and ecosystem would count since the language is common in energy-related HPC tasks[1] and they even got grants from DoE. See also this discussion [2] specifically on fusion research with Julia.

[1] https://juliacomputing.com/industries/energy/

[2] https://discourse.julialang.org/t/julia-in-fusion-research/2...

thanks everyone for the insightful answers! as a follow up is there anything I can do now to prep myself for working in space/energy? like should I study physics/chemistry on the side?

For context on fusion reactors I recommend edx's plasma physics and fusion course. From that it seemed fusion reactors that give a net power surplus, maybe, are just not possible in such confined envs on earth. 80 years of incredibly smart people with billions chasing the problem, with the first Tokamak reactor going back the 60s. Even ITER, the world's most expensive scientific endeavor ever, is still a question mark if it will work.

Would love to be convinced otherwise to be hopeful of confined fusion reactions on Earth.

The parameters that go into the efficiency/gain of a power plant are well known: gain goes up as radius to the power of 1.3, and magnetic field strength to the fourth power.

ITER was conceived in an era of much lower-field superconducting magnets, so it had to increase the size instead. This massive size has been the big cost and schedule driver.

However, since ITER was designed there have been big advances in the production of high-magnetic-field superconductors. These are really recent - ReBCO tapes have only started to be sold by commercial producers in the last year or two. With higher fields, we can get performance equal to or better than ITER at much smaller size (and hence price). This specific effort is an MIT project, relying heavily on MIT research in building magnets with the new superconductors.

I highly recommend this video for a look at the different scaling factors: https://youtu.be/h8uYNhevRtk?t=571

Fusion is N years in the future and it will always be so. N=30.

With the news from Commonwealth Fusion Systems, N is still 30. And it will keep being 30 for a while.

Why? People are averse to radioactivity. Fusion produces a huge amount of neutrons, and those neutrons transmute the nuclei they encounter and result in radioactive waste. If you decide you can put up with radioactive waste, you might as well use fission.

For other disadvantages of fusion, you can check [1].

Where fusion can shine is in fusion-fission hybrid reactors [2]. This can not only be used for energy generation, but also to dispose of the accumulated nuclear waste.

[1] https://thebulletin.org/2017/04/fusion-reactors-not-what-the...

[2] https://en.wikipedia.org/wiki/Nuclear_fusion%E2%80%93fission...

This is true, but is not the reason.

The reason is that there has never been, at any stage, any serious intention of building practical Tokamak power generation systems.

Power density is too small, plant lifetime is too short, scale is too large, reliability is too low, recovery from failures too costly: each by one or more orders of magnitude, and multiplied, while renewables costs are still in free fall. Even fission, overwhelmingly cheaper, is not competitive.

Instead, Tokamak research has always and will always be a jobs program for hot-neutron physicists, intended solely as a way to maintain a pool of candidates for weapons work.

Other, aneutronic fusion methods could possibly be practical someday for power generation or for propulsion, but are actively neglected because they are useless for the true purpose of the program. If your stable of physicists stop thinking about hot neutrons, and find job prospects not involving skills useful mainly for weapons work, they are lost.

Money spent on Tokamak work could be spent on other methods that have any actual prospect of success. Production capacity of high-temperature superconducting tape could be used to cycle solar and wind power between geographic regions without losses. But they are diverted.

Huge Tokamak construction projects have another organizational purpose: corruption. Any sufficiently large public-works project offers practically unlimited opportunities to siphon off monies to profit individuals not engaged in any societally useful capacity.

This is why fission project costs balloon from, e.g., projected $2B to $10B+: not from regulation, not from change orders; it is because ballooning cost is the whole point of the project. Any finally useful outcome is mainly a carrot to keep the money flowing.

ITER is such a project. It does not even pretend to "work" until 2050. It does not even pretend to any prospect of useful power output, ever. Yet, its construction consumes many $Bs every year. Each $B spent makes it harder to abandon, justifying endless overruns.

Most projects must ultimately deliver something, even if vastly overpriced: a jet, a tunnel, a power plant; or finally be cancelled and written off.

Fusion work has successfully resisted any demand for delivery, over many decades, and openly admits plans on not delivering for many more. It is perfectly corrupt. Note, not a penny evaporates: it all lines open pockets. I haven't heard of millionaire physicists (pace Myhrvold), so somebody else gets it.

There are nuggets of truth in here but they are combined in a way that makes no sense in net.

Eg. it's true that some fusion programs are rooted in nuclear weapons work, but that cannot be easily said of ITER, because if that were the goal we'd have FIRE instead, the Fusion Ignition Research Experiment that used actively cooled conductors and higher chamber pressures. (And then also a bunch more since FIRE would be built.) ITER's advantages don't make sense if the goal isn't actually sustained fusion.

Yes, ITER is slow, over budget, and had delays, as is typical of project of its scope, but this doesn't mean it's a fake project any more than James Webb or SLS. Not delivering power doesn't mean it's fake science any more than the LHC. Certainly its jobs program aspect doesn't transfer to new fusion startups, which are in large part privately funded.

ITER is, as you say, slow, over budget, and had delays, and never promises ever to deliver one solitary erg of energy, commercially viable or otherwise; yet it is planned to consume all fusion research money well into the 2050s, and with delays, beyond. FIRE would have failed more spectacularly, and then the gravy train might have dried up.

To say a fusion program is "rooted in nuclear weapons work" suggests it is meant to produce results of interest for weapons, but that is wholly unnecessary. The connection to weapons work is more tenuous. An ITER physicist has no job prospects besides either ITER or weapons work. Without a stable of ITER physicists, who would be available to hire for weapons work, whenever it pops up?

I think James-Webb has a telescope built and waiting to be scheduled for launch, so there's that.

SLS certainly is a fake project. I will go on record, here, forecasting that SLS will never launch one solitary astronaut. It never had a purpose to exist other than to maintain employment in a sufficiently wide variety of Congressional districts. The planned "lunar station" had no reason to exist except that it was the farthest SLS could get to. Once Starship is flying, in a year or three, even the US Congress will find it difficult to continue funding SLS.

Most fusion startups, like 90+% of startups, are about diverting VC money into ready pockets with no expectation of return. Principals of startups are mostly unaware of their role in this dance.

Your argument seems self-refuting, tantamount to claiming that politicians know more about fusion physics than fusion physicists do, and through illegitimate means had convinced the body of scientists that ITER and FIRE were physically sound. (For sure, neither approach is directly an economically viable path to fusion, but that is not the intent.)

Saying that SLS will be too late to the market—certainly a reasonable stance, and I'd happily have it cancelled—is not the same as saying it is fake. Similarly, if James Webb is real because it is almost ready for launch, then it is possible for something to be real in spite of being years overdue and multiples over budget.

No detail of SLS was dictated by any plausible mission. It had to have strap-on multi-part solid boosters solely to keep Morton Thiokol busy. It had to have SSME-style engines to keep somebody else busy; but not use actual SSMEs, for the same reason. They had to be super-expensive engines designed for multiple uses even though they would have been scuttled after a single use. Things can be "late to market" without being fake, but being late to market is an operational requirement on fake things such as SLS.

As noted, military-industrial concerns demand funding go exclusively to hot-neutron schemes. The politicians' interest is only that billions of dollars be disbursed to obscure entities over a long period that can generate sufficient regular, reliable, legitimate-appearing kickbacks.

"Commonwealth Fusion Systems announced today that later this year it will start to build its first test reactor, dubbed SPARC, in a new facility in Devens, Massachusetts, not far from its current base in Cambridge. The company says the reactor, which would be the first in the world to produce more energy than is needed to run the reaction, could fire up as soon as 2025."

The key takeaway, for those skimming.

I wish this SPARC better luck than the CPUs.

I wonder if the facility will be called a SPARCstation.

Of course, they were from Sun - the sun does fusion.

haha this gave me a chuckle. underrated comment.

You mean to say that fusion is only 5 years away? Amazing!

A five time improvement over the previous estimates

Six.

ITER is aiming for Q>1 in 2035 IIUC, so 3x.

Wasn't ITER supposed to be built by now? DEMO was supposed to start in 2033 as the successor to ITER.

It depends on which baseline you're referring to and how many more times the US would back out of its financial commitment after that baseline was made.

ITER, like many oversized government programs where funding and structure are almost entirely political, has had delays. First plasma (not full power) was originally scheduled for 2016, and is now 2025.

See also: The SLS program (the "Senate Launch System").

If you try to coordinate design and manufacturing among 35 participating states, delays tend to happen.

Perhaps also noteworthy from a historical perspective is that the whole thing was proposed by the Soviet Union, which doesn't even exist anymore. The US also pulled out of the collaboration in '98, necessitating a redesign; they rejoined in 2003, but congress periodically tries to pull the plug...

I was more sarcastically referring to the - I thought - common saying that fusion will always be 25 years into the future.

The test plant is 5 years away, but yeah, net-gain in energy is the target.

5 years away and always will be is a 4x improvement over the previous 20 years away and always will be!

Fusion research has been steadily progressing for many years. The "50 years away and always will" quip is baloney.

So how far away would you say we are now? Because I remember "10 years away" being said in earnest since back in the early 90s.

I don't know, and I don't think anyone does. It depends on lots of things including funding and what other unknown unknowns there are on the tech/eng tree.

What I do know is that superconducting magnets have been getting cheaper, more powerful, and more compact for quite some time and that's one of the limiting factors on fusion reactor designs. I also know that there is no known physical barrier to net-positive fusion, only engineering barriers.

If I had to totally guess I'd say someone will show net-positive fusion for a brief period of time before 2030... assuming the funding is present. It will not be a fully viable power plant yet but a proof of concept. This will be followed by a huge bump in funding and a race to produce power plants.

... but that's a guess.

It's like asking "when will there be a human base on Mars?" We know it's possible and I think it will happen, but I don't know how long it will actually take. We could probably have one in 5 years if someone wanted to write a blank check.

I would bet that we could have a fusion PoC in 5 years if someone wrote a blank check and fully funded many different credible efforts.

I'm pretty confident it will make it over the line into viability eventually, but it's a really tough problem and it is going to take decades. There's still a lot about the engineering required we just don't fully understand, and it will take a lot of tests and investigation on a lot of different fronts to make meaningful progress. So I agree, we're in it for the long haul on this one.

Why are you confident it will be viable? Remember, commercial viability means not just that it works, but that it's better than any competitor. Why is it that fusion will win this competition? It's the nature of most technologies to fail in the face of something better.

I was reading somewhere that the earliest point for commercial viability will be in the 2060s. By which point we can assume there have been build a lot of renewable energy infrastructure (or else the climate crisis is still in full effect; and I can’t really think about what kind of society will have evolved from that disaster). This should bring down the cost of renewables even more then today as economics of scales grow exponentially. Making the quest for comersial viability of fusion even tougher.

I honestly wonder if fusion will ever be commercially viable. And if all these experiments will simply lead up to us realizing that: “cool, so fusion power is possible on Earth. Now what should we do with it?”

But hey. Maybe it will be the energy of the future on Antarctica or the Moon or something instead.

It just a best estimate, there’s no way to be sure at this stage. Plenty of smart people think it will and plenty think that it won’t. I think there are so many different promising alternative approaches that hope to solve this problem, that on balance I’m not prepared to bet against every single one of them not panning out. It’s just a guess though.

I agree renewables might make it mostly moot, but there’s still the issue of base load. Storage might solve it, but if fusion can be made safe and cheap enough it could still have a niche.

Then there’s Mars and the belt. There’s no guarantee we’ll colonise it permanently, but if we do fusion could be really useful as it’s far enough out that solar is significantly less efficient. Also fusion reactors could be handy power plants for spacecraft. Cheap reusable heavy lift systems may make all of this feasible.

Finally, the future is a long, long time. If we don’t wipe ourselves out, eventually we will every technology that is viable will be achieved (not possible, viable).

Fusion is a complicated thing. Even if we reach net positive energy generation before 2030. How much longer will mature power plant design take? Lets say something lasting at least 10 if not 20 years with reasonable maintenance(that is not replacing the whole thing every other year). Not to forget that nuclear plants can run 60 years.

And then comes questions of life-cycle energy inputs and costs. How long will it be to be net positive on these? That is we spend less energy on cooling the coolant for superconductors and overall building the thing.

There have been tokamak and stellarator fusion power plant studies published every few years for the past several decades. Answers to all of your questions and more are in them.

Don't keep us guessing, what is the date these studies came up with?

When we decide to fund it.

> I would bet that we could have a fusion PoC in 5 years if someone wrote a blank check and fully funded many different credible efforts.

If that's the case, then why hasn't anyone written that check? The potential profit from commercialized fusion seems enormous (unlike a mars base), and there doesn't seem to be a shortage of capital seeking large returns.

Few people have the ability to write the check.

Musk is the only one that even comes to mind, and he’s on the record saying he won’t touch turbulence.

Still, there is hope.

I guess it would be a really really big blank check, and then you'd still have to get from a PoC to a commercially viable plant and convince the public it's not a hydrogen bomb and it's quite safe and then you'd have to outcompete renewables for decades to get a decent return while you miniaturize your device for use anywhere else than a stationary plant, and then someone starts making a paint with tiny organic solar cells in it and absurdly-efficient batteries and while you can still make a killing for niche applications, it's a pittance compared to your investment. I guess in the end both risks and upfront costs are just way too high to do things that way.

Depends on the funding. The standard response to the 20 years away for 50 years quip is https://commons.wikimedia.org/wiki/File:U.S._historical_fusi...

Why do you think the funding has been so low?

The investment prior to any payoff is significant. Few countries, or even groups of countries, have had the excess capital and idealistic interest to solve the problem.

Raegan gutted virtually every US program and refunding has only come back to things that are politically relevant. “That giant science machine that might one day make lots of heat” isn’t high on the “political value” list.

There was a lot of juice back in the day for research priorities around aviation, nuclear weapons, etc but those were all things that were militarily significant.

Eternal energy security is not militarily significant?

Not really. Energy scarcity is a very useful geopolitical tool when you're the one with the tap on the oil, the weapons, and the money.

Who knows. If I had to make a guess, here are some ideas:

You cannot yet make money with it, so there's no capitalist lobby. Success is not certain, and timeframes are too long for politicians to score points in the election game. It's not as cool as space, and the green faction isn't too keen on the whole nuclear thing. So far, fears about peak oil turned out to be largely unfounded (but do note that we probably did pass the peak as far as conventional oil production in concerned).

If I kept at it, I probably could come up with more theories...

I don’t think the green faction disagrees with fusion because “nuclear = bad”. I think the fact that current estimates point to the 2060s as the earliest point in time for commercial viability. By that point we will have had to have moved on from fossil fuel (or at least gained net 0 carbon emission) for at least 10–20 years. So nuclear fusion is not a solution to the climate crisis.

In other words the green faction is simply disinterested in nuclear fusion, like we are disinterested in the Large Hadron Collider, sure its a cool experiment, but nothing we should be considering to further our goal of fighting our current environmental disasters.

From personal experience, "nuclear = bad" is a thing.

That said, I agree that as things stand today, fusion research is no panacea to climate change. However, note that the United Nations Framework Convention on Climate Change was ratified in '92, and if we'd decided to go all-in on fusion back then, who knows where we'd be at today...

Just because some people believe blindly that “nuclear = bad”, that doesn’t mean they won’t change their opinion of fusion once they receive an adequate explanation on how fusion is different from fission. I’ve personally never met anybody in the wild that disagrees with fusion power because of our current track record with fission power. I.e. the believe that “nuclear = bad” does not equate to the believe that “fusion = bad”.

So honestly I don’t believe that there exists people in the wild who’s opinion is: “No to fusion! Because nuclear = bad”, and if they do exists, I don’t think they are of anywhere near size and numbers required to influence public funding.

> Fusion research has been steadily progressing for many years. The “50 years away and always will” quip is baloney.

The upthread comment was 20, not 50, and I’ve heard 15 or 20 years away frequently since the 1980s and seen it in things dating back to the 1960s, so, no, its not baloney.

Nor does it necessarily mean that progress isn’t being made, its more of a comment that the unknown unknowns are being converted in known unknowns as fast as a known unknowns are being converted into known knowns.

Just got a new Mac. As I read your comment, my Migration Assistant is doing similar estimates. “13 minutes remaining” then “41 hours remaining” then back again.

It’s been whipsawing between these estimates for the last couple hours.

I think I now understand the state of fusion research.

Think of it this way: how hard is it to estimate the time required for a software project? How successful are most such estimates?

We don't know how long it will take, but at the same time there is visible progress occurring on many fronts: understanding plasma behavior and how to control it, better superconducting magnets, better control systems, solutions to the neutron embrittlement problem, etc.

There's not going to be a single breakthrough... or rather the fusion breakthrough already occurred. We've already learned about fusion and how to trigger it. That happened in the early 20th century.

Instead of a single breakthrough it will be continued progress on all fronts until at some point everything matures enough that someone manages to build a proof of concept reactor. At that point investment will flood into the space because it will have been sufficiently de-risked.

fusion is the energy of the future and has been so for the last 50 years

Just goes to show that something witty is good at convincing a ten year old that something false is true.

I think it will happen. I mean, Duke Nukem’ Forever actually did eventually get released. Anything is possible!

Hopefully not like an asymptotic curve.

5 minutes away and always will be.

People love to joke about it but I remember 20 years ago, the joke was about it being 40 years away.

In SimCity 2000 it didn’t show up until around 2050. Game was from 1993, so that was a 57 year prediction.

Given we know how long it has been 40 years away, 20 years away, and now 5 years away, we can estimate how long it will be until it is 0 years away. Unless it goes down by a power of 2 each time.

Zeno's fusion reactor

Well, 40 years ago it was 30 years away, so for a while it must have been going backwards.

The "only X years away" jokes are predictable, but we'll know soon enough if they can do 20 Tesla - and if they can, then the time tables should be reset.

Technically speaking we already have « fusion ».

The problem is we don’t have « stable fusion » this one will take another 5-6 years.

Then we need « positive yielding fusion » today fusion has negative yield... that’s another 5 -10 years at least.

Finally we need «commercial scale fusion reactor » like France did with their Nuclear Reactor massive investment from post war to today in order to make cost and delay acceptable.

That would be 2040 at least for industrial nuclear fusion.

I have no idea what humanity will look like in that timeframe.

TFA is about stable, positive yield fusion in 5 years.

The key difference here is that it's a venture backed effort, which signals something very different from large state-financed research efforts like Iter.

Namely, that there is a path to financial viability

MIT’s product isn’t the reactors: it’s HTS coil winding. That’s the proprietary tech they are developing and planning to sell.

VC backed and

> Namely, that there is a path to financial viability

I see what you did there.

IF (if) everything goes to plan than 5 years is quite fundamental considering the amount of implications about safer cheaper energy.

"produce more energy than is needed to run the reaction" is a great milestone, but "economically competitive method to produce electricity" is the real goal and requires a good bit of work beyond that

But even without that it could be immensely useful for space travel and colonization. And also for remote environments.

at this rate we might get fusion before James Webb space telescope becomes fully functional!

ref: https://xkcd.com/2014/

A Micro Fusion system SPARC which is similarly named to the CPU from Sun Micro-System.

Worth noting, though it's not in the article, that they're aiming for a commerical version (ARC) in 2030.

"fire up" != energy positive and ripe for production

How many fusion startups have made neutrons? (I know this is a low bar... Fusors can do it pretty easily, but I still want to know.)

A tokamak can only constrain a plasma in 2 dimensions. To get confinement to last longer the magnetic field concentration needs to increase. An entirely new way to constrain plasma to 3 dimensions is needed for practical fusion applications. Otherwise accidents will be disastrous for plant operators.

> Otherwise accidents will be disastrous for plant operators

What's going to go wrong? I've seen talks that breaches will be rapidly cooling and will be contained by a modest amount of concrete.

Plasma disruptions

https://www.jp-petit.org/NUCLEAIRE/ITER/ITER_fusion_non_cont...

Field quenching of the super conducting coils. The stronger the field the worse it can be when it collapses.

Quenching is less of a concern with HTS coils. They are operated with liquid helium for higher critical current. The HTS is embedded in copper.

One argument going on right now is insulated vs uninsulated coils. Insulated are the norm, but uninsulated are virtually impossible to damage from quenching. The copper is a virtual open when the core has zero resistance. When the core quenches then suddenly the windings are all parallel and the coil turns into a single turn copper coil that immediately dumps the current.

Tokamaks are in the limelight, but the older brother hanging out in the shadow will take the stage in our lifetime :)

Keep an ear out for new stellarator projects (assuming an increase in research funding).

What's your opinion on stellerators? From what I heard about Wendelstein 7-X they were very happy with the results achieved in the experimental phases so far and are currently upgrading it to achieve longer runs in the future.

Disclaimer: I work on a stellarator. Specifically, the first MHD optimized stellarator and the first with 3D modular coils. W7-X is the only other one in the world. I feel that the current state of stellarator theory is encouraging. Coil optimization has integrated lots of important factors into their cost functions, including tolerance. Flux surface geometry optimization has integrated theoretical turbulence models into the cost functions. A pure MHD optimized machine will not be a fusion reactor, but we don't know enough about turbulent transport to build a working machine today. We need mid-scale machines to develop the theory. Tokamaks get the vast majority of the funding because their performance metrics are higher, but imo stellarators are capable of matching tokamaks in performance with much less control issues.

> To get confinement to last longer the magnetic field concentration needs to increase.

Which is what the new project will do - using a classic tokamak with the much higher field magnets that can be built with high-field superconductors.

How is Lockheed Marting doing with its shipping-container-sized fusion reactor?

It grew in volume by a factor of 100, its power density fell to about 1/40th that of a fission reactor, and they stopped talking about it.

Even so, if they could make it work it would be quite amazing. After all, miniaturization is a different challenge from getting the thing to work in the first place.

It would be amazing in the same sense a dancing bear is amazing. Not because it's useful, but because it can be done at all.

Minaturization runs up against limits on power/area through the wall (and minimum thickness of T breeding blankets) that will force any DT fusion reactor to have power density a small fraction of a fission reactor.

I knew it was doomed from day one. They never could show that the reactor support parts weren't in line with the flow of the plasma. It never got much past the Polywell concept from which it was derived.

Yet, it was a big success in that substantial cash was diverted to places where it would not otherwise have been. Not a penny of it evaporated; all went into waiting pockets.

I still don't understand why people want fusion.

Like fisson already does basically what you need and is easier in every way.

Yes, the energy density of fusion is higher but the energy density of fission is already so absurdly high compared to chemical.

There are only a small number of cases where I can think of this making sense, and even then it would likely not be worth it.

The problem with nuclear power is the lab to operations process, regulation and engineering cost. Fission will likely not improve on either of those compared to fission reactors now being developed.

If we can't can't get a Molten Salt reactor with a CO2 Brayton Cycles turbine into commercial deployment, I have little hope for Fusion.

And if we do, then its hard to see how Fusion reactors beats it on price.

That said, I want fusion for crazy rocket concepts.

Fusion is much less likely to have accidents that poison the local landscape for the next 10,000 years.

As odd as this may sound, this is not an impediment to current nuclear fission deployment. The risks are small, and many populations are very willing to accept them. See France, as well as most places in the US that already have nuclear reactors nearby. Or for that matter, many many sites in the UK.

I am skeptical of this, assuming you need some sort of public support in order to build a nuclear power plant (I'm guessing voters are involved somewhere, even if it's electing their state governor).

Most people are scared of nuclear power, so it seems politically problematic (at least in the United States).

I went to school in Pittsburgh where there are nuclear power plants nearby and people still felt more comfortable with coal being shipped over from Virginia.

Current sites that would welcome new nuclear, as evidenced by enthusiastic local support matched by a new construction project:

* Vogtle, Georgia

* VC Summer, South Carolina

* a cluster of small towns in the west that are the first customers for small modular reactors

* Wylfa, UK

* Hinkley, UK

The greater challenge with nuclear is getting the funding to construct, followed by actual engineering, procurement, and construction.

The funding problem is because of the safety problem. There are high regulatory costs associated with building new (fission) nuclear power plants, because:

* If you don't regulate the materials and fuels used for fission nuclear power plants, most countries could easily build nuclear weapons, and

* If you build fission power plants badly, or maintain them poorly, everyone and everything around them dies in a large radius, and in an even larger radius gets severely sickened. And this radius is poisoned effectively forever.

If there weren't safety problems inherent to fission nuclear reactors, they would be much cheaper to build as well as being much cheaper to operate — and thus easier to fund. That's part of why fusion reactors are interesting: theoretically they should work just as well if not better than fission reactors at converting fuels to energy; the fuel is more prevalent and cheaper; and there should be lower costs associated with building and operating them since the risks are lower.

We just don't know how to build them yet, and figuring that out is expensive.

> The funding problem is because of the safety problem.

I have yet to see convincing evidence of this. It seems like an excuse. Perhaps it's code for "the regulators won't let us get away with screw ups", which is what happened at Flamanville.

The chance of that happening with the fission reactors currently being built is almost vanishingly small.

And even the largest nuclear accidents ever did not lead to anything close to that.

This is just fear mongering nonsense. And btw, even with Fusion you still produce huge amounts of high energy particles that can be just as dangerous and can be used to do bad stuff as well.

Fusion is not magic.

On the other hand, they will almost certainly have tritium leaks far larger than from fission plants. A single 1 GW(e) DT fusion reactor would burn enough tritium each year to contaminate 2 months of flow of the entire Mississippi River above the legal limit for drinking water. Leaking even a small fraction of that could have serious consequences.

The question is why wouldn't you want fusion. More options are always better I think.

I'm not saying I don't want it. It just seems as humanity we totally fucked up. We found a revolutionary new energy source and we totally fucked up the deployment of it.

We should live in a nuclear age already, fission powered space craft, trains, ships, power stations, remote electricity. There is no fundamental reason why fission should not be used an all of those.

Yet we almost don't use it at all, and phasing it out at the same time as we face climate change.

At the same time huge money is spent on Fusion that is much less likely to actually help. With the money spent on ITER you could literally run a matcher competitive competition to build 3-4 new fission reactors and likely multible new powerful turbines.

A molten salt reactor with a brayton turbine would likely be far more revolutionary then whatever ITER can ever be.

In general I just feel like fission is disliked and future has this 'wow the future could be magical', and I'm saying, the present could be magical, we don't need to wait for some magical technology. All that is required is some engineering and a general acceptance that fission is good among politicians, regulators and people.

If some start ups want to work on it, I'm not against it. The point is more that even if this magical technology break-threw happens, deploying it in the real world will run against many of the same problems as fission does.

The problem, in my view, is mostly regulatory capture and this incessant drive to let markets do their thing.

As it stands, fossil fuels should be taxed so high that building, maintaining, and running nuclear power plants is cheap in comparison. Yet we have continued investment into fossil fuels even though we're now fully aware of the damage they're doing. And people say "oh, well we don't have nuclear, because it's so expensive." You know what else is expensive? Entire cities being under 6ft of ocean and having to relocate hundreds of millions of people.

In other words, the known externalities are not imbued in the price, because yay capitalism. I think a little market tampering is warranted when planetary survival is at stake. And obviously, the ramp-up should be gradual, ie, we should have been starting this 20 years ago, when it was also painfully obvious that digging up huge amounts of carbon and burning it is a bad idea. Oops.

Recently I've been seeing people comment that raising a building 6 feet using jackstands OR filling the first story in with cement isn't "actually" that difficult or uncommon 100 years ago.

https://en.wikipedia.org/wiki/Raising_of_Chicago

https://en.wikipedia.org/wiki/Regrading_in_Seattle

https://en.wikipedia.org/wiki/Seattle_Underground

https://www.asce.org/project/galveston-seawall-and-grade-rai...

Feeding them all, that's what I worry about.

I'm having trouble determining the connection between my comment and yours. Can you help me out?

I was responding to your off-hand comment "You know what else is expensive? Entire cities being under 6ft of ocean and having to relocate hundreds of millions of people."

6ft of water seems somewhat manageable, relatively speaking, so long as you're willing to move up a story or go full neo-Venice.

Oh, interesting. I didn't connect those two.

I think my hesitation would not be that 6ft of water is not manageable in the single case, ie, one building. But several thousand of them at the same time? A whole city? Good luck coordinating that in any reasonable amount of time, especially when roads are all flooded.

Abandonment is a much more viable option at some point along that particular path.

I wouldn't want it because it would be far too expensive.

- no radioactive waste

- no nuclear meltdowns / runaway processes

- more abundant fuel (on earth and the rest of the solar system)

- less pre-processing of fuel

- fuel cannot be used to easily make weapons

Should we use fission right up until we have viable fusion? Of course, we should definitely be building more fission reactors. But I can't think of a single reason we'd continue using fission once we get to fusion.

- there is waste, it's just shorter lived than fission waste and lower in quantity. The expectation is that fusion reactors will have to be regulated in almost exactly the same way as fission reactors because they are nuclear sites, with nuclear waste, and proliferation concerns.

- there is lot's of pre processing of the fuel to breed the Tritium in a molten salt blanket that surrounds the reactor and separating from the salt and then feeding it into the chamber

- there is plenty of fission and fusion fuel. Yes, there is more hydrogen around.

- tritium is used in nuclear weapons as a booster, to dramatically lower the amount of necessary fissile material - each fusion reactor is a fast neutron source, which means it can be used to make weapons grade materials. Conveniently, it has a breeding blanket for tritium, in which other fertile fuels can be place to make weapons material: proliferation concerns are a real problem for fusion

- tritium is used in nuclear weapons as a booster

It is, but tritium is not put into bombs. Lithium is.

- proliferation concerns are a real problem for fusion

Unless all fissile materials are banned. It is very easy to check for the existence of fissile materials. If there were no legitimate, safe reasons to have any fissile materials in use on the planet, then a global ban on fissile materials is on the table. A treaty where every nation checks on the other is reasonable. It is hard to build a secret fusion reactor, just as its hard to build a secret uranium centrifuge.

It is, but tritium is not put into bombs. Lithium is.

Both are put into bombs.

The main concern when it comes to tritium supply, regards tritium used for boosting of fission charges. Both applications are crucially important, but fusion boosting appears to require significantly larger quantities of tritium. Tritium and deuterium for boosting are supplied to the weapon from an external reservoir (gas bottle) as part of the arming process of the weapon.

Since about 5.5% of existing tritium decays every year, the tritium assigned to each weapon must be regularly replenished. This is done by removing the weapon’s tritium reservoir and exchanging it with a newly refilled reservoir (5). Figure 1.3 shows what may be such a reservoir.

From Norwegian Defence Research Establishment report "Tritium production":

https://publications.ffi.no/nb/item/asset/dspace:6780/20-013...

Also see this Savannah River Site page about tritium supply for weapons:

https://www.srs.gov/general/programs/dp/index.htm

And for a deeper dive, this fascinating blog post:

"U.S. Tritium Production for the Nuclear Weapons Stockpile – Not Like the Old Days of the Cold War"

https://lynceans.org/all-posts/u-s-tritium-production-for-th...

You are both correct :) Tritium is used for boosting fission bombs, lithium (lithium 6/7 deuterides) as fusion fuel in thermonuclear bombs.

So interesting story about this.

We originally didn't know Lithium-7 would be useful in thermonuclear weapons. It was assumed that it would be inert and that only the Lithium-6 would react with neutrons from the fission primary and breed tritium for the fusion secondary.

Then we tested a bomb [0] and the yield on it was accidentally 2.5x greater than anticipated. So large, in fact, that it is still the largest bomb ever detonated by the USA. It turns out that Lithium-7 will also breed tritium if the neutrons are powerful enough, and emits an additional neutron to continue the reaction. Reactions that we might never have discovered (or probably not until later) if it hadn't been for this mistake.

The end result was a lot more fuel for the bomb, and the explosion was so large that many of the measuring instruments were vaporized. The large yield also contributed to a radiological disaster [1], which was then the inspiration for the original Godzilla [2].

Anyways, that's how a math/chemistry mistake lead to the most famous kaiju movie (series) of all time.

[0] https://en.wikipedia.org/wiki/Castle_Bravo

[1] https://en.wikipedia.org/wiki/Daigo_Fukury%C5%AB_Maru

[2] https://en.wikipedia.org/wiki/Godzilla_(1954_film)

> more abundant fuel (on earth and the rest of the solar system)

This is by far the most important reason in the long term. Between stars and even at our own outer planets where solar panels aren't reasonable fusion is the only long-term large scale energy source.

It's the difference between being stuck as a Kardashev I or II civilization or approaching III.

> Should we use fission right up until we have viable fusion? Of course, we should definitely be building more fission reactors. But I can't think of a single reason we'd continue using fission once we get to fusion.

The power density and relative simplicity of fission (including mere thermocoupled) is still worthwhile for robotic probes or initial sources of power in distant places, but we'll be able to make our own fissionables indefinitely once we have solid fusion power.

> no radioactive waste

It can create things that take 100 years, a closed cycle thorium breeder takes 200-300 years.

And the waste from that process is actually quite useful to extract isotopes for medical, nuclear batteries and other applications.

> no nuclear meltdowns / runaway processes

Neither can a properly designed fission reactor. And in a molten salt reactor all dangerous gases are chemically bound in the salt and if removed from the reactor would freeze instantly not realising them into the air. So even if some basically unforeseeable chain of event lead to a runaway process, it would not release gases into the air and would stay contained on the reactor site.

A fusion reactor is actually more likely to release dangerous gases into the air in case of an accident.

> more abundant fuel (on earth and the rest of the solar system)

Thorium is incredibly common on most rocky planets. Its already a waste in rare earth mining, so likely you wouldn't even need a single new mine. Every country has enough thorium in the ground to power itself. Mars has plenty of Thorium as well.

Fission does not have any practical issues in regards to fuel availability.

> - less pre-processing of fuel

Depending on the fusion reactor you still need some preprocessing. Depending on fission reactor you need more or less.

If you have a continuously refundable thorium breeder you actually need very little pre-processing other then devolving the metallic thorium into salt.

While that might be an advantage, I don't see it as some gigantic advantage that it worth the additional complexity of fusion.

> fuel cannot be used to easily make weapons

Most fusion reactors that are considered today absolutely can be used to create nuclear weapons.

I would argue starting a nuclear weapons program if you have control over a fusion reactor is far easier compared to when you have a thorium breeder.

With neither is it easy in any way.

Practically is mostly a non problem. This is a buggy-men, and would still be with fusion.

Well, the cost is the reason why you might not want to do fusion if fission works. That said, I'm not anti-fusion. I'm just miffed that we rush into fusion when we have so much improvement on fission that could solve the exact problems fusion is trying to solve.

It is, of course, the regulatory cost they want to attack. Fusion can’t continue melting down. Unlike fission. The fact that it’s much harder to get to work at all is considered kind of an advantage here because it means the reactor can’t accidentally keep producing energy when you don’t want it to. Second, if you get really good at fusion there are some reactions that don’t produce neutrons means you don’t get all this activated material. And that lack of neutrons also makes it proliferation-resistant (and can allow more compact ways of generating electricity than a typical thermal cycle). But even with the low hanging fruit type of fusion with tritium & deuterium, you don’t get these long lived transuranic isotopes. Also, fusion has some important very long term applications in human spaceflight (& interstellar travel). And the fusion fuel infrastructure is much less susceptible to being diverted to making weapons.

But overall I agree with you. Fusion makes fission look really easy, and there are advanced fission designs and processes which address most of the above issues.

If we built a fission reactor with the low power density of a fusion reactor it would be so big it couldn't melt down, just from its own thermal inertia.

Arguably a fusion reactor has a much higher chance of harming the population around as a breach in containment could release highly radioactive gases.

> But even with the low hanging fruit type of fusion with tritium & deuterium, you don’t get these long lived transuranic isotopes.

You don't get these in a good fission cycle either.

> Also, fusion has some important very long term applications in human spaceflight (& interstellar travel).

Yes but and fission has a lot of applications in human spaceflight too. And not some theoretical interstellar travel, but rather in things that are actually useful and that we need now.

Mars surface power most importantly. Nuclear Electric Propulsion second most importantly.

Its all fine to dream of interstellar travel, but by any logical view to world, Mars is a closer term thing then interstellar space travel.

Also, if we just want to send interstellar probes. Lasers driven by fission are much more likely to be a good solution in the next 100 years.

Fission has a few unavoidable issues: long radioactive isotope half-life, weapons proliferation, and the runaway scenario. Fusion has none of these issues plus the fuel being abundant. It's the answer to the question "how do we meet the growing energy demands of humans over the next thousand years?".

The only real danger is that of a tritium leak, but the short half-life makes the prospect of a leak less concerning.

Since the reactors would ostensibly be continuously operating, the danger is the same. 10+ year half life is about as dangerous as 10000 years because you are always generating the waste.

Er, no. If you generate the same volume of waste per year, but for a fusion reactor the waste stops being a problem within a century and for a fission reactor it takes 10,000 years, at any point in time after the first 100 years, waste from a given fusion reactor will be less prevalent than waste from a given fission reactor.

It also doesn't matter that no specific nuclear reactor will have a lifetime of 10,000 years. The problem is that per megawatt of energy generated, fission theoretically creates (much) longer-lived waste than fusion. Over a longer-than-one-hundred-year timeframe, equivalent amounts of energy generation result in vastly different waste carrying costs. Fusion's waste carrying costs are much lower.

And obviously that number is even more in favor of fusion if it only takes 10 years. (ITER claims 100 years though: https://www.iter.org/sci/Fusion)

The danger is not the same between fusion and fission.

Firstly, 10 years worth of energy is inside a fission reactor and is capable of releasing most of that energy in an instant if not properly controlled. This cannot happen in a fusion reactor. A year's worth of fuel is in a gas tank on the wall and needs absurd conditions to ignite. It cannot happen spontaneously.

Secondly, the exhaust is helium-4: a stable isotape of a valuable element.

Thirdly, the neutron bombardment in a fusion reactor activate the materials they hit. If they hit lithium then they make tritium: a much needed isotape for fuel in first generation fusion reactors. The other materials they hit are chosen to have half-lives of less than 100 years. So you have a nuclear site that no one's allowed to touch for a while then you can recycle the materials. It's nothing like the transuranium nuclear waste from fission plants.

If you had a half life of 5 minutes you'd only need to store the waste for a couple hours before burying it somewhere and there's a certain (smallish) amount of high-grade waste at any given moment, no matter how long you run the reactor.

10 years isn't 5 minutes, but it means you just need to keep it secure for a few decades before burying and forgetting it rather than many human lifetimes.

Any leaks will be (to some extent) self-cleaning, insofar as they'll decay substantially within a human lifetime, so if you stop the leak you can wait a couple decades and it will have cleaned itself up. That's much better than the long-life stuff fission produces.

Why is the runaway scenario unavoidable for nuclear fission? Pressurized water reactors for example have negative void and temperature coefficients, and from what I've read about potential future reactors designs (e.g. high temperature gas reactors and some of the molten salt reactors) passive/inherent safety is a major selling point.

All of those solutions involve better ways at keeping neutron multiplication factor below 1. Fission reactors need fuel that naturally have an eta above 1. Fission reactors are also designed to have large quantities of fuel inside of them. So if the control systems fail, even if those control systems are built-in chemically, then you have a disaster that lasts for timescales that humans would prefer not to be on the table.

What's a scenario where passive safety systems, say those of a HTGR with TRISO fuel, fail and cause it to go prompt critical? I'm genuinely curious about this, as everything I've read suggests this is essentially impossible due to the design of the reactor, fuel, and coolant.

I don’t have an answer. Not knowing failure modes doesn’t mean they don’t exist. HBO’s Chernobyl series highlighted how dangerous surpressing information about fission reactors is. No one knew how the design could fail until it did and then it was painfully obvious. I’m not saying that HTGRs can meltdown nearly as readily as RBMKs, but the risk of the unknown needs to be given respect when the stakes are high.

It’s difficult to be sure of safety in complicated systems when the only people with enough technical expertise to fully vet the systems have an interest in their success. I’m not saying it can’t be done, but I think it slows policy down significantly.

How could someone demonstrate the safety of these systems if their very association with those systems is a sufficient reason for you to doubt them? If the research and experiments of nuclear engineers, scientists, and regulators from around the world cannot be trusted to develop or assess the safety of fission reactors, why does this change with fusion? I also have not seen evidence that anyone is attempting to suppress information about nuclear safety. Overall nuclear power has an outstanding safety record and ranks among the lowest deaths per TWh of any energy source (and this includes Chernobyl)[1][2].

For the record the HBO series on Chernobyl, while a good show, greatly exaggerated parts of the story. There was no threat of a megaton-level thermonuclear explosion that would destroy Kiev or make huge parts of Europe uninhabitable from the melted core coming in contact with water. The soviets did know about the RBMK's propensity to have a runaway reaction, and the rest of the world never allowed those types of reactors to be built.

[1] https://www.statista.com/statistics/494425/death-rate-worldw...

[2] https://ourworldindata.org/grapher/death-rates-from-energy-p...

Assuming all that is true, reality is extremely unpredictable. Imagine a country is at war and they accidentally drop bombs on the reactor that crack the fuel and change the chemistry enough. A volcano erupts under the plant. Imagine a nuclear weapon going off nearby and causing a meltdown (for example, if the attacker was using a "low yield" neutron bomb). Imagine an astronomical phenomenon that happens to pass through the plant.

Low probabilities, but man they would suck.

We keep finding new ways to make fission safer and reducing risk of runaway scenario but I'm pretty sure we'll never reach zero risk. Sure we might reach it on paper but human error can always happen. Chernobyl operators thought their reactor design had zero risk of exploding, current reactors are much safer but I'm pretty sure the risk isn't zero.

With modern reactor designs the inherent safety mechanisms mean that humans are not in the loop to reduce reactivity or remove decay heat.

Here's an example from Argonne National Laboratory:

> In the first test, with the normal safety systems intentionally disabled and the reactor operating at full power, Planchon's team cut all electricity to the pumps that drive coolant through the core, the heart of the reactor where the nuclear chain reaction takes place. In the second test, they cut the power to the secondary coolant pump, so no heat was removed from the primary system.

"In both tests," Planchon says, "the temperature went up briefly, then the passive safety mechanisms kicked in, and it began to cool naturally. Within ten minutes, the temperature had stabilized near normal operating levels, and the reactor had shut itself down without intervention by human operators or emergency safety systems."

https://www.ne.anl.gov/About/hn/logos-winter02-psr.shtml

There is obviously a difference between:

- There are many passive systems that work in concert to prevent the fission material from having a runaway chain reaction that continues on its own,

and

- It is literally impossible within our understanding of physics for the reaction to continue without the continued application of power to the reaction chamber.

No matter how 'safe' the former gets, it's just asymptotically approaching the latter. There will always be more assumptions and caveats involved in preventing a self-sustaining reaction from continuing.

In particular, re. that article, a lot seems to be resting on the sodium cooling pool being present while there's something else going wrong. So what if an earthquake breaks it open and dumps it out. Or a bomb.

Their risk isn't zero but even the 60s the inventor of PWR said this was not very safe design and designing inherent safely systems are far better.

Of course you never have zero risk. That literally impossible and not a standard you would use for literally anything else in human existence.

The fact is, you can design nuclear power plants that are so safe that the chain of events you had to come up with to get any radiation outside of the reactor safety boundary is so ridiculous that the probability of them happening is barley measurable.

Sure if you have human error and 3 black swan events on the same day, the risk is not zero.

But even if you come up with these crazy events the damage from those events would be a far smaller then Chernobyl and Chernobyl was also far less damaging then in popular imagination.

The risk that somebody dies during the construction of the reactor confinement building is probably 100000x higher, but nobody seeks to prevent ever building large structures.

> Chernobyl operators thought their reactor design had zero risk of exploding, current reactors are much safer but I'm pretty sure the risk isn't zero.

This is where we are with nuclear. Any debate goes back to Chernobyl. Again, in no other area do we go and say 'well the soviet thought this in the 60s so therefore we can never moved past it'.

There is fundamental physics and chemistry involved and just because some soviet operators didn't know that does mean its unknowable.

Humanity should be living in the nuclear age. Climate change would not even be a thing if everybody had done what the French have done in the 70s. And we would be much better in terms of space exploration if the whole world were not so reluctant about using anything nuclear.

I guess people are increasingly frustrated with the failure of the current nuclear reactors to cost scale in the modern economy, so they look forward for a fundamental shift in technology. That is, there is no hope in fission, it has not stood the test of time so far, and it there is no reason it will in the near future. If we want nuclear we can only hope for fusion.

Note that this is me projecting. I don’t have a horse in this race. I’d be perfectly happy with nuclear free Earth; with renewables being our primary method of generating energy; a future which as of now looks the most likely. And if people develop fusion at some point in the future... cool.

Yes, it's magical thinking born of frustration, not something that one should truly expect.

I imagine that the popular view of fusion will be better, because of the impossibility of meltdown and the much easier to manage spent materials. If that translates into less NIMBYism, the regulatory costs should be much lower.

If the inputs can then ever be scaled, it could present a gateway to powerplant "mass production", which would be truly revolutionary. Especially for those crazy rocket concepts!

If a fission reactor had a power density as low as a fusion reactor, it would also be impossible to melt down, just because the thermal inertia of the core would be so large. Of course, it would also be uneconomical, because of the cost of that larger core (just as a fusion reactor would be uneconomical.)

re: regulation. This is the problem with fission. but not necessarily so with fusion. The reason is that unlike fission, where it's impossible to a small company to get a handle on fuel and build a fission plant in a less-regulated country, with fusion that becomes a lot more approachable.

But I am just as skeptical as you about the future of fusion and fission in the US and Europe.

People want fusion because fission reactors leave long-lived radioactive waste, can melt down dangerously, and use the same fissile materials as nuclear weapons, making combating nuclear proliferation difficult.

That's why there are extremely high regulatory costs associated with fission reactors.

(I'm not saying we should wait for fusion reactors, but there's a lot of good reasons to develop them, and once fusion reactors are available there's a lot of good reasons to stop building fission reactors at that point.)

The issue is that the current fusion choices (deuterium) produce neutrons during fusion, that hit your containment vessels, and ... create long-lived radioactive isotopes while slowly embrittling the containment vessel. When you replace it, it is in fact, a long lived radioactive husk that you need to safely contain for a long time. There are other fusion options that do not, but they are MUCH harder to fuse than deuterium (what we are still trying to achieve)

While true, it's still better than fission in terms of waste. (And much less liable to Fukushima-style meltdowns, and less likely to lead to nuclear weapon proliferation.)

Well, actually, a nice source of a large number of neutrons is a HUGE proliferation thread: weaponizable Pu-239 is produced out of easily-available U-238 by neutron capture. Only generation 2 (or 3) fusion, when we move past deuterium, will be proliferation-safe.

It's a threat comparable to fission reactors only if fusion reactor containment vessels are built out of uranium, which they aren't.

To put it another way: operating a nuclear reactor today is expensive due to regulatory constraints meant to prevent nuclear weapon proliferation. If the fuel for your reactor can't be mistaken for nuclear bomb parts, and the components of your reactor can't be mistaken for nuclear bomb parts, it's a lot cheaper to build and operate. And it's a lot safer for someone to sign off on "Yep that's a whole bunch of lithium for a fusion reactor" than looking at a bunch of uranium and being like... Well...

> operating a nuclear reactor today is expensive due to regulatory constraints meant to prevent nuclear weapon proliferation

The way I read that, you seem to imply that this cost dominates all others. If you do mean that, i'd like to see a citation please.

Here you go: https://www.americanactionforum.org/research/putting-nuclear...

"These figures have profound implications for the industry’s bottom-line. Based on a review of per-plant profitability, there are at least six plants nationwide where regulatory burdens exceed profit margins."

Regardless, as I mentioned, the entire process is safer from a proliferation perspective.

Those regulatory burdens described there exceed profit margins on operations. Those margins ignore the sunk capital cost of the plants.

For new nuclear construction, these regulatory costs would be a small compared to the cost of actually building the plants. Of course, new nuclear plants would be outrageously unprofitable.

You can't ignore operating costs because of sunk capital costs. That's the sunk cost fallacy! You can sometimes ignore sunk costs if operating revenue is good. But if it's negative, that's the problem.

I wasn't ignoring operating costs. I was pointing out you can't ignore capital and financing costs.

This is exactly my point, a properly designed molten salt thorium fission reactor is really terrible to build nuclear weapons (no nation would ever do it) and a Fukushima-style meltdown is literally impossible.

The waste from a closed cycle breeder reactor only has to be stored less then 300 years and you can put it back into a mine before that if you want.

People want fusion because of mental inertia. Fusion was long sold as the future, so the idea that's it's desirable has become a social default. But it's a fossil belief, perhaps true decades ago, but with little to justify it now.

Fusion has long seemed unattainable, but now we seem to be approaching the point where we can actually build a useful reactor. Why wouldn't we use fusion if we could? It's a lot better than the coal and gas we're still using. Wind and solar will be a big part of the power grid going forward, but the installation costs are substantial and we'll need a lot more battery storage to fully transition to renewables.

Even assuming SPARC (or one of its competitors) works, it'll be awhile before the technology becomes mature and we can assess whether it's actually cheaper/better than fission or wind/solar/batteries. But from where we stand now it looks promising. Why the pessimism?

> now we seem to be approaching the point where we can actually build a useful reactor.

I think not. Reaching ignition in a fusion reactor would be akin to what fission achieved in 1942. There would then be enormous engineering obstacles to overcome, particularly to produce a design that could be competitive with other sources of energy. Fusion has grave disadvantages (low power density, complexity, reliability, difficulty of testing components) that must be overcome. I see nothing from existing efforts that suggest they will be able to surmount these obstacles.

There are several advantages:

A fusion reactor doesn't create create long-lived radioactive waste (or any kind of pollution).

A fusion reactor cannot be used to create nuclear weapons.

A fusion reactor doesn't require any form of mining for it's fuel.

A fusion reactor cannot meltdown in any way.

Due to these inherent safety features, the costs associated with the regulation and engineering a fusion power plant could be much lower than a fission plant.

It will be easy to use neutrons from DT fusion to get weapon-grade plutonium. This may be easier to police, i.e. "there should be no uranium nowhere near a fusion plant", but that's a much subtler statement.

And of course mining for lithium as a fuel is still necessary, so you should perhaps say "no additional mining" or something.

The need for lithium is a valid point. The amount required for fuel is very small though. A 1 GW plant would use more on initial filling than a 30 year lifetime of refilling and even that first filling would be on the order of a metric ton (unverified ballpark, dependent on reactor size, blanket thickness, and plumbing overhead). 80,000 metric tons of lithium are mined each year. We consume about 18 TW.

So I estimate that converting all electricity sources to fusion would use about 1/4 of a year’s worth of lithium, but would be enough to make 30 year plants, which would still have most of their lithium left over for recycling/reuse afterwards.

And enriching lithium is necessary. A single ARC reactor would use a good fraction of all the 6Li produced for the US thermonuclear weapons program.

The tritium produced in a fusion reactor program would make it much easier to engineer high yield fission bombs, via boosting.

> And of course mining for lithium as a fuel is still necessary, so you should perhaps say "no additional mining" or something.

True, but the quantity of lithium required to breed tritium for power generation is ridiculously low. Operating a DEMO-like reactor for 30 years would consume 2 tons of lithium, which is nothing compared to the annual consumption for battery manufacturing (around 30000 tons).

The blanket of DEMO, if using PbLi eutectic, would require 52 tonnes of 6Li. In addition to the raw lithium itself, building even a single reactor like DEMO would require maturing a new 6Li separation technology (the old one caused unacceptable mercury pollution) and industrializing it.

https://www.sciencedirect.com/science/article/pii/S092037961...

You can not actually police that sort of thing. Thorium is literally everywhere in the earth core.

You could basically dig hole right next to your fusion reactor and mine Thorium right out of it, expose it to neutrons and create Uranium-233.

I suspect a reactor that uses FLiBe can produce pollution: the BeF2 in the solution is toxic, and presumably would have to be replaced periodically. https://en.wikipedia.org/wiki/Beryllium_fluoride#Safety

If the Be is not recycled then Be consumption becomes unacceptable. A single ARC reactor (2014 paper) would use 40% of the world's annual production of the element.

> A fusion reactor doesn't create create long-lived radioactive waste (or any kind of pollution).

It can create things that take 100 years, a closed cycle thorium breeder takes 200-300 years.

And the waste from that process is actually quite useful to extract isotopes for medical, nuclear batteries and other applications.

> A fusion reactor cannot be used to create nuclear weapons.

That is just flat wrong. Most fusion reactors that are considered today absolutely can be used to create nuclear weapons.

I would argue starting a nuclear weapons program if you have control over a fusion reactor is far easier compared to when you have a thorium breeder.

While it is possible in theory, in practice nobody would ever do it. Your whole workforce would suffer from to much radiation and because of certain other parts of the material, it would be incredibly easy to track in terms of proliferation.

This is a buggy-men, and would still be with fusion.

> A fusion reactor doesn't require any form of mining for it's fuel.

Thorium is incredibly common. Its already a waste in rare earth mining, so likely you wouldn't even need a single new mine. Every country has enough thorium in the ground to power itself.

> A fusion reactor cannot meltdown in any way.

A fusion reactor is actually more likely to release dangerous gases into the air.

> Due to these inherent safety features, the costs associated with the regulation and engineering a fusion power plant could be much lower than a fission plant.

That's a nice fantsay to have as we don't yet have fusion reactors so one can just make assertion. But as I explained above, a fusion reactor would need at least as much or more regulation as the fission reactors I describe above.

And if you follow the industry you will know that those reactors have a very hard time clearing the regulation.

> I still don't understand why people want fusion.

> That said, I want fusion for crazy rocket concepts.

There you go, you answered it yourself. Fusion rockets would open up the Solar System the way the steam engine opened up the oceans.

Also on Earth, fusion would be cool. It's less dirty than fission, and the fuel is FAR more plentiful.

Fusion could be a great equalizer in terms of global access to large amounts of energy: if you have moderately good access to the ocean, you have a nearly inexhaustible source of the hydrogen isotopes available in seawater and required as fuel for a fusion reactor.

Contrast this with both fossil fuels and fission materials. Those resources are the foundation of modern geopolitics. Seawater is not, and way more people have access to it.

Literally every country in the world has enough thorium in the ground for 100s of years.

Go look up how much thorium is in the earth.

You can literally go in-front of your house, dig a whole and in theory you can have enough thorium for a year.

Thorium is literally waste material that comes out of rare earth mining, its not a limited resource in any way.

From my point of view, fusion is incredibly important for long term space exploration... but I would agree, not much else. Even the "not so long" term space exploration can be done with fission.

Modern super conducting magnets are amazing. It’s the next generation of miniaturization. We could see fusion in the home by the end of our lifetimes.

>Fusion in the home

Why would you possibly want to make reactors that small when we already have such extensive electric grids?

Because it is cool?

Seriously, having such thing on a car is a god send. And a lot of people are looking for live off-grid nowadays; plus US power grid does not look like will sustain without a lot of capital.

I'm now imagining an implausible scenario with a SPARC reactor loaded onto a trailer and being used to power Jay Leno's steam powered car.

Spaceflight

All I want is truck-size fusion reactors.

That would open up the Solar System the way the steam engine opened up the oceans.

Ideally you'd have a custom design for rocket engines, where the reactor is semi-open: you feed in fuel through one end, and have the nozzle generating thrust at the other end. Even better if you could optionally close that, when you need only electricity but no thrust.

I see why you might make the analogy. Steam ships opened up trade routes that were impractical for sail power, but the way you phrased it is odd as the oceans seemed pretty open to sail already. I think fusion rockets would be a lot more revolutionary actually.

SpaceX's Starship is the "age of sail" in space.

For the "age of steam" we need nuclear power.

DT fusion reactors have no advantage over fission reactors in space; in fact they are inferior. Putative space fusion reactors would have to use advanced fuels.

Honestly I think the grid will change over coming years. The cities will remain similar but rural towns and remote homes will be switched to local power or limited connectivity. It’s too expensive to maintain a modern grid. Connecting power used to be closely related to phone, but this is less the case with cell and satellite service. Also building utilities can be amortized in a way that assumes the entire population in a region are captured customers. But as individual homes start generating power instead of consuming it, those methods of paying for the grid no longer make sense.

The modern grid is very affordable. What you describe is not an actual problem that we have.

I am doubtful that is something I will see in my lifetime, but there are loads of reasons why people would want it. For a start in New Zealand we pay a lot more to the grid operators than I pay to generators for power most of the time, the only exceptions being large spikes in power prices due to shortages.

Obviously, to power DeLorean time machines

https://backtothefuture.fandom.com/wiki/Mr._Fusion

That would take a different sort of breakthrough. Fission, unlike fusion, does not produce actinides or fission products. But it does produce neutrons, and neutrons produce activation products, and I don’t want neutrons or activation products in my house in any quantity. And reactor designs using lovely materials like FLiBe involve having those materials around. I also don’t want them in my house.

To be clear, I would be okay with a neighborhood fusion plant so long as the safety measures were well designed. There would be no risk of massive catastrophe along the lines of a fission plant, but I would want the risks of an activation product release or a release of non-radioactive but still nasty substances to be appropriately mitigated.

That is definitely impossible. Fusion is a highly radioactive process, a large part of the fusion energy is emitted as high speed neutrons, which interact with any material they find and turn it brittle and radioactive. Even if they are relatively easy to shield from, the shield will pretty quickly become highly radioactive. Not to mention, these things will leak tritium - radioactive gas that is impossible to contain. It won't be in huge quantities, so it's not expected to be a major problem for power plants, but if it's running 24/7 a few meters away from you, that will likely change.

Not to mention that a failure in the magnetic confinement could still spew plasma, which would definitelybe hot enough to kill or maim anyone close by.

Robert Bussard tried to build a small superconducting reactor 30+ years ago. He had a company called Inesco (International Nuclear Energy Systems Company, and also Italian for igniting, per his collaborator Bruno Coppi). There used to be something about it in Bussard's Wikipedia biography but it looks like it got edited out. I don't know what eventually happened to Inesco but there is still info about it on the web.

All the 50 years, 30 years, 20 years away jokes are silly, because the time-frame always had the caveat: Given a certain level of funding.

Because the funding never came, the time frame was never going to work out.

It's like someone asking for a dev estimate, and then coming back in that time and asking where it is. If you were assigned to something else clearly it didn't get done.

We’ve been spending 0.5 to 1B a year on inertial confinement alone for decades.

https://www.laserfocusworld.com/lasers-sources/article/14175...

That's because inertial-confinement fusion experiments are applicable to nuclear weapons and get their funds via the Nuclear Weapons Stockpile Stewardship program.

Isn't that all about modeling thermonuclear weapons?

I'm working on a concept that would use the field from a COTS permanent magnet MRI machine. I've got more information at http://www.ddprofusion.com

It looks like 3d confinement. Is that what you have envisioned?

I'm not quite sure what you mean by "3d confinement". The device would contain a non-neutral plasma in a device similar to a Penning trap.

I anticipate that a semi-stable oscillation can exist that momentarily forces all trapped ions to occupy a small one dimensional region (very small diameter cylinder) along the central axis of the device.

A magnetic field causes charged particles to revolve around a set axis. Depending on their charge they revolve around in opposite directions. You can enclose this in a torus to make them go end to end but collisions cause these particles to jump loops and the plasma expands. The stronger the field the tighter the gyroradi. 3D confinement means they all run in circles in a Tokamak. But the more collisions the greater these particles move away from the center. Moving these particles back to the center to equator of the device is not possible with a Tokamak under normal operating conditions. Defeating this limitation would open new avenues of confinement. Hence the 3rd dimension of confinement is out of our reach at the present time with tokamak style confinement.

An entirely new kind of confinement could be possible. But it will depend on principles we have scarcely dreamed of.

In my proposed device, the field lines are uniform and the single species ions all orbit identically. They are able to follow their entire cyclotron trajectory, and all these trajectories intersect the same area at the same time.

Penning traps are known for their excellent containment characteristics in all directions, but the maximum density of a non-neutral plasma is not usually sufficient for fusion to occur before the Brillouin limit is reached.

My device exceeds the Brillouin limit only for short periods of time, but in a way that might be able to oscillate stably.

Lots of talk about shrinking the magnets required (which is a big part of ITER’s cost). Useful of course but the article doesn’t mention the big problems of neutron embrittlement.

I think the idea (from a couple of videos of talks I've watched) is that in ARC they want to use magnet coils that are soldered so that about once a year or so they can heat the containment vessel and then the top opens like a lid and then replace the inner parts that are closest to the plasma and take the brunt of the neutrons.

The containment vessel itself (and the coils I think) is protected somewhat from neutrons by a fluid (FLiBe) that absorbs the neutrons and convert it to heat (which is used to boil water and run a steam generator). They're also trying to figure out if there are some stainless steel alloys they can use for the containment vessel that are less reactive so that they have less of an issue with it turning radioactive.

I don't think they're looking at solderable magnet coils for SPARC, since it's more of a prototyping platform and not something that's meant to run continuously for a long time.

If it is a startup it should aim at a product: does anybody know what is the plan to harvest the energy (how to convert it too electricity)

I guess it's generated as heat, which boils water, which drives steam turbines.

So the reactor is submerged in water?

They will run coolant through the reactor walls, and send that coolant to a steam generator.

I clearly know nothing about fusion reactors, but couldn't be have a reactor with upgradeable magnets as they continue improving?

The magnets are the reactor.

But, building a whole new reactor complex would involve moving a great deal of money over a long period, which is the real point of the whole operation. Big Construction is synonymous with Big Corruption, the true reason for fission plant construction cost overruns, as for municipal tunnels and for military programs.

Meanwhile, renewables get cheaper every year, and projects produce marketable power immediately.

A recent development is floating solar on hydro and municipal reservoirs, and building it into roofs over canals, reducing evaporation, algae, and toxics production, without consuming land, and offering easy tie-in to existing distribution networks.

Such synergy is nowhere near exhausted. For example, a farm can erect a wind turbine far from distribution networks, and use the intermittent power to produce ammonia from just air and water. Ammonia is easily and cheaply stored indefinitely, and is directly useful as fertilizer injected directly into furrows behind plow blades. It is also directly useful as fuel for farm machinery, which can easily accommodate the somewhat larger tankage vs. diesel fuel.

These are the kind of advances we need in fusion research. Who knows what these new magnets will unlock.

Fusion: let’s make a tiny star and hold it in place with magnets.

So dang cool.

The more I read about fusion, the less I think it's anything near to a star. The kind of fusion is completely different. A sun wouldn't work on earth because it doesn't produce enough energy per square meter. It isn't even in de same ballpark. The only reason why our sun produces so much energy is because it's so incomprehensibly large. A compost heap produces more heat per square meter than the sun.

Fusion reactors are incredibly cool. But they're not "harnessing the power of a sun".

> "harnessing the power of a sun".

isn't that solar?

That's why fusion is so cool. We're making something that runs a higher temperature than the sun's core. It's not even strong enough for us! We can't rely on that super slow quantum tunneling fusion; we're gonna make stronger yet thinner stars at home and literally suspend it in extremely limited space.

What we need is a SpaceX style company to accelerate development of fusion reactors. Elon make it so!

Although perhaps not with so many explosions...

>"Newer high-temperature superconductors—so called because they can superconduct at relatively balmy liquid nitrogen temperatures above 77 kelvins—were not around when ITER was designed. But they can carry much higher currents, tantalizing fusion designers with the prospect of smaller, cheaper reactors.

Yet they are brittle, persnickety materials, so “a lot of people had given up on them,” says Rod Bateman of Tokamak Energy, the U.K. startup that is also betting on the technology. “They were just too unreliable.”

In the past decade, researchers have developed ways to deposit thin layers of

superconducting rare-earth barium copper oxide (ReBCO) on metal tape.

The tapes can be manufactured reliably in long lengths, and perform best at about 10 K. But in terms of low-temperature engineering, “10 K is a lot easier than 4 K,” says magnet engineer John Smith of General Atomics in San Diego.

The ReBCO tapes can be bent but, being flat, are challenging to wind into coils, Mumgaard says. “You have to stop treating it like a wire and asking it to do the things that wire does.” Commonwealth has developed a cable with stacked layers of tape twisting like candy cane stripes.

The company believes the cables can carry enough current to generate a 20-tesla field—1.5 times stronger than ITER’s—in magnet coils just a few meters across."

Weird Idea: Might have future applications in warping space, like in a warp drive... (but don't ask me how that would be possible at this point in time!)

Reply on news.ycombinator.com