Lockheed Martin has obtained a patent for a compact fusion reactor | Hacker News Reader

Lockheed Martin has obtained a patent for a compact fusion reactor

thedrive.com

257 pointsjoak200 comments

The real deal is here, from MIT: https://www.youtube.com/watch?v=KkpqA8yG9T4

Apparently there was a recent breakthrough in superconductors, which allows significantly more current in the inductor coils while still maintaining superconductivity. This in turn allows for much stronger magnetic fields, hence tighter confinement of the plasma and therefore more fusion. A standard tokamak with these new superconductors should produce more energy out than energy in, and be a viable source of energy.

Eye opener is the chart at 18:03: "from 1970 to 1995, [fusion] power and energy increased much faster than computing power (Moore's Law)" [1].

Magnetic-confinement fusion stalled, according to Dr. Whyte, because to get more power (and efficiency) we need stronger fields. To get stronger fields we needed bigger magnets. To support bigger magnets one needs bigger structures. Big magnets and big structures will absorb more magnetic and neutron flux; that increases maintenance frequency, complexity and cost.

Better superconductors let us get stronger fields without bigger magnets. That changes the cost function.

[1] https://youtu.be/KkpqA8yG9T4?t=18m3s

"In 2011, Japanese scientists stumbled across a discovery which increased a metal compound's superconductivity by immersing iron-based compounds in hot alcoholic beverages such as red wine.[40][41] Earlier reports indicated that excess Fe is the cause of the bicollinear antiferromagnetic order and is not in favor of superconductivity. Further investigation revealed that weak acid has the ability to deintercalate the excess Fe from the interlayer sites."

I guess iron based superconductors are a relatively new and hot area of research.

https://en.wikipedia.org/wiki/Iron-based_superconductor

The more I read about this, the more I kept waiting for a Douglas Adams joke. Science is just silly sometimes.

The lack of investment in this field is absolutely ludicrous given the potential payoff. I suspect some amount of lobbying by entrenched power and fuel interests that would be rendered obsolete, but I think more blame can be placed on that awful meme that brainwashed a generation of scientists and engineers into dismissing fusion as a boondoggle:

"Fusion energy is 20 years away and always will be."

This shows how much damage a superficial catchy meme can do if it manages to slip in and override more nuanced and informed thinking. It's why I distrust catchy sloganish rules in other fields too, such as "premature optimization is the root of all evil" (leads to erooM's law in software) or "never roll your own cryptography" (discourages people from learning about how to create secure systems), etc.

Your post hit a lot of notes I agree with. Programming dogma doesn't make a lick of sense sometimes. I wonder if the "don't implement cryptography yourself" rule is disingenuous, and really about not necessitating more work from our governmental APTs to get peepholes into everything. I'm also amazed at how often people cry premature optimization when there's potential to reduce hardware cost by orders of magnitude by not writing slow, memory hungry junk by default.

I feel you. While there's plenty of history (hundreds of years) to indicate you shouldn't rely on rolling your own (though Jefferson did, and did a helluvagood job.) It's crazy easy to fool oneself about how great one's own crypto is. However, what you can do is roll your own and then encrypt your data once again - after that initial encryption - with excellent standard methods. This can allow decryption in parallel in some circumstances - if you want the best speed - by encrypting a one-time-pad with the roll-your-own and sending that in advance (and decrypting that pad in advance, so you just have to xor it into the mix later.) When encrypting fer real, first use the one time pad on the data and then encrypt it with regular methods.

20 years ago they said "Fusion energy is 50 years away and always will be." Soon enough it's going to drop to "Fusion energy is 5 years away and always will be." Which is just silly.

We'll still be able to confidently say Virgin Galactic is only 6 months away from sending up paying passengers.

Right. The light at the end of the tunnel recedes until it doesn't.

exactly, these slogan almost look like a doctrination campaign. but this is off topic.

Apparently there was a recent breakthrough in superconductors [...]

Recently in this case means mid 1980s. Is the ITER design really that old that they could not make use of YBCO superconductors?

The recent breakthrough is high current YBCO tape, not YBCO itself.

https://en.wikipedia.org/wiki/Superconducting_wire#Coated_su...

IIRC YBCO can't be feasibly made into wires. However new cuprates have been discovered which are less brittle:

http://en.wikipedia.org/wiki/Bismuth_strontium_calcium_coppe...

I am not an expert but according to Wikipedia BSCCO is not suitable for high magnetic field strengths. In the video - which I only skimmed - the main point seems to be that they can now build vastly smaller fusion reactors because now super conductors supporting higher fields strengths are available and the energy density of a fusion reactor scales with the fifth power of the magnetic field strength. They also specifically mention rare earths which does not match BSCCO which were, last but not least, also discovered in the late 1980s.

The biggest question in my mind about the ARC proposal is how well these new superconductors stand up to radiation. This is a concern for ITER as well, which uses more traditional superconductors. The ARC team is looking into it, but I'm going to remain unenthusiastic about ARC until they've really proven out the concept.

Around 33 minutes into the presentation he does address radiation briefly, noting that it's favorable because most of the issues relate to the ceramic insulators in traditional SC magnets. In their design, the steel structural component of the tape serves as the insulator; they just wind up the SC tape and the current travels along that 1um SC layer.

More generally, the demountable coils suggests that designs could be made where the magnets are replaceable with some degree of economy.

Preface: I know nothing about this.

How can it "produce more energy out than energy in"? I don't know if I am not aware of what you mean by that, but that doesn't seem possible?

Current fusion reactors require more energy put into them to start and maintain the fusion reaction than is harvested from the actual fusion reaction. Better superconductors means it will take much less energy to start and maintain fusion and thus the reactors should produce energy rather than consume it. It doesn't break the laws of thermodynamics since the energy produced is from fusing hydrogen into lower energy state helium, just currently we use more energy than the process produces to make it happen.

It depends on what you count as "energy in". A solar panel produces more energy out than it takes in - if you don't count the sunlight as "energy in." A fusion reactor (hopefully) produces more energy out than energy in if you don't count the fuel used. Current designs require us to have another massive power source keeping the reaction going.

Aren't there some reactors that produce more fuel than they spend?

Deuterium-Tritium fuelled reactors can produce Tritium by wrapping the reactor walls in a "blanket" of lithium. The lithium captures neutrons and transmutes into Tritium. It should be possible to produce an more tritium then the reactor requires, but no one has actually done it yet for fusion reactors. Fission "fast" reactors can do something similar by turning uranium into plutonium. This has been demonstrated by several countries but so far has been uneconomic.

Breeder reactors are possible because elements above the atomic weight of iron are storing a lot of energy that's released if you can bust them up (fusion.) Iron is the least (stored) energetic state, so below iron, fusion releases energy. It's a real nasty process to control, though since neutron radiation tears the shit out of container metals over time.

Thanks for the details. I don't think I even have to confess that my knowledge here is rudimentary. I just finished Atomic Awakening and I remember the phrase popping up a few times.

Energy IN in this context is the energy needed to make the fusion possible and controllable, and energy OUT is what you get from the reactor. For this to work as a generator you want out to be bigger than in, obviously.

I see. So the OP is just saying that it is above the breakeven point, as in we are not losing energy from running it. Seems like a pretty low bar for energy production :). FWIW, that is a pretty misleading statement when not contextualized and some people could become very confused.

The energy produced by the fusion reaction is greater than the energy required to make the fusion reaction occur. The extra energy comes from inside the atoms being fused, it's not appearing out of nowhere.

That's the whole point of fusion: it produces energy. But you can't do anything practical with it until you have a reactor that produces more power than it takes to run.

The parent is pointing out that mass-energy is conserved. There's nothing that can produce it.

To clarify, it produces energy equal to the difference between the mass of the fuel and the mass of the final byproduct multiplied times the speed of light squared. Energy and mass are conserved, but some mass is converted into useful energy.

Doing so in a way that produces sufficient energy to sustain the fusion reaction without creating an uncontrolled reaction ("boom") is the trick that always seems to be 30 years away.

> Doing so in a way that produces sufficient energy to sustain the fusion reaction without creating an uncontrolled reaction ("boom") is the trick that always seems to be 30 years away.

Uh, no, not really, a run away fusion reaction has never really been a concern. In fact, that's one of the biggest advantages of a fusion reactor vs. a fission reactor. Fission is a self-sustaining reaction, once it starts you have to work to stop it (via injecting a mediator to interfere in the fission reaction), where as fusion requires constant energy input in order to maintain the reaction. The part that "always seems to be 30 years away", is achieving a fusion reaction that produces more energy than it takes to maintain (allowing some of the output energy to be siphoned off to maintain the reaction). There have been a number of techniques attempted to achieve this with the holy grail being so called "cold" fusion, where cold is defined in this context as something less than the surface temperature of the sun. It sounds like the ultimate solution to the problem though is simply better magnets, not cold fusion at all.

Assuming this pans out, the real thing needed to make this viable as something other than a novelty is how much more efficient the reaction can be made. After all, if the output energy is just barely over the input energy you'd need to scale out to ridiculous extremes to produce enough usable energy, but if it's a significant amount higher then that makes more modest size plants viable.

It's the opposite of fission. Fission produces energy by "destroying" matter (splitting it into smaller parts). Fusion produces energy by creating it (combining it into larger parts). The difference in mass between the inputs and outputs manifests as energy. In simple mathematical terms, thanks to Einstein, E=mc^2

I don't quite understand what mass is lost. I thought fusion involved overcoming the electrostatic forces to get two protons close enough that they are pulled together by a nuclear force, thus releasing the nuclear potential energy. But, you start with two protons, and you end with two protons, so what's the mass that's converted to energy?

Apparently I was mistaken. They don't just make a weird hydrogen isotope. They make plain-old hydrogen. 4 protons become 2 protons and 2 neutrons. The difference in mass between a proton and a neutron is 2.3×10^-30 kg. So, naively, you'd generate 4.1×10^-13 J per helium atom. Or, roughly 11 GJ (3 Mwh) per litre of helium at STP.

Seems a little low, though I suppose 1L is only 1/5th of a gram of helium. I still might be missing something, too.

Previously its been possible to do contained fusion but the energy cost of containment was as high or higher than what could be gleaned from the reaction. Youre right to assume there can't be any new energy out of nowhere, it comes from the atoms smashing together and becoming a new element.

Campfires "produce more energy out than energy in" too - the kindling is energy in, but once you've established a chain reaction (no, not an out-of-control chain, but yes a self-sustaining chain) you get warm. Exothermic.

But have they made any strides in manufacturing tokamaks, or are they still insanely difficult to build?

This superconductor breakthrough also seems relevant to Stellarators such as Germany's new(ish) Wendelstein 7x: https://en.wikipedia.org/wiki/Wendelstein_7-X

Well, this is the stride. If I remember (I watched the video weeks ago) building them smaller, with the steel part of the tape doing a lot of the work means you can build and swap modules to create a far smaller tokamok or stellerator that you can immerse. The same topologies can and would be used - perhaps with more stable containment (for reasons I forget.)

A practical small sized fusor is a fantasy at this point. The only reaction that stands any chance to break even is the Deuterium-Tritium fusion, which generates one neutron for each tritium atom fused. Since tritium is not found in nature and has a 12 years half life, it needs to be produced by capturing the free neutron. The classic proposal is to line the reactor with a thick blanket of molten or ceramic lithium that will breed a tritium atom for each captured neutron.

This whole design is purely speculative, there is no practical instance of this process at industrial scale and there are significant doubts tritium self-sufficiency is even attainable -the neutron capture and tritium recovery efficiency must be close to 100% or you need a large ratio of neutron multiplication that brings its own problems of nuclear waste and contamination. Tritium is a particularly hazard with the nasty habit of replacing hidrogen in living tissue, seeping out of the tinniest pores and embrittling the reactor vessel and ducts.

Assuming all these problems are solved (which are themselves already researched for decades and worthy of 100s of patents), you will still end-up with a factory sized tritium production facility, not something container sized. BTW, did I mention tritium is the key ingredient for moving from clasic fission nukes to thermonuclear weapons? (never mention the classic proliferation appeal of any environment with plenty of neutrons)

And this is just one subsistem, one problem to solve out of a vast number. But hey, they've got a patent.

This is a critical part to consider in "current" fusion research. Most of it rely on the deuterium / tritium fusion which will probably require a classic nuclear plant to generate the fuel. This makes the whole "infinite clean energy" claim pretty bogus for now.

Not only the "infinite" part in that statement is indeed bogus - but the immense neutron flux makes me skeptical of the "clean" part as well.

We know how to contain strong neutron flux. We know how to contain short to medium lived activated materials, we know to avoid cobalt is hardened materials, we don't have to deal with water based corrosion in a fusion containment, though neutron embitterment may be a similar issue.

It is very unclean, especially since it requires conventional fission reactors to produce the tritium. It’s theoretically possibl to breed tritium in the reactor blanket of the fusion reactor, but that’s a totally unsolved problem. The high neutron flux of the D-T reaction also destroys all known materials in short order by disrupting the atomic structure of the materials through atomic spallation. There are plenty of other hurdles to making a fusion power plant as oppposed to just a research reactor, but those are big ones.

There is currently a good stock of tritium that would power many such reactors for a while. So it's not insane to put off the tritium problem for later.

The global inventory of tritium is estimated around 20Kg, that's barely enough for research purposes and ITER. Fueling DEMO, the first commercial fusion reactor will require about 10Kg, and at 2-4GW thermal it will require about 300g of tritium per day, completely depleting the existing stockpile in a single month of operation.

So the existing stock is insuficient for even a single commercial reactor. Talking about container sized fusors without ensuring tritium self-sufficiency is a particularity distilled form of insanity.

Not wholly speculative they've built and tested some machinery with that tape as I remember.

It's a very strange project. If some startup was doing this, it would sound completely bogus. But it's Lockheed Martin's Skunk Works. They have a reputation for making the impossible work. Lockheed Martin is funding this internally, and, in one of the few public statements about this, their CEO said that progress was sufficient that the company was putting in more money.

Be aware that the CEO is probably factoring in PR-value in the calculation of “sufficient progress”.

Having worked at Lockheed, that is most certainly not how they think. The bottom line is pretty much all that matters and the self-investment even more striking for a government contractor.

Are you scoffing at fusion startups in general? Because they do exist.

https://tae.com/

http://www.helionenergy.com/

https://www.cfs.energy/

https://lppfusion.com/

Hire bright scientists to work on fusion, move them to some other boring research project after two months. ... But perhaps I'm pessimist about corporations.

Can't speak to this specific project, but your cynicism isn't really misplaced overall. You forgot to mention that the cool project is used as leverage to lower salary expectations...

If you want to be really cynical then the project they’re moved to is still fusion, but the NIF style used for nuclear weapons research rather than power plants. Then it’s not just about salary, it’s about tempting people who would never work on weapons research into that space.

I don't know enough to say what their odds are, other than very long. But one day, probably in my lifetime, we'll master fusion power, and it will begin a revolution that will make the industrial revolution look like a minor event. I don't think people realize that really cheap, clean energy solves pretty much every problem of scarcity. We'll enter an age of plenty like the world has never known. Limitless energy means limitless water, food, and materials. Of course it's not really limitless, but if it's even an order of magnitude cheaper than what we have now, the possibilities would be mind blowing.

We could pretty much save the world with fission today, but we don’t because politics. I’m not confident that we’ll actually adopt fusion even when the tech is ready.

We don't adopt fission because it's disastrous when things go wrong, so it scares people. Plus it produces a waste problem that lasts for an incredibly long time.

Fusion doesn't suffer from those problems.

D-T fusion suffers from all of those problems.

To the same extent as fission? You're being disingenuous here with that comparison.

Fission is required for D-T fusion at this point, otherwise you don’t have any tritium. It’s the people who ignore that, and our failure to breed tritium within a fusion reactor who are disingenuous. If you don’t know even that much about the science, I’d strongly recommmemd refraining from accusing people of being less tham forthright.

Per the article, it sounds like the fusion reactor itself could indeed be the thing producing the tritium by way of a lithium blanket. Of course, that means it'd need to have some way of generating the neutrons (the design seems to include two spots for "neutral beams", but I missed any mention of what's supplying those beams).

It is today, but my understanding is that fission is just our best source of neutrons to breed tritium from lithium 6. Fusion should also be an excellent source of neutrons, so I don't see that as being a hard requirement. I'm not a physicist, so I may be dead wrong.

I can imagine the first reactors would be prohibitively expensive.

Naturally. You don't pump money into something like this for years without expecting to get a huge payback. That patent had a lot of military applications in it.

How could fission save the world today? It's not that cheap.

Part of the cost is relatively excessive safety margins. “Relatively excessive“ because despite the reputation fission is the safest power supply we have now, including solar because installing things on rooftops is not without risk.

https://www.nextbigfuture.com/2011/03/deaths-per-twh-by-ener...

The risk is much higher because it is not only about human deaths, but also about unusable land for decades and decades.

It will be impossible to re-populate land up to six miles from the Chernobyl for the next 10,000 years, and it will probably be the same for Fukushima

Along with that, zones that can be repopulated are suffering much higher risk of cancer and misc health issues.

The safety margins are big because the risks are tremendous, not because people love to waste money.

>It will be impossible to re-populate land up to six miles from the Chernobyl

There are still humans living within the exclusion zone, still humans working at Chernobyl where three reactors continued to operate after the accident, the vast majority of gamma from the site is from an isotope with a half-life of 30 years, and the background radiation within the exclusion zone is provably less than the background radiation you find when living in high altitudes.

Chernobyl was the absolute worst case in that it had no containment whatsoever, and Fukushima was an absolute worst case for a western reactor in that it couldn't SCRAM and cool properly with multiple backup systems failing, but the implication that large tracts of land are uninhabitable for tens of thousands, or even hundreds of years is patently false.

In addition, there are no attributable deaths to either accident among the general population. Radiation doses in both cases were very low in the context of the general population surrounding these plants.

The fact is that more people died from the sudden evacuations and stress of relocating than died, or will die, from the radiation levels.

Hmm, I got the 10000 years number from this source : https://www.telegraph.co.uk/news/worldnews/europe/ukraine/12...

But I can't seem to be able to find the original, even on Greenpeace

You might be right on this

"In addition, there are no attributable deaths to either accident among the general population. Radiation doses in both cases were very low in the context of the general population surrounding these plants."

I think you need to spend more time reading about chernobyl.

Many, many of the cleanup workers at chernobyl in the weeks and months following the disaster were normal people who were essentially gang-pressed into service and handed a shovel.

All of those people wrapping tree trunks in plastic and burying them were not all soldiers or paid, professional firefighters. These people were worked until they literally fell over and died in hospital shortly thereafter. I would characterize these people as part of the "general population".

I recommend _All that is Solid Melts Into Air_ and _Voices from Chernobyl.

I know all about the liquidators, and I sympathize, but that's why I stressed "general population" several times. It's a given that the cleanup crew isn't part of the general population.

>These people were worked until they literally fell over and died in hospital shortly thereafter.

It's more accurate to say that they were literally cooked. It's horrible, but that's what the Soviets did, and continue to neglect many who are still alive. Their sacrifices prevented the absolute worst outcome of that disaster, and they certainly deserve all recognition they can get. And for leaving the liquidators out of that comment, I apologize.

The real stories surrounding the accident and the cleanup certainly are terrifying. The control room in particular and the imagery of several individuals being totally and instantly vaporized will always stick with me.

> The risk is much higher because it is not only about human deaths, but also about unusable land for decades and decades.

By that argument, I should also be accounting for the climate change effects of carbon fuels, the mining impact of basically everything including what renewable plants are made from (IDK about most, but turns out uranium’s easier and safer to mine than coal), and the environmental damage caused by us using so much energy.

I argue the reason for tight regulation is an entirely different risk: political risk. People fear it, demand control over it, vote for politicians who implement it. It’s not like any German reactor could’ve suffered tsunami-induced damage, but tsunami-induced damage in someone else’s reactor resulted in no more German reactors. A tsunami which, for the record, killed at least 15,895 people and caused a lot of environmental damage from all the consequent chemical spills. Yet no grand public international outcry against chemicals which can be spilled by a tsunami.

Humans are interesting, what we consider to be a risk or not. :)

We could save the world with solar right now and it can be started immediately and is indefinitely scalable. The problems of energy supply are political not technical.

Not just today, we could've decades ago.

Fusion power will probably not be particularly cheap unless we really do figure out how to build very small reactors. If big Tokamaks like ITER are the way forward we need to build these very big, very expensive power plants whose costs need to be amortized over the electricity produced.

I think(read: am guessing, smartly) when you factor the entire cost of fossil fuels (environmental damage to harvest, subsidies, tax breaks, etc) fusion will be cheaper.

Fossil fuels have a number of intentionally "hidden" costs.

Are you aware of how CO2 is used by plants? that's a "hidden" benefit of the large-scale use of fossil fuels.

At current CO2 levels plants are starved for CO2. At 150 ppm there would be a massive extinction of land-based plant life.

I don't think plants are "starved" for CO2, and increasing levels of CO2 most likely do not linearly increase productivity of plants [0].

There's also the other ramifications of increased CO2 concentration like: ocean acidification, warmer overall climate globally and increased incidences of severe weather, rising seas resulting from the warmer temperatures, and not all plants will enjoy the higher temperatures (or the droughts, hailstorms, strong winds, too much rain all at once), especially those in the tropics which already get quite warm.

>Our results suggest that future climate change will push this ecosystem away from conditions that maximize NPP, but with large year-to-year variability

[0]http://www.pnas.org/content/pnas/early/2016/08/30/1606734113...

edit: less blunt.

If you reject the fact that the vast majority of plants are currently starving for CO2, you reject science.

Here is one of many scientific papers on the subject: "Carbon dioxide starvation, the development of C4 ecosystems, and mammalian evolution" [1]

[1] http://rstb.royalsocietypublishing.org/content/353/1365/159

I don't think you are reading the studies you are sharing.

>Global expansion of C4 biomass is recorded in the diets of mammals from Asia, Africa, North America, and South America during the interval from about 8 to 5 Ma.

That's a 3 million year period, ending 5ma. You say CO2 starvation is down to 150ppm, but atmospheric CO2 levels have fluctuated between 180-200 (ice ages) to 300ppm (warm periods) for the past half million years or so.[1] Meanwhile C3 plants (those supposedly which suffered during your linked expansion of C4 biomass) are around 95% of plant biomass currently. Doesn't seem like plants on the whole are starving for CO2 at this time.

[1]https://climate.nasa.gov/climate_resources/24/

> I don't think you are reading the studies you are sharing.

I'm impressed by your method of argumentation.

What I mean to say is there is more to the data than any single paper, and with a bigger picture in mind there seems to be no shortage of CO2 for 95% of plants on earth(C3 carbon fixation, evolved earlier), and for the other 5% (C4 carbon fixation, evolved more recently) they can be fine in a much greater range of CO2 concentrations.

It's pretty simple and there is ample scientific evidence and writing on the topic: at current CO2 levels, plants are starving. There is even satellite-based measurements of vegetation by NASA that demonstrate a global greening of the planet in tune with atmospheric CO2 increases over the recent decades. You are denying scientific facts.

K. Go gaslight someone else bud.

Is this some kind of a joke? Global warming, full stop. Plants are not starved, they were ubiquitous until we chopped “most” of them down.

It's science.

Land-based plants evolved when CO2 levels were much higher than today. [1] In fact levels have been steadily declining from 3000 ppm 150 million years ago; the current, holocene rise is a relatively minor bump. [2]

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1692178/

[2] http://caos.iisc.ernet.in/faculty/pghosh/content/Publication...

And by the way global temperature 150 million years ago was the same as today, +/- 2deg C.

Could be. But the bio-make up of the planet was different then, yes?

Perhaps that's true. I appreciate the insight. However it doesn't really address my point, as it was directed at another comment.

The guys over at Tokamak Energy are aiming to produce a small reactor:

https://www.youtube.com/watch?v=jRyUWOUk_48

Since energy is such an important factor to the life of humans, I'd say the dollar/euro price is a measure that is too narrow. The question is : is fusion necessary ? how many will die if we don't have it ? how many will be born if we have it ? Have we got a future without it ? what will be the quality of life without it ?

Increased birth rate is not necessarily a good thing for human development at this stage. Our planet cannot support an infinite number of people.

Since we are already facing unprecedented growth that won't be stopped except by famine, world war or some other calamity, it's important we find a way to stretch our limited resources further and develop technologies that will help us expand beyond earth like nuclear fusion.

Yup. But I was pointing more at the fact that the notion of "price" is not very useful here. If we're talking about mankind's future, money should be brought back to the level of "minor technical detail".

Price is a proxy for "how much resources would it take to roll this out at scale." That's important regardless of how your economy works.

Price also tells us whether the new energy source would require government support, or require less resources than building and fueling a new fossil plant, or (if extremely cheap) would prompt people to shut down even brand-new fossil plants because the new energy source is cheaper than the cost of fuel.

Money is what we currently have as a means of getting a consensus about what is important for mankind's future and how to allocate resources based on importance and expected payoff. It's not the best system, but it's hard to introduce alternatives.

I agree that it should be that way.

Price is important to motivate people to develop nuclear fusion technology, and to finally build commercial fusion plants when the time comes.

Nuclear fission is a superior power technology already, but we've largely stopped building new plants because of the exorbitant upstart costs compared to other technologies.

It may well start out like that, but like most technology I expect it will just keep getting cheaper and smaller with time.

Fundamental engineering limits aside, fusion progress competes with battery progress, and we put a lot more money into battery research.

Correct me if I'm wrong, but haven't most large infrastructure projects become more expensive over time?

I mean more expensive in inflation adjusted dollars.

The amount of progress over time is a bell shaped curve. For a period of time after the technology becomes practical, it will get more funding & smart people involved. During that time most of the major advanced in the tech will be made. This phase may last a long time. After much of what can be done with it is discovered and optimized, the amount of progress will slow down.

Ah, yes, thank you. I was thinking about the far side of the bell curve after most of the easy gains have been made and bureaucracy sets in.

Once we get it going it will attract lot's of investment, global competition, and mass production. Every fusion reactor you build is essentially a money printing machine and capitalism excels at optimizing the production of things like that.

The moment that anybody demonstrates a real, working fusion reactor, a gargantuan amount of money will flood the sector.

At that point, fusion will go onto a Moore's Law curve and we will see exponential improvements for quite a few years.

Batteries have been around for a long time. A lot of money was (and is) spent, and yet we see slow improvement (incredibly slow on a given technology, advancement usually come from new chemistry, or new physics e.g. lightsail).

Nothing resembling Moore’s law style advancement, despite a lot of money to be made.

It is likely exaggerated but the headlines claimed the big Australian battery returned the investment within a couple of months. There’s a LOT of money to be made on energy storage.

Why would fusion reactors be moorable?

> Nothing resembling Moore’s law style advancement

As you say, batteries have been around for a long time. The rapid growth/improvement part of the battery curve happened back in the late 1800s/early 1900s.

The lead-acid grid lattice design, still the...errr...gold standard when it comes to the most amount of joules stored per buck, was invented in 1881.

(modern technologies like the various lithium battery chemistries win when it come to storage for a given mass -- thus their use in things like portable devices and cars, but lead-acid still wins when it comes to storage for a given cost).

> The lead-acid grid lattice design, still the...errr...gold standard when it comes to the most amount of joules stored per buck, was invented in 1881.

Li-ion has caught up. If you have $400 to spend on batteries both li-ion [1] and comparable lead acid [2] (deep discharge, long cycle life) cost around 3 Wh/$.

[1] https://www.imrbatteries.com/samsung-29e-18650-2850mah-2-75a...

[2] https://www.powerstream.com/BBep.htm (EP100-12)

The magnetic fusion triple product* did in fact increase exponentially from 1970 to 2000, at about the same pace as Moore's Law. Then we reached the limit of the superconductors at the time, and could only move forward by building an enormous reactor which still isn't finished. Now we have better superconductors; fusion output in a tokamak reactor increases as the fourth power of magnetic field strength.

We also have lots of different designs to experiment with, and much better computers for running simulations. As the computers improve, fusion progress will speed up, if the funding is available.

* (The triple product of temperature, density, and confinement time is the critical fusion metric; for every fusion fuel there's a triple product above which you get net power.)

Why would it be exponential? I like Kurzweil's historical exponential growth across IT (pre-dating silicon), and long-term exponential progress, technological, and longer-term biological, but I haven't seen the mechanism convincingly articulated...

Sure, more money means more improvements faster, but at best that can only amplify already-exponential progress. Unless, it leads to even more money? Or, that each improvement scales all factors? Or, that one improvement makes it easier to find subsequent improvements (a kind of positive feedback loop; accelerating returns).

Why should money make fusion have Moores-like growth? Not even silicon has it any more...

> Why would it be exponential? ... Why should money make fusion have Moores-like growth?

Because fusion research is critically underfunded and always has been. Like any project, there is probably a point where we'll hit diminishing returns, but right now we're barely keeping the lights on, much less hitting diminishing returns.

https://commons.wikimedia.org/wiki/File:U.S._historical_fusi...

> Why would it be exponential? ... Why should money make fusion have Moores-like growth?

I'd say the key to the parent's premise, is they said it'd be exponential for quite a few years (rather than indefinitely). That's likely correct. The early improvements would probably leap substantially in regards to the output possible. We saw the same thing in nuclear reactor tech.

I'd love for that to be true but physics reality doesn't care how much money you have, not everything is possible.

Moore's law is an observation, it is not an input into some process.

It is both.

The original observation is basically a winner-take-all combined with the fact that transistors scale as the square of the minimum dimension. So, if there is linear improvement in dimension, there is exponential improvement in density.

The reason why Moore's Law continued on for so long was that companies were willing to spend exponentially increasing amounts of money to hit the next technology node because of the winner-take-all nature of the product. Anybody who got to the next node forced everybody else to the next node or wiped them out of business.

This is going to be the same thing with fusion. Linear improvements in the fundamentals translate to exponential improvements (fourth power or better) in the outputs. The first folks to fusion are going to force everyone to fusion or wipe them out of business.

I agree with you that if they can find those improvements that will all hold. But there is no guarantee for that at any give budget, though the larger the budget the bigger the chances that some improvement will be found.

Sorry to be pessimistic, but although I agree that technological progress has the potential of dramatically improve our lives, I fear that instead we will get more capital concentration.

Put another way: social forces will balance out any positive effects of technological progress, so that in whole, humanity's well-being remain more or less constant.

Which makes me sad reading an article prioritizing all the military applications, because Lockheed Martin.

It's like a real life Iron Man movie script.

I prefer the Superman version of limitless energy.

https://www.smbc-comics.com/?id=2305

If this technology becomes practical and we don't have the military applications figured out, other nations will. Sorry Citizen, you must speak German now, Changeover Day was yesterday.

Energy is already incredibly cheap. In developed countries it's something like 10-20% of household spending, which is a really nice bonus if you get most of that to spend on something else, but it's also not going to dramatically change the lifestyles of those people (who basically enjoy access to limitless food and water, with quite abundant access to other materials).

It's also set to get dramatically cheaper as solar really takes off in the coming years.

Being able to use the same amount of energy cheaper is only half the equation. I'd be more interested in what we can do with 10x as much energy at the same price, or 100x. Could open up entirely new industries or make things that are currently infeasibly expensive novelties into commonplace materials.

One obvious win would be desalination plants to allow mass scale water production in areas with little or no fresh water but access to ocean water. Right now that's only economically feasible in limited cases, to my understanding.

Massive scale hydroponics at all latitudes, too. Why be limited to what the sun can grow, when we can grow our own sun right here on Earth?

You may be lucky enough to enjoy limitless food and water, but a substantial part of the world does not and it's going to get worse before it gets better as our population approaches 10 billion in the next 30 years. A lot of that water is going to need to come from desalinization, which is basically an energy problem. Vertical farming may help us when it comes to agricultural water demand, but it's also very energy intensive. Both problems reduce to energy problems.

Hunger and access to water are largely already political problems, and like I said, energy is going to get cheaper fast in the coming years (shifting them further into the political realm).

Vertical farming is going to provide perishable produce to rich people, so it's kind of weird to end on that point given where you started.

I stand by what I said. You're looking at how we meet our needs for food and water today. I'm looking at how we'll do so for 10 billion people, with a larger percentage of them consuming more resources like us Westerner's enjoy.

Vertical farming hardly enters the picture today, and we currently pick a lot of the low hanging or unsustainable fruit where water is concerned. But we're going to have to do a lot of desalinization going forward, and we're also going to need to take a good percentage of farming indoors because we just don't have the land and water resources to do otherwise.

Vertical farming works for more than just lettuce and tomatoes - the thing is it's uneconomical so far. The technology needs to get cheaper and more widespread, and the inputs of energy, water, and fertilizers need to get cheaper. Advances in materials, building technology, and transportation would also help. Cheaper energy helps with everything.

> really cheap, clean energy solves pretty much every problem of scarcity

How does it make rent in San Francisco affordable?

How does it make enough grass-fed premium beef for everyone in the world to eat steak every night affordable?

Well, if energy were cheap enough you could just dig huge, multi-level tunnels, fill them full of grow lights, and raise all the grass-fed beef you wanted.

It's the same with water -- there's plenty of water in the ocean. If energy were cheap enough, you could just distill seawater and get all the freshwater you wanted.

"Limitless" energy would mean the ability to grow crops and meat and everything with very little worry about efficiency.

Flying apartment buildings. Lolz

> I don't think people realize that really cheap, clean energy solves pretty much every problem of scarcity.

Except the one people care about, scarcity of mates.

Less scarcity of sex robots? Only half joking.

Nuclear powered sex robots!

how's there a scarcity of mates? for each male there is a female (approx.)?

Not in China.

Last time this came up, in Google Solve for X in 2013, it was quickly picked part in physics forums by scientists.

It's a brainchild of Thomas McGuire, aerospace engineer who has studied some fusion in the graduate school. His team don's seem know what they are doing.

I wish there would be betting market for this kind of stuff.

edit: Here is actual critique from plasma physics laboratory http://www.ipp.mpg.de/3787558/cfr

> I wish there would be betting market for this kind of stuff.

https://www.google.ca/search?&tbm=fin&q=NYSE:+LMT

I would like to bet against.

Lockheed stock is not going to react negatively if they waste few millions for this.

Okay sure but you don't get 1:1

Buy puts :)

As the parent says, it's presumably not material to the company.

Thanks Nokinside! This is a great and easy to follow takedown of the concept on several dimensions (which I've chosen to summarize inline):

First, that making the internal device apparatus more complex is counter to the production of energy (you must put stuff in the way of the plasma that has to go fast) and it's also counter to the stability of the device (the plasma that's going fast hits the stuff you put in the way, and the neutrons you produce also hit the stuff you put in the way). Second, that more complex configurations don't in any known configuration actually trap plasma better (i.e. without escape vectors along field lines), and third, that the reactor-size problem comes down to physics of temperature gradients, not (mostly) of plasma containment.

Now, I have to wander a little bit down the path of rebutting common fusion concerns raised in sibling comments:

The link then mentions that tokamaks (powerful magnetic confinement) and stellarators (geometrically engineered plasma-self-confinement) are two existing designs that can manage not to leak plasma.

ITER, as seemingly everyone knows, is a tokamak big enough to potentially be net positive but with an absurd organizational cost and build time. Wendelstein 7-X is a big stellarator potentially able to give us good info on further stellarator designs (and has been a motivator for some cool industrial design advances).

I wish that more people on forums bemoaning ITER would learn about ARC at MIT, which is a very well-considered modernisation of the "big tokamak" design, with potentially-available high power superconductor magnets. MIT recently made a large push in favor of this, with op-eds and articles in major newspapers announcing large investment in an MIT-affiliated spinoff company by the Italian energy conglomerate ENI to pursue the initial risky magnet-design and then the routine tokamak construction after.

ARC is our best hope today, and ITER is a great long term fallback. Wendelstein is a great research tool, which may someday lead to a power production system. Then there's also the dozen-plus small startups or projects nestled in big orgs, which are mostly research-stage projects, and the whole ecosystem acts as a feedback loop for progress. Everyone's learning together.

Of course, even once (er, if) ARC hits net positive rapidly repeated pulses in a few years, there's a lot of issues with tritium and other industrialization and policy concerns (as mentioned elsewhere in this submission's comments) to figure out. But huge money will start pouring into fusion as soon as we have a system that "works", and right now it really seems like we might before 2025. That money will necessarily cause rapid progress on the industrial and engineering work, because that type of work is comparatively easy, and the upside will be both very real and yet also unimaginably high.

Fusion is not taking up all of our resources elsewhere, not by a long shot, so while there's even a thin sliver of a chance of it working out it is not reasonable to wholly stop pursuing it in favour of some other plan like all-solar+wind. And right now there is a heck of a lot more than a thin sliver of a chance!

...but, it seems Lockheed's effort would be a big surprise if it's the thing.

Lockheed Martin has quietly obtained a patent associated with its design for a potentially revolutionary compact fusion reactor, or CFR.

This is a patent application, not a patent. The difference is significant.

A patent application can be filed by anyone on any idea, without government review.

A patent is granted by the US government after a review process.

Part of the review deals with whether the application contains enough information for "one skilled in the art" to reproduce the invention.

I never understood this. They casually suggest they can solve the energy crisis but think the big break-though will be making it compact? Its like saying I can turn lead into gold but I'm waiting for the travel version? It makes no sense.

It does because the non-compact version is ITER and that costs $50bn. Smaller = cheaper.

Smaller = cheaper and faster to build = more iterations = faster learning about these systems.

The fact that you can patent something before you know if it works is an abuse of the original intent of the patent system and shows how much it has been corrupted.

Why? If it doesn't work, the patent is worthless. And the patent office still gets their fees.

I'll just leave this here (mentioned in the middle of a list in the article):

https://en.wikipedia.org/wiki/Lockheed_C-5_Galaxy

------

EDIT: I'm surprised at the downvote. Anyway I'll explain my thinking for sharing that link. I was surprised at the importance of being "container-sized" size since within limits who cares how big a fusion reactor's building is. On land. In the sea or air is a different matter - and when you see the word "Lockheed Martin" most people do think of planes.

This is mentioned in the article, in this sentence:

>According to the company website on the CFR, the reactor could be powerful enough to run an aircraft carrier, power a plane the size of a C-5 Galaxy airlifter, provide electricity to cities with anywhere from 50 to 100,000 people, and maybe even speed up a trip to Mars.

So Lockheed Martin already has planes of a large enough size that such a power plant could make sense.

That's pretty amazing if you imagine it. I mean, nuclear submarines already have reactors - why not a plane?

A quick google "airplane nuclear reactor" returns:

* https://en.wikipedia.org/wiki/Nuclear-powered_aircraft

and

* https://en.wikipedia.org/wiki/Aircraft_Nuclear_Propulsion

but no practical models. (The first of these two links starts with the words "A nuclear-powered aircraft is a concept for an aircraft intended to be powered by nuclear energy." and goes on to say none have been produced.)

So if Lockheed Martin already makes planes that big, then it might be natural and amazing for them to explore containerization of the power plant for that reason.

After all, what other space or size/weight constraints are there for nuclear power plants? Where else does it bother anyone whether it's container-size or the size of a two or three or six story building?

So, this is the reason for my leaving the link to the specific airplane mentioned in the article.

The problem with the fission powered aircraft was the weight of the shielding to protect the pilot. My dad worked on the project, although I never talked to him about it.

Ok, but why?

I doubt fusion power plants in a plane is something worth considering much in the next 30 years at a minimum.

That said - it'd be for power plants that can be flown around the world and utilized on the ground once landed, as a large recharging supply source for the future in which a lot of military hardware use batteries.

Or for powering large aircraft that can stay aloft for very long durations and are armed with high powered solid state lasers used to shoot down objects. Or alternatively used for AWACS. Can you fit a fusion power plant in a large cargo plane along with the tech necessary for a large solid state laser? I can't imagine, maybe as they both miniaturized over decades, but this is all just fantastic as a premise.

The article says

>If this project has been progressing on schedule, the company could debut a prototype system that size of shipping container, but capable of powering a Nimitz-class aircraft carrier or 80,000 homes, sometime in the next year or so.

I feel completely justified in my pointing out the consequences of that statement.

Based on your reply, it might as well have said, "at this rate by the time someone born today is in college, their watches will have six or eight fusion reactors each, depending on whether they are also using it for personal trasportation."

In other words, pixie dust. Don't blame me for having reading comprehension :)

Why did I mention it? Well, I thought that it's obvious. Look, fuel costs are extreme, I did a quick Google search and got for example:

>"For the airline I work for that would mean that an oversimplified average flight's fuel cost is about 40 percent of the overall cost. It's the single most important cost."[1]

Almost nothing is as power-dense as a nuclear reactor's core, but you would need an absolutely massive plane before you could see the economy of scale to use it. Unless someone got it down to a manageable size - say, container size. And had a large enough plane to place it on.

If you read the article I linked (again: https://en.wikipedia.org/wiki/Lockheed_C-5_Galaxy ) it actually would benefit immediately from a kind of power plant that didn't have to store such huge fuel weight. It says: "We started to build the C-5 and wanted to build the biggest thing we could..."

I am not saying it would go in a C-5, only that this company has experience building absolutely huuuuuge planes. And the data sheet gives you tantalizing visions for the future. (Well, it gives me tantalizing visions for the future.)

This is how the jet fuel economy works on the C-5 currently: ""After being one of the worst-run programs, ever, in its early years, it has evolved very slowly and with great difficulty into a nearly adequate strategic airlifter that unfortunately needs in-flight refueling or a ground stop for even the most routine long-distance flights."

So you see, at the moment it's a huge gigantic plane that Lockheed Martin has huge experience with, and needs to refuel mid-flight for even routine long-distance flights.

It also has a payload of 270,000 lbs (120 imperial tons.) That's not counting the 51,150 gallons of fuel capacity.

Do you think that would carry a container?

We are looking at the kind of thing (or maybe a somewhat larger version) that might actually be able to use an on-board nuclear power-plant.

Again: where else on Earth would anyone need a container-sized nuclear power plant? (Genuinely.)

So I'm just connecting the dots here.

By the way, just for your information, do you know what percent of all human carbon dioxide emissions are caused by plane travel?

Google says: "In 2013, aircraft were responsible for about 3 percent of total U.S. carbon dioxide emissions and nearly 9 percent of carbon dioxide emissions from the U.S. transportation sector. " [2]

Sticking a container-size nuclear power plant into a plane of that size, yeah, with lots of parachutes or safety mechanisms, would be amazing.

Above a certain size, the plane would not have to be spending its energy, on carrying its energy (fuel.)

I can see that this is a visionary far-out ideal. Maybe they're afraid to bill it as such directly.

But then: what other use is there on Earth for a container-sized nuclear power plant? (Besides a ship or submarine.) The air (or space) is the only place with such ridiculous constraints.

And anyway I didn't come up with this "vision", the article literally mentioned it explicitly. I just supplied a link to it so you everyone could read through the article and data sheet for themselves.

[1] https://www.quora.com/What-percentage-does-fuel-cost-take-ou...

[2] https://www.c2es.org/content/reducing-carbon-dioxide-emissio...

I can see fusion being crucial to colonising mars, solar power is diminished and for earth we could use LFTR (for earth) as it should be cheaper.

The development of fusion at a the cheapest it's even been would bring competition to fission and also push the development of LFTR too.

The claims appear to define nothing more than, what sounds to me like, a basic plasma confinement.

Also it says it's a pending application.

It's there something interesting or especially novel here?

I haven't seen the patent - but there were minor novelties, maybe the steel-backed tape, modularity method, immersion method, etc. Probably enough to get a patent, but I suspect not a patent of great value since there will be other slightly different implementations unless their "metal-backed" tape is their secret sauce. That's possible.

"Just build it smaller" can't be patented, that's not a method. Just use tape from that's made by X corp, I can't see that being a patent either.

I was looking for the novel point for a long time.

It seems that this is just a confined plasma by magnetic mirrors.

Mirrors do not confine plasma well, at least for fusion.

I want to see one actually work before getting hyped up. There has been “good news” thru the years but nothing happens with it

In this thread from a few weeks ago: https://news.ycombinator.com/item?id=16590030, there were a lot of commenters saying that Fusion (/ cold fusion) is not much more than vapourware. Is this news evidence to the contrary?

Progress is steadily being made on fusion, not sure why anyone would say it's vaporware. Cold fusion is something totally different...

Ok, thanks for that. Wasn't sure there was a distinction between the two (layman here with a casual interest in the topic)

Cold fusion would happen without plasma involved, i.e. room temperature. Fusion happens with plasma: at least 1000s celcius. The plasma is one of the primary reasons fusion is hard (the magnets are there precisely because of it).

That's the allure of CF, pity it's impossible.

Search for "Pons and Fleischmann" to learn more about Cold Fusion :)

Appreciate that!

Actually there was a recommended video by MIT on the subject when opening the video above about fusion

It seems there is something, but not really what Fleischman & Pons saw, (and a lot of quackery), but there seems to be some legitimate effects going on.

The MIT video referred to is -

Cold Fusion - Real, But Is It Ready? - Prof. Peter Hagelstein https://www.youtube.com/watch?v=CiNDqaFPO4A

It's not vapourware, just expensive to develop.

The path is ITER, DEMO (30% greater plasma density) then commercial reactors based on DEMO

That's the most expensive path. New superconductors let us shrink the same design for much cheaper development; MIT's ARC project and Tokamak Energy are pursuing this.

As much as I'd love to have safe, reliable, economic fusion reactors, the recent progress in solar and batteries makes me wonder if they best place for a fusion reactor isn't 93M miles away. ;-)

Is it me or is Lockheed Martin trying to take over the world? They are producing their own semiconductors, moving into radar and now defining new markets to make money from.

I mean, they're in the business of war. It makes complete sense that they would be producing defense-oriented technology as a defense contractor.

This is true. Unlike others in that field they seem to be trying extra hard

This is a very big deal if they can make it practical (that's a very big if).

I sometimes wonder if we actually need Fusion right now. There's a big sun throwing it's energy at us all the time.

Cost of ITER is gonna be around $20 billion.

That's 8 Topaz solar stations, together putting out 10 TWh annually. That would give 1.5 San Franciscos electricity at the cost of maintenance/staffing.

"right now" - Probably not. But with something like fusion for free energy would could just start pulling carbon from the atmosphere, we could start to transmute elements, it would be great for interstellar travel and energy would be the limiting factor of a global AI.

Fusion power will be as 'free' as wind power. Both technologies struggle with high installation cost that has to be distributed over the number of customers and the lifetime of the plant. For fission power, the installation is already too high to compete with wind.

but fusion throws out lots of energy. the argument is that the energy will be very very cheap compared to the installation cost (because you just get so much).

Fusion reactors also consume a lot of energy. ITER plans to have a gain factor of 10, making 500MW of heat -- not electricity -- out of 50MW. According to wikipedia, a gain factor of 22 would be needed for commercial operation. The record is a gain factor of 0.67..

If fusion will ever become economically viable is an open question.

But that gain factor declines steeply if sunk costs decline, as they would for smaller reactors producing more energy per pound installed.

Fission also throws out a lot of energy, and the plus side is it's much easier to get to work (which is also the drawback...).

...and that's a good point. After working on fusion for a while, fission certainly looks attractive! I mean, it just works! And for all the hemming and hawing about safety, it's actually remarkably safe (in real terms) compared to just about any other energy source.

which sounds like a lot until you remember that total world energy consumption is over 110,000 TWh. So using your numbers, you would need 88,000 topaz solar stations at a cost of $220 trillion.

Makes ITER look like a drop in the bucket.

It doesn't get you anywhere to compare numbers between research and industry because ITER won't even generate electricity to the grid.

I'm merely saying that if we have $20 billion laying around you can imagine the practical route of powering the Bay Area completely on renewables right now.

How many $20 billion experiments are we away from fusion reaching those numbers?

And also, why not expand Fission production. We have 5000 years of global energy powering Uranium fuel supply in the oceans. Power everything with nuclear and you have a cool 1000-2000 years of no scarcity to figure out Fusion.

Fission is crazy expensive to even start up, and with expensive safety mechanisms still a risk. Then you get to the expensive and very risky / guaranteed to leak long term storage of the very dangerous waste.

Fission is expensive to start up primarily due to regulation. Modern designs are extremely safe (fail-safe instead of fail-unsafe) and produce little in the way of dangerous waste. Politics is what killed fission.

What is the state of the art in fission that is already available?

The sun throws its energy at us all the time, but half the time there's a planet intercepting the catch.

Cheap storage that scales far enough to run civilization isn't really a solved problem. We may solve it, but it's a research problem just like fusion is.

Another possibility is to get the planet out of the way by putting our solar power stations in geosynchronous orbit. If SpaceX delivers on the extremely low launch cost they're promising for the BFR, this looks surprisingly economical with current solar power satellite designs; a good book about this is The Case for Space Solar Power: https://www.amazon.com/Case-Space-Solar-Power-ebook/dp/B00HN...

Cheap storage is a vastly easier problem than cheap/sustainable fusion, though.

I actually agree that space based solar power may make more sense than fusion, though. The path for space based solar power is through well-understood engineering. The path for fusion (while I'm certain it's possible) lies through less-well-understood plasma physics.

Cheap storage is much easier than either problem, IMHO. We've already pretty much solved it to the extent needed for civilization, it's just not as cheap as our existing sources of industrial energy. But we're very close.

If you actually spent significant time looking at what fusion and space based solar power require, then cheap storage looks much easier.

Cheap storage doesn't look all that easy. None of our current technologies can pull it off; we simply run out of raw materials for existing battery technologies, and we're close to geographic limits for hydropower storage. There's research on new batteries using more abundant materials like sodium, but like many things, battery breakthroughs have a history of taking a long time to pan out.

I read the above book cover to cover so you could say I've spent a decent amount of time looking into space solar. It's well worth a read. The early designs from the 1970s would have been hugely expensive even if launch were free, but new work since the late 90s has changed matters enormously. One key innovation is a change from a monolithic design to a self-assembling modular design, with a limited number of component types that are churned out in factories in large quantities. Another is retrodirective arrays, which use a ground signal to allow an array of small microwave transmitters to return a coherent focused beam to the signal source. The book estimates a retail cost of 15 cents/kWh; substituting the estimated BFR launch cost takes that down to 4.5 cents.

Tokamak scaling laws are very well established at this point, and MIT's ARC design actually looks quite practical. The construction is modular, the inner wall is 3D printed and replaced annually, the coolant/blanket is FLiBe molten salt, and the whole thing is about ten times smaller than ITER with similar power output. The JET reactor is about the same size and was built in four years.

> Cheap storage doesn't look all that easy.

It really, really does if you look at the challenges of making either fusion or space based solar power cheap enough in real life. In fact, it's so easy we're already doing it in places. For the other two, we're decades away from useful commercial output.

I've also done considerable calculations about space-based solar power. It's obvious why Elon Musk doesn't consider it a good idea. Even if your launch is free. (I still hope people try to make it work, though...)

Doing it in places is not at all the same as doing it at the scale where you run out of the resources you were using.

Since you're interested enough to have done those calculations on SPS, I really think you'd like that book, which works out the cost and efficiency numbers in great detail.

The only comment I've seen from Musk was "You'd have to convert photon to electron to photon back to electron. What's the conversion rate? Stab that bloody thing in the heart!"

Meanwhile he wants to convert photon to electron to chemistry to electron.

To answer his question, the overall conversion rate is 40% with today's tech, and probably 60% with some more R&D. That's not bad given that you don't need storage at all, and at all times you have 30% more energy hitting your solar panels than if they were on Earth at noon on a sunny day. You're in sun 99.5% of the time.

The system works especially well with other renewables, because the ground stations are a small portion of the total cost; you can build extras, and point the power to the places you need it most.

Photon to electron to chemistry to electron definitely has a huge efficiency advantage over photon to electron to photon to electron. The battery has about >90% round-trip efficiency while the wireless power transmission over thousands of miles has on the order of 33% round-trip efficiency once you add everything up. That means that in order to get the same energy, you need almost as big of a solar array as you would in a good desert location even though you get 24/7 sunlight!

And it's not even the conversion efficiency that's the problem. It's the cost of the conversion equipment. The power electronics, the microwave amplifier, the array, the receiver array, rectifiers, and power electronics as well as transmission all has a MUCH higher cost than the actual solar cells. Additionally, the minimum size space based solar power satellite and receiver station is super expensive, and the situation only starts looking like it might be worth it when you approach multiple Gigawatts per installation.

In some ways, space based solar power is based on the idea that solar cells are expensive and scarce and their output should be maximized. Nowadays, that's a strange thing to believe because solar cells go for 16 cents per Watt on the spot market, so we tend to emphasize the constancy. But really, even that is falling prey to technological advances in battery technology.

As far as "ground stations are a small portion of the total cost" and "point power to the places you need it most," that's simply not true. The ground stations would rival an equivalent solar array in cost, not even counting the space-based portion at all! But I suppose the in-space portion WILL be crazily expensive, so you might still have the ground-stations a "small portion of the total cost" while still being crazy expensive.

And due to the diffraction limit and required safety margins, your ground stations will have to be huge. You're not just going to beam power into the middle of cities with high aircraft traffic and safety concerns. The exception to this would be if you used much shorter wavelengths, such as mm waves or lasers, but there the cost of everything (amplifiers, optics, etc) is much greater, the realistic round-trip efficiency drops to like 10-20%, and you become much more susceptible to weather. Oh, and what you're building now looks a HECK of a lot like a weapon.

Well all I can say is check my source, which covers all of this, with extensive references on efficiency, cost, and lots of other practical concerns. Some of the key research was done by NASA in the past decade.

The cost of a 2GW ground station is $700M, which is pretty decent for a peaking plant that doesn't require fuel.

The idea isn't so much that you have to minimize solar panel size, as that you can entirely eliminate the need for storage, which is a big deal once we try to get past fossil backup. To see the scale of that problem, read A Nation-Sized Battery, by Berkeley physics prof Tom Murphy. Even if he's too pessimistic by a factor of ten, storage looks like a daunting problem.

https://dothemath.ucsd.edu/2011/08/nation-sized-battery/

Yeah, I've read that blog many times. The author wastes his intelligence by refusing to pursue creative solutions to problems vs just trying to find ways to make the problem unsolvable.

The answer to season storage for solar, for instance, is to make the solar array larger, not to have a nation-sized battery. That means you only need a day or so of battery, not a week or months.

Also, why would you want to eliminate storage? Just like nuclear power, you'd want to use storage at very least to help convert a constant baseload power source into one that can follow day vs night demand. That is ultimately cheaper. And his complaint that batteries might require service? Well first of all he's off by at least an order of magnitude in cycle life, and second of all, yeah, why wouldn't we do a lot of service on batteries like we do on the rest of our energy infrastructure? That's a weird thing to focus on.

As far as material shortages: I find this highly doubtful. Lithium is not fundamentally rare. "Proven reserves" might be, but that is almost entirely a function of demand (provided your mineral isn't fundamentally rare, which lithium isn't). Other metals used in batteries, like cobalt, can be substituted by other more abundant minerals if desired, especially in grid storage. (LiFePO4 is one such chemistry.) That the author of that blog seems to not realize this pretty obvious fact strikes me as naivete dressed as "skepticism."

While we don't strictly need fusion it would be nice to have. We cannot easily generate all the power we need when and where we need it from renewables. we can generate a significant fraction of it tho. Plus for space based application fusion would be preferable.

  > There's a big sun throwing it's energy at us all the time.

That's the thing. Fusion energy research hasn't been paying off, solar energy research has. With the momentum solar currently has, it makes sense that everybody is betting on solar.

Efficient fusion power would be absolutely fantastic, but I'll believe it when I see it.

Perhaps we've just been looking in the wrong place for it. Spending all this time in laboratories, when we had it all along in the centre of our solar system.

It seems to me we are in a desperate situation - we need to decarbonise energy, but so far have only made a tiny tiny scratch with intermittent, low-density renewables. So I don't get the sanguine attitude to whether we need fusion. It looks very much like we need something else, if we're ruling out fission.

While it might not make sense for the grid there are other applications as well - like nuclear subs, battle cruisers and carriers.

And not a single thought was given to using this on railways?

What's made you leap to that assumption? One would think they'd be keen to make the thing work before they start defining its applications.

The diagram in the article.

Reply on news.ycombinator.com