> This is actually an easier problem to solve than electricity generation as the deuterium can be used as fuel and as a propellant ejected straight out of the plasma core at variable ISP up to around 100,000.

Jesus.

For reference, chemical rockets tend to have ISP in the low-to-mid hundreds, and ion thrusters are in the thousands—not ten thousands, but single thousands. A 100,000 ISP rocket is like a car that can drive from Bangkok to Vladivostok to Lisbon, then back to Istanbul and down to Johannesburg without refueling or recharging. It’s like a phone battery that lasts years rather than days. And it’s all powered by the most plentiful elements in the universe.

I’m pretty sure you could generate electricity by flying this rocket around in a circle tethered to an electric turbine if you wanted to. I mean, all of your electricity would probably be in space, because a high ISP usually implies that the rocket doesn’t work as well on the earth’s surface (not as a physical law, just as a limitation of our technology), but of all the monumentally impressive things you can build from a fusion reactor, a 100k ISP rocket has got to be up there.

As for ITER, I’m not necessarily bearish on it, but if there’s a credible chance of an alternative design that can be proven for anything <$1 billion...I hate to say it, but any person, venture capital fund, or government with more than $10 billion to their name would be hard pressed to find any better use for it. Even if the damn thing is 95% likely to fail, that’s a 5% chance you literally end up having a stake in the single most valuable technology in the single most valuable sector of the economy—one that will provide untold benefits to mankind.

30 million dollars, really? Like—does Y Combinator know about this? Never mind the “benefits to mankind”—if a single business can be first to market with an economical fusion power plant, that will become the single most valuable business in human history. There’s like five tech companies that are up there in the ranks of “most valuable business in the world”, and they’re approximately the same value as any of about ten oil and gas conglomerates, and those lists usually don’t even count state-owned companies that are even more valuable. (“All the oil in Saudi Arabia”—that’s wealth.) Fusion power would immediately be more valuable than every single oil and gas company in the world put together. Even if nobody bought fusion plants, you could just build the damn things yourself, use them to synthesize hydrocarbons from the damn atmosphere, and undercut the market while simultaneously fixing climate change by making fossil fuel usage a closed loop with no new net CO2 introduced into the atmosphere and an economical method for extracting the excess we already have.

For reference the temperature of fusion plasmas are in the 100 million degrees C range. So very high ISPs are not that crazy. There are many other problems to solve in building a usable fusion rocket, but high ISP comes for free.

One thing you should realize about a fusion reactor is that the easiest (which is still really difficult to do) reactions for fusion release a lot of their power in neutrons. These can not be harnessed directly to produce electricity, unlike moving electrons or ions theoretically could be. They can be used to boil water, just like a coal plant, and use steam to run a generator to make electricity. This part of the power plant is expensive. And the fusion reactor is expensive. You don't just get a Mr. Fusion on your car roof outputting Gigawatts of power.

As to it being a great investment, an electric power generation method is only going to be a great one if it is much cheaper than current methods. This is why ITER is such a boondoggle with its $10-$20 billion dollar price tag. It is very unlikely to be the path to putting fusion power onto the grid. Also, who knows if people will be just as afraid of nuclear fusion plants and nuclear fission. If that is the case they will not be economical to build. Maybe a name change is in order?

Relativistic magnetohydrodynamics is a hard subject matter. Investors with $30 million dollars to speculate on such a project are not going to know enough to distinguish a possible project from unlikely ones. Physicists who can know enough to see that some plan might work are unlikely to have the funds to invest. I saw Sam Altman had a fusion book on one of his lists at one point and Elon Musk might know a physicist he would trust, but getting those people to take a look is not easy. I'd be happy to make introductions though. :)

> One thing you should realize about a fusion reactor is that the easiest (which is still really difficult to do) reactions for fusion release a lot of their power in neutrons. These can not be harnessed directly to produce electricity, unlike moving electrons or ions theoretically could be. They can be used to boil water, just like a coal plant, and use steam to run a generator to make electricity. This part of the power plant is expensive. And the fusion reactor is expensive. You don't just get a Mr. Fusion on your car roof outputting Gigawatts of power.

I would kind of expect a completely new technology to be expensive to begin with.

> As to it being a great investment, an electric power generation method is only going to be a great one if it is much cheaper than current methods. This is why ITER is such a boondoggle with its $10-$20 billion dollar price tag. It is very unlikely to be the path to putting fusion power onto the grid.

I'm not sure how that follows--this might not be a great comparison, but the cost of building the first two atomic bombs was orders of magnitude higher than the cost of building the most recent two atomic bombs. Even if it costs $20-50 billion dollars per fusion power plant, it still might work out if each plant can produce enough economy of scale.

> Relativistic magnetohydrodynamics is a hard subject matter. Investors with $30 million dollars to speculate on such a project are not going to know enough to distinguish a possible project from unlikely ones. Physicists who can know enough to see that some plan might work are unlikely to have the funds to invest. I saw Sam Altman had a fusion book on one of his lists at one point and Elon Musk might know a physicist he would trust, but getting those people to take a look is not easy. I'd be happy to make introductions though. :)

Yeah, that makes sense. There has to be some reason that a $30 million plan to build a proof-of-concept viable fusion reactor hasn't been funded yet.

Tokamak fusion plants 10 times the output of the biggest power plants today are likely to work, but no one would want such huge plants on the grid. When it needs to shut down for maintenance, that amount of power cannot be easily replaced. If it was used only for hydrocarbon synthesis, hydrogen generation, or seawater desalination, this could work, but not for electricity generation.

> Tokamak fusion plants 10 times the output of the biggest power plants today are likely to work, but no one would want such huge plants on the grid. When it needs to shut down for maintenance, that amount of power cannot be easily replaced. If it was used only for hydrocarbon synthesis, hydrogen generation, or seawater desalination, this could work, but not for electricity generation.

The United States has, according to a quick Google, over 8,000 power plants. If you replaced them with 800 fusion plants, wouldn't you still have enough redundancy to take them offline for maintenance? Sure, having a smaller number of power plants increases line loss, but there has to be some economy of scale where tokamak fusion starts winning.

Thinking even bigger, if you have a 100x or 1000x scale fusion plant and used it for hydrocarbon synthesis, as you suggest, couldn't you use that plant to supply several downstream power plants that burn hydrocarbons and still, potentially, end up net-ahead?

I dunno--my impression is that the theoretical yield of fusion is so high that you can very, very quickly start reaping benefits even if you're doing so in extremely inefficient ways. Once you get above break-even, even if you're only a couple percentage points above break-even, if can build a huge enough reactor, it adds up really, really, really fast. Fast enough that some rich technocratic dictatorship could, say, build a bunch of tokamaks and effectively corner the world oil market without drilling for a single barrel of natural oil. You'd have to think that fusion starts doing things to the cost of energy that Moore's Law did to the cost of computing, to the point where people actually wrap user-facing applications in gratuitous web browsers.

You still have the cost of steam turbines and generators just like a coal/oil/gas power plant. The main cost savings is the fuel cost, which is a small but significant fraction of today's electric power cost.

There are theoretical proposals for direct conversion of electricity using the charged particles of certain fusion reactions (usually using Helium-3 mined from the moon) in different reactor designs but those are even further out than the toroidal magnetic confinement reactors (tokamaks, spherical tokamaks, and stellarators.)

I know it is really hard to believe that if fusion power was possible we as a society would just not choose to develop it, but that is the case. This philosophy/issue is discussed by many people, in many contexts and can be summarized by Peter Thiel's quote, "We wanted flying cars, instead we got 140 characters"[1]. The world had been on a runaway upward trajectory of technological improvements that peaked with the atomic bomb, nuclear power, and the Apollo program and then basically stopped.

In the 1970's a huge demographic block started coming into power, the baby boomers, and for many different reasons they questioned all of this ever increasing "improvements". Humans seemed not to have enough wisdom to wield wisely the new levels of power tech improvements were giving them. The Vietnam "war" against a country that never attacked the USA, DDT and other pesticides were killing all the bugs and destroying natural ecosystems, Russia and the US were building up an arsenal of nuclear weapons that numbered in the tens of thousands and could destroy humanity in 15 minutes, nuclear power was getting started and promised to give us electricity too cheap to meter to fuck up the planet even quicker, we kept tearing down and cutting apart beautiful cities with more and more interstate freeways, modern architecture with its anti-human scale and design, Agent Orange, etc. It was probably a good idea for humanity to take a breather on improving tech. The geeks were denied the power to keep implementing more powerful tech and nuclear powered rockets and airplane research were shut down. Denied working with real world, geeks turned to computers and no one cared. How could one do anything scary with pure calculations?

Well, whole libraries full of books have been written on the subject. For power generation just look at how it is priced for residential costumers. In California the price per kWh goes up the more you use, unlike almost any product. Our philosophy on power generation in the "tech capital of the world" is that we should do less of it, not more. Cheap power is a negative thing. It takes a strong sense of purpose to invest large amounts of money into a project that your peer group is going to say is a bad thing.

As the boomers fade from power, I see some hope of humanity too gain back some of its dreams for a better future using more powerful tools but with more balance on keeping things good for humans. Elon Musk is a good example for a third way, I hope.

Big power plants are a problem. The US generating capacity is about 1,000 Gigawatts [2] so if fusion power plants are huge at 20 Gigawatts, that is only 50 plants. One per state. You can see how one of those going down is going to be a problem. The average plant size is only 0.06 Gigawatts if your 8000 power plant number is correct. (All calculations liable to correction as this is just a quick web comment.)

[1]Thiel's view expanded on his founders fund website:https://foundersfund.com/the-future/#

[2]https://www.eia.gov/electricity/monthly/epm_table_grapher.ph...

> The geeks were denied the power to keep implementing more powerful tech and nuclear powered rockets and airplane research were shut down. Denied working with real world, geeks turned to computers and no one cared. How could one do anything scary with pure calculations?

Hah! Still, I don't think this is a total loss. Sure, we use computers for a lot of dumb stuff, but we also use them to cure cancer, control rockets, design airplanes, find our way when we get lost, produce art, and distribute information. And there was so much low-hanging fruit in computing that this focused effort let to outsized economic growth and basically brain-drained the entire rest of our society.

> Our philosophy on power generation in the "tech capital of the world" is that we should do less of it, not more. Cheap power is a negative thing. It takes a strong sense of purpose to invest large amounts of money into a project that your peer group is going to say is a bad thing.

> As the boomers fade from power, I see some hope of humanity too gain back some of its dreams for a better future using more powerful tools but with more balance on keeping things good for humans. Elon Musk is a good example for a third way, I hope.

It's not entirely clear that this cynicism will die with the baby boomers. They've had decades to instill and pass down their pessimistic attitudes while the hope and enthusiasm that took us to the Moon gradually passed from cultural memory. Even now, the most audacious project that the average millennial can think of is a "Universal Basic Income" so he gets a lifelong stipend to sit at home playing video games.

Musk is a good example of a way forward, but if anything, his story proves my point. Musk is an optimist and someone who actually believes in the American Dream, and it's no coincidence that he's also an immigrant. Optimism and ambition used to be the defining qualities of this country. Someone who believes in them strongly enough to come here usually believes in them more strongly than most of us who were born here. (Probably a corollary to the old observation that recent converts to a religion are more obnoxiously pious than those born in it!)

But America still has enough optimists, and if there ever is a small-scale fusion reactor design that gets proven feasible, it's a lot easier to convince the last surviving well-funded ambitious optimists than to convince the greater public. And if we the West doesn't make it out of this hole, there are still other hungry up-and-coming civilizations out there in the world who will get around to it in due time themselves.

Is this really only a networking issue, a matter of people with the right information being unable to reach people with the right amount of capital to invest? Sounds pretty unlikely, but if so, that's insane.

It's quite misleading to talk about 100,000 Isp for fusion. That's because unless the quantity of material you are expelling is at least as large as the fixed mass of your engine and tankage, high Isp actually hurts you. You'd be better off diluting the outgoing mass with inert material and accepting a lower Isp, but higher total impulse delivered by the stage.

How much fuel do fusion reactors burn?

A 1 GW(th) plant will burn about 100 kilograms of fusion fuel per year. ITER, which has a mass of 23,000 tonnes, will produce 400 MW(th). So, going by that, it would take nearly a million years for a reactor like ITER to fuse its own mass in fusion fuel. That's obviously silly. ITER is clearly not optimized for use in space (and burns DT, in which 80% of the energy comes out as neutrons that are useless as plasma exhaust), but any real fusion rocket will be designed to operate with an Isp much lower than 100,000 seconds.

Agreed. That is why I said variable up to around 100,000 ISP. For low thrust application having high ISP is great for saving propellant. The key feature for a fusion rocket is that one can have tanks of deuterium as the propellant you can mix into the plasma that is also the fuel.

I think the general public has just completely given up on seeing any kind of new, radical cool tech but with the decline in the anti-tech baby boomer generation a lot of progress of long dreamed projects are starting to happen. SpaceX has lead the way with reusable rockets.

Here is a link to a video from a real company that is actively researching and building fusion rocket prototypes[1].

[1]Fusion drive page: http://www.psatellite.com/tag/dfd/ [2]video: https://www.youtube.com/watch?v=F3ao5SCedIk

And here is a link to a technical paper if you are interested[3].

[3]https://w3.pppl.gov/ppst/docs/Cohen2017ibis.pdf

That's a little reassuring in the sense that 100,000 Isp is, by at least an entire order of magnitude, too good to be true.

But, who knows? Energy is such a limiting factor that maybe you could, with enough energy, make a gigantic space-ark in space and fill it with hydrogen and have a massive fusion-powered rocket with 100,000 Isp. Maybe that's how we reach Alpha Centauri.

Beamed energy propulsion makes a lot more sense.

One huge problem with any extremely high Isp rocket is heat dissipation. Any heat hitting the vehicle has to be radiated (there is insufficient reaction mass flow for that to cool the engines, as it does in chemical rockets). So it pays to keep the powerplant separate and beam low entropy energy at the vehicle.

> One huge problem with any extremely high Isp rocket is heat dissipation.

Oh, yeah, shit. I keep forgetting that's the defining difference between science fiction and reality when it comes to space travel.

Billionaires are investing. See TriAlpha.

Cheap unlimited energy leads to heat pollution per the laws of thermodynamics. Asimov Foundation series first book warned about this in passing. Be careful what you wish for!

TriAlpha's scheme was roundly criticized 20 years ago (basically, ions scatter much more readily than they fuse, so their colliding beam scheme can't work.) They responded to these criticisms, but in my opinion their responses were inadequate. And now they're pivoting to cancer treatment! It's not difficult to see where this is going to end up.

I'm not a fusion scientist either :) But I was fortunate to sit next to one in a bar in Princeton NJ a couple years ago. The gist of that conversation was in sync with what you said.

Long to short, it's close. Money could help accelerate the research into the necessary details.

They are close to achieving breakeven, maybe (for certain values of "close"). Congratulations, that puts them where fission was in 1942.

But a workable, practical fusion powerplant? Not close at all.

If the only thing the world needed to make fusion work was less than half of the Juicero Series A, we would have had fusion a long time ago

They're saying it (might) be the only thing the world needs now, after all the advances it has made up to today, not that it takes $30M to develop it all from scratch.

There is a point at which a single step is enough to win a marathon; that doesn't mean it would have been enough when the race started.