I'm not that familiar with the subject, but this phrase sounds suspiciously like the 1943 IBM executive saying there is a market for only five computers in the world - i.e. it completely discounts technological progress. Of course, this progress has been painfully slow until now, but comparing it with fission is still unfair - fission is well-researched and it getting more expensive is tied to the inherent dangers of radioactivity (and the ever more complicated measures which try to mitigate them), and the costs of spent fuel storage.

Fission getting more expensive is tied to the very lucrative fully-legal corruption industry. Once a fission plant construction project is started, the money starts flowing, and is practically impossible to stop. If it is ever finished, the money would stop, and nobody actually involved wants that ever to happen, at least not until their kids graduate college. So the project blows through deadline after deadline, and cost overruns pile up without limit, until somebody manages to pull the plug. Once it looks like the money will stop regardless, there is some incentive to deliver a reactor; we saw that in Finland recently. South Carolina is the more typical cancellation after billions are spent, except in that case with a few indictments.

Normally nobody gets indicted, and nobody ever has to give any money back.

Wow this is brutal but feels like it's correct. What a frightening state of affairs.

Is the construction industry really so corrupt? Is government and industry unaware or unwilling/unable to detect this earlier and stop it?

I spent some fair amount of years as both a government employee in DoD and then some years as a defense contractor amd I can tell you the military industrial complex is a real thing. In my contracting days I worked with a handful of DoE engineers that were more familiar with reactor projects and as I understood it, reactor peojects basically follow the same type of workflow as any big DoD prohect. So OPs comment is much in line with my understanding of the issue. It is an unfortunate state of affairs.

I've worked in fusion research and defense. MIC corruption and incompetance is quite high but virtually not present in the fusion research community. The researchers, you know, give a shit. They're not there to make a buck and milk the gravy train. If a machine gets greenlit then they pour everything into getting it to succeed because it might be the only research advancement greenlit in their life. Failure is always an option. Look at NCSX.

What the fusion research community has instead is people who have committed their careers or even lives to a particular field. Getting them to admit they made a mistake is very hard. What did Upton Sinclair say? "It is difficult to get a man to understand something, when his salary depends on his not understanding it." Most of them will rationalize as hard as they can for as long as they can.

You sometimes see bitter comments when they are approaching or are in retirement.

The fusion researchers are not who would finance a fusion reactor. The people who finance those would be the ones financing fission plants, and then milking the project for as long as it can be made to run. Since nobody expects a fusion plant construction project to take less than 20 years and $50B after it is proved possible, they will be on it like flies.

This seems... dubious at best? In terms of large nuclear reactor projects recently, we have the Finnish and Flamanville ones as examples of major overruns... but they're the first two examples of a type. Other EPR builds are going much better; one of the Chinese ones is finished and the UK one is basically on track. The timing was bad, in that EPR starts coincided with a general loss of interest in nuclear over Fukushima, but there's nothing too unusual here.

Historically the first build of a type has _generally_ been problematic.

It is, anyway, a bigger problem in the US than in some other places, and, in the US, a bigger problem in public projects than in private projects. So, we see it operating today on NASA's SLS rocket, but not on SpaceX's Starship.

But there has never been a privately financed nuke plant.

The South Caroline indictments were because some of those involved decided to pull the plug. The indictments were revenge for pulling out and derailing the gravy train.

> Tokamak fusion can only ever be far more expensive, per watt and per watt-hour, than existing fission.

Why?

The energy flux density in the most compressed plasma imaginable is orders of magnitude less than in fissile uranium. So, to match power out, it has to be big. Really, really big. And a fusion reaction chamber is a much more expensive affair than a fission pile: super high-tech materials and construction, enormous magnets (blasted by the hot-neutron flux output), piping for thousands of tons of melted lithium that have to be replaced (by robots! because super radioactive) every couple of years as the neutrons destroy them. The (radioactive) lithium itself, and containment for if it breaks out of the piping. Continuous processing of the lithium to extract synthetic tritium for fuel. Heat exchangers for lithium to a working fluid that won't pick up radioactivity, and more heat exchangers from that to water for steam. And then the usual steam turbines, which demand regular, very expensive maintenance. Cooling towers to condense the steam, after.

A hundred billion dollars estimated for one several GW plant, and decades for construction. Meanwhile, solar and wind get cheaper every year, with no bottom in sight. The longer they take on fusion, the less competitive it gets, and it starts out way behind, with no possibility of ever catching up.

This was with previous designs of fusion reactors. Read about proposed Ark and currently constructing Spark reactors. By using stronger superconductive magnets it was possible to reduce the size of the plant by order of magnitude. Spark even has a chance to get to Q>1 before ITER and it’s chamber is less than 2m in diameter. Plus it is possible to replace its parts without disassembling the whole reactor.

[1] - https://en.m.wikipedia.org/wiki/SPARC_(tokamak)

That might be nice if they had solutions to the other problems not addressed (see above). But operating it will necessarily still be enormously more expensive than solar or wind, which cost practically nothing.

Somebody with one leg can do a lot better in a race than somebody with none, but that is not the measure to consider when the competition is with two-legged runners. (Hint: they win.)

In Europe after the last autumn when for months there were little wind and cloudy sky there is a realization that energy based purely on renewables will require much bigger storage and much more solar or at least significantly more expensive solar that works under clouds. Doable, but cost is non-trivial.

So if any of fusion startups can make operating plant within the next 20 years, it can be used. But if that takes longer, then I agree that large-scale fusion will not happen.

Long term darkness/calm is solved with hydrogen. Europe has plenty of underground geology suitable for mass hydrogen storage (it's just like natural gas is stored.) Most of the energy still would go directly from solar/wind to the grid, or through efficient batteries, so the low round trip efficiency of power to gas to power doesn't matter much.

I suspect long-term, large-scale hydrogen storage is not "just like natural gas"... those pesky little H₂ molecules are awfully small, and may leak out rather easily compared to hydrocarbons.

Hydrogen storage is difficult only in metal tanks at high pressure and temperature. Store it underground or liquify it, and the problem dissipates.

Heating and lamp gas composed of of hydrogen and carbon monoxide (!) were once delivered throughout cities through cast-iron plumbing. That was displaced by "natural gas" as increasing numbers of people were found to have been asphyxiated after a lamp flame in their house died, typically from a pressure fluctuation.

Natural gas then had to have mercaptan odorants added after leaking gas accumulated undetected and exploded. Then, valves that would only stay open when hot came into regular use, for pilot lights. Now pilot lights are gone, replaced by spark generators. Hydrogen will displace natural gas in industrial settings, but synthetic methane or propane will probably be needed for domestic use.

There are already underground storage caverns for hydrogen, and more planned. It works fine.

If hydrogen is produced from electricity at scale, it won't be used for energy storage. It'll get sold to chemical plants first. They'll pay good prices.

Yes, surely that will be a major market. It just makes the case easier for renewables, as that's a major source of dispatchable demand that can smooth variation in renewable output without having to pay a cost for turning it back to electric power later.

And, anywhere it is used, it will also be stockpiled. And sold back whenever the spot price rises.

Or to steel smelters.

Electricity -> Hydrogen -> Electricity is insanely wasteful. If this is the only large scale long term storage we can come up with, we have a huge problem...

It would be a serious problem if hydrogen were used for diurnal storage. But for rare event backup, efficiency is really unimportant, since so little of the energy produced will go through that channel. What matters for use cases like that is minimizing capital cost, not maximizing efficiency.

I'll add that the alternative proposal, excess renewables with curtailment, has 0% efficiency on the curtailed power. Using that power for anything at all, including making hydrogen, is more efficient that just dropping it on the ground.

Hydrogen is so useful for so many other things that its (present) poor round-trip efficiency doesn't matter much. And, that efficiency will only ever get better, not worse, and will get better faster the more of it there is.

But hydrogen is very far from the only storage choice, it is just one that is attractive for numerous ancillary reasons: we need to make the hydrogen regardless, so might as well store it too.

You are making a commonly seen mistake there.

Yes, making magnetic fields stronger can give one higher power density than ITER. But one runs into a limit because of power density at the wall. Once you're at that limit, higher magnetic fields don't really help.

ARC's volumetric thermal power density (volume of reactor, not of the plasma) is just 1/40th that of the primary reactor vessel of a commercial pressurized water fission reactor.

60% of the mass of an ARC reactor is in the steel structural supports for the magnets, so it's about at the upper limit for steady magnetic fields in a fusion reactor.

Compare that to a simple molten salt fission reactor. It will simply never be cheaper.

Molten solt reactors do not solve the problem of nuclear waste. It is very expensive to deal with it as it has to be stored for hundreds of years unless some technology would appear to separate that cheaply. Fusion produce like order of magnitude less radioactive waste which will be short-lived.

Plus a fusion plant can be very quick to start and shutdown or throttle down which is very useful for base load.

Molten salt reactors inherently don't. But you can burn up higher actinides resulting in waste that only needs to be stored a few 100 years and pose very little danger even in the most absurd scenarios.

You don't need to separate all the different parts. You reprocess our old waste into fuel for modern reactors. In those reactors higher actinides will be burned up. And the resulting output does not have higher actinides and only needs to be stored as long.

Moltex Energy is currently engaged with the Canadian regulator for such a facility, the process is called WATSS. Doing this is actually cheaper then the deep storage storage facilities proposed in many countries. Sadly because of political bullshit reprocessing has been made illegal in the US.

Such things would have been doable decades ago, and the money for it is already on bank accounts in most countries. But politician force the 'dig a hole for everything' solutions instead.

> Plus a fusion plant can be very quick to start and shutdown or throttle down which is very useful for base load.

Next generation fusion would also be very reactive to changes. Far more then traditional.

But the reality is you actually want to run full power as much as you can, no matter if fission or fusion.

Nuclear waste is not the major problem with fission, capital cost is.

Fusion solves a minor problem of fission while making the major problem worse. It's a great example of misplaced focus.

Cooling down the magnets of a fusion plant is not a quick process, btw. Hell, just getting a fusion plant built with RAFM steel up to operating temperature is a delicate process. Once irradiated, the temperature window that material can operate at is rather narrow (300 C to 550 C; below that it's too brittle and above that it creeps.)

Exactly, capital cost of large civil engineering project and the financing of that is actually the driver of the cost.

If you can build a nuclear plant with 500MW output and reduce the amount of concrete needed by significant factor, all of a sudden your are quite competitive.

Just for comparison, this is one of the furthest along commercial GenIV projects designs:

https://www.youtube.com/watch?v=1pZC_ajI8Hs

No massive cooling towers. No gigantic containment cylinder. No gigantic low-pressure steam turbine.

In a PWR the core containment vessel is incredibly expensive and can only be built a few places, but its actually not the cost driver. Even so, compare to those, modern GenIV core vessels are vastly cheaper.

Not having to weld 8 inch gigantic steel vessels is far easier. See:

https://i.ytimg.com/vi/OgTgV3Kq49U/maxresdefault.jpg

And, not building them at all is easier yet. And, massively cheaper.

> So, to match power out, it has to be big. Really, really big.

Fusion has all of the characteristics of something you'll want to do if you are in space. Add to that the fact that its fuel is much easier to acquire than fission and that solar isn't available everywhere and you have a winner.

Meanwhile down to Earth, yes, most likely it will never be an important energy source.

For use in space, you need a reactor whose energy does not come out as hot neutrons. Even gamma rays would be better.

D-He3 (deuterium and synthetic helium) is attractive, except you have to make the He3 by making tritium and waiting for it to decay. pB (hydrogen and boron) is attractive, but much harder to make work than D-He3.

There is work on a D-He3 rocket, tagged "Princeton FRC", that they hope to loft for a trial maybe in 2035. Elon could get it tried out earlier, and ought to. We should be cranking tritium production up immediately so there will be enough He3 when we need that.

I must strongly disagree, particularly for DT fusion.

DT fusion reactors would be very heavy. ARC, for example, would have the weight of a couple of US WW2 destroyers. DT fusion delivers its energy as heat. A fission reactor would be far smaller, simpler, cheaper, and more reliable.

You might make the case for advanced fuels, but that likely means you have to get 3He from somewhere. I suspect beamed power will beat even advanced fuel fusion in all missions in the future (a weak prediction as I will not be around to see.)

But you dont need a lot of fuel for fission and you also need many other things to keep a ship/station running. Isnt fuel in the ballpark of just one of many consumables that you need?

Fission-powered ion propulsion is the best mature tech we have.

In space, the fission doesn't need containment, or even much shielding if it is far enough away from the rest of the ship.

In the cloud tops of Titan, Venus, Saturn, Uranus, or Neptune, a nuclear power plant may be as simple as a naked pile suspended near the bottom of a big fabric tube with wind turbines at the top. The radioactively-heated, buoyant air inside holds it up, and, escaping through the turbine, generates power. (Enough is held back so it stays up.)

Perhaps surprisingly, gravity is almost Earth-normal on all four planets. Perhaps more surprisingly, the temperature at a not unreasonably-high pressure is also close to Earth-normal (except on Titan). But getting back to orbit from any of the last three would be very, very difficult, so such a reactor would only ever power robots.

Getting back to orbit from Titan is easy. From Venus, it is pretty hard; and once in orbit you would need to refuel to get anywhere else.

Titan is perhaps the best place in the solar system for nuclear energy. An open Brayton cycle system on Titan wouldn't even require a very hot reactor, when the inlet gas is dense nitrogen (mostly) at around 94 K.

I find it hard to believe that people would be seriously researching this if we already knew there could never be any sensible commercial application.

More likely we just don’t know the way to make it viable yet.

In some hypothetically possible world, policy choices are based the best reasoning from the latest information. We, alas, live in this world, where policy is mainly a product of history. The H bomb legacy looms large here.

We are in the position of the person whose car keys were lost somewhere, but we are looking only under the D-T fusion streetlight because we can't see anywhere else.