The time has finally come for geothermal energy

The US Navy consistently builds reactors on-time and in-budget

There’s something to be said for a standardised design with replaceable parts.

The US and Russian Navies deciding to remain mostly petroleum-fueled is one of the strongest arguments against nuclear becoming very cheap: surely they would do it if it wasn't ruinously expensive, because it eliminates the national security risk of a petroleum blockade and simplifies at-sea logistics immediately.

Don't presume too much about the US Navy's fleet decisions. Using that same logic you could presume that smaller, aged and poorly maintained fleets are advantageous for naval supremacy since that appears to be the choice of the US Navy for a couple generations now.

Or you could presume that the complete inability to build a merchant marine fleet was also a strategic advantage!

I don't know much about militaries or nuclear reactors, but I know that reactors are used in some submarines and in some aircraft carriers -- situations where you want a vessel to to remain at sea for long periods of time without refueling, and weight is not a primary concern.

That's pretty niche, though. Think about trucks, tanks, aircraft, generators for outposts, etc. It might be cool if you could safely package a zillion nuclear reactors for those use cases, Terminator style, but I'd guess that reactors are a better fit for centralized, permanent power generation.

It's not just the US Navy. It's also the Russian Navy, the French Navy, the Chinese navy of the PLA, the British Navy, the Indian Navy. If nuclear power were cheaper than oil, or anything other than much more expensive, at least one of those would have gone all-nuclear.

The Aircraft Reactor Experiment https://en.wikipedia.org/wiki/Aircraft_Reactor_Experiment, yielding 2.5 megawatts, was about two meters tall and one meter in diameter; the fuel was 15kg of U-235, but I think the reactor as a whole may have weighed several hundred kg. (It couldn't have been more than about 40 tonnes, just because no material would be dense enough, but I think it was much lighter than that.

Oh, here it is: https://media.githubusercontent.com/media/openmsr/msr-archiv... page 24/139 budgets 48400 "pounds" for "reactor and reactor shield", which is 22 tonnes, about 110kW/tonne.

The smallest nuclear submarine was NR-1, which had a total displacement of 400 "tons": https://en.wikipedia.org/wiki/American_submarine_NR-1 so the reactor must have weighed less than that.

The 10MW version of SSTAR was supposed to weigh 200 tonnes, 50kW/tonne, while the 100MW version was 500 tonnes, thud 200kW/tonne: https://en.wikipedia.org/wiki/Small,_sealed,_transportable,_...

A 4.95-kg americium thermal reactor design outline has been published: https://www.researchgate.net/publication/239521070_The_Small...

Arleigh Burke class destroyers have 80MW of propulsion and displace 8300 to 9700 tons, compared to which the SSTAR's 500 tonnes is almost insignificant. So weight isn't an issue for ships, and weight doesn't get ridiculously high until you're down below the megawatt scale.

So, you may be right that existing proven reactors won't scale down to a single truck or tank. There isn't a known physical reason it's impossible, or even impossible to do safely, but it hasn't been achieved.

Probably you are right that many small reactors would be more dangerous, but warships exist so that they can go into dangerous situations. You have to weigh the risk of a reactor problem against the risk of being unable to fight because you have no fuel. And we've certainly seen that many militaries have little concern for sailor safety.

Despite all this, no navy has switched all their ships to nuclear fuel. The only explanation I can come up with is that it's unsustainably expensive.

The US Navy doesn't build them, private companies do. None of the SMR companies that I am aware of have attempted to use designs from naval reactors. Perhaps it has something to do with the fact that they use highly enriched fuel which is a no-no for civilian reactors.

In any event, if your more general point is that naval reactors indicate its possible that SMRs can be built on time and budget, I don't disagree. It certainly is possible. NuScale was the furthest along of the SMR companies and had their project implode before it got off the ground. So, its also possible for them to crash and burn.

So, I'll stick with my classification of SMRs as "maybe it will work out, maybe not".