The idea of natural resources running out is always pushed about everything and is always wrong.
However, in this case its wrong on an even more fundamental level. There is a gigantic amount of fertile (i.e. you need to hit it with one neutron to make it fissile) nuclear material around. Literally enough for 1000s of years. And volcanic activity is bringing up new such material from deeper in the earth.
Any non insane nuclear strategy will eventually switch to breeder reactors and if you do, the amount of fissile material is essentially infinite.
We already have gigantic amounts of fissile material in the form of spent fuel that we can use to start breeder reactors.
This video goes threw how you can do this:
https://www.youtube.com/watch?v=bkZ0ZBAVDu4
> other fissile material to use (thorium?)
There is both a huge amount of fertile material, both in the form of thorium (T-232) and what is usually called natural uranium (U-238).
The things consider limited is Uranium-235 if I'm not mistaken. This is the only surviving naturally fissile material. But should it get down to it, I'm sure we could extend the amount of known reserves, just as we did for almost everything else.
There is a pretty interesting story here about 3 different strains of nuclear that exists and of these the naturally fissile material of 2 of them is 'extinct'. Only U-235 is still sticking around. If not for that, it would have been much harder to start nuclear age.
In the beginning there was a lot of effort put into breeder reactors because the amount of known U-235 was considered really small. So breeders had to be developed however as usually happens more U-235 was discovered once people started looking. And this sadly lead to way less emphasis being put on breeders.
> there's still the constant need of water for cooling
This is wrong. Almost any GenIV reactor is perfectly fine with being air cooled. In fact most reactors are perfectly fine in the icy tundra and the hottest desert on the planet.
Consider that for a reactor that runs on 600C even a desert feels like being in an ice bath.
As with everything nuclear, the real problem is that we stopped progressing in the early 1970s and are still using blown up submarine reactors.
The real problem is not nuclear, but technology stagnation because of societal issues.
> nuclear proliferation
An over estimate problem. Depending on the design you can make it practically impossible to do anything. And if you can do anything it would be about 100x more effort then just developing the bomb the same way everybody else has.
There are absolutely nuclear designs that you can deploy in most places in the world.
> the small risk of malfunctions
While true, the same can be said for wind and solar. Even current nuclear doesn't kill people.
Future GenIV reactors potential for harm is very local. You could potentially imagine some accident that kills some people inside of the power plant.
But unless you combine the story of 3 evil bond villains its hard to see how with such reactors you could seriously harm people outside of the plant boundary.
So I would say in terms of risk, building 1000s of large towers to harness the wind, this is much less dangerous.
> or maybe deploy fusion if we ever finally crack that egg
Practically it makes no sense and will not be cheaper then fission for a whole host of reasons.
Also, it still has proliferation concerns and still has the potential of accidents that could leak radiation.
Fusion might make sense for a few space application, but practically for earth is mostly irrelevant.