The most optimistic estimates by the commercial companies working in this space is the first commercial reactors coming online around ~2050. It takes time to build these reactors and integrate them into the existing power systems. I may have misstated my position which is that I don't think fusion will make up any meaningful part of power by the end of the century. The reason is a) requires fundamental engineering and materials breakthroughs b) funding for it is a pitiful amount compared to what it probably needs c) it's competing with other cheaper energy sources d) we don't have any net positive energy reactors yet.
The basis of that prediction is just looking at how long it took fission to take off and there it was being driven by military needs (submarines) and not civilian.
> and there it was being driven by military needs (submarines) and not civilian.
I'm surprised that support for fusion isn't part of US DoD policy; abundant, cheap, clean energy would be a huge military asset. Those explosives zooming back and forth in Ukraine are really just energy balls.
Fusion would in no possible world be cheap. Whatever else they would be, fusion reactors will be the most expensive structures per unit of volume built by man.
Pardon me if I'm talking out of my ass, but this is what I tought:
* The main input to a fusion reactor is electrical power.
* The main output of a fusion reactor is even more power.
So that seems to me to mean that once you've incurred the capital cost, the machine produces free energy. My reasoning sounds naive and simplistic, because I don't know what I'm talking about. But what's wrong with my reasoning?
1. A fusion plant would be extremely expensive to build. The most optimistic estimates are 10x per GW vs. fission. But we really don't know how to build one; nobody has identified a material that would work.
2. It is extremely expensive to operate. The most optimistic estimates are >10x fission. A thousand tons of lithium "blanket" would have to be sifted daily to get a few grams of tritium for the next day's operation. Nobody knows how this could be done.
3. It destroys itself with neutron irradiation in only a few years. At best, major parts of the structure would have to be replaced using robots because of the extreme radiation in the parts being replaced. Similarly, for repairs. Nobody has built such robots, so they are custom one-offs.
The fuel cost of fission is a negligible part of its cost. The fuel cost of a fusion plant would be negligible, assuming enough tritium could be obtained at all. The ITER project expects to run out and does not know where they will get enough for future experimentation.
The necessarily super-expensive fusion would be a strict liability. It will, in consequence, not be built. All work toward that is pure waste.
The DoD does spend a great deal of our tax money on fusion, but not for "abundant cheap clean energy". It is, rather, for uses like what is occurring in Ukraine: death and destruction; and for vaporizing cities and ports, distributing radioactive fallout over wide areas.
As a sort of representative summary of the situation, the ITER project has been in progress since 1988 - 34 years - and hasn't yet achieved fusion for more than about 5 seconds.
On top of that, by design, it will never be net energy positive - the tens of billions of dollars and decades spent on it are purely to produce a proof of concept for sustaining a fusion reaction. Turning that into something that can actually act as a source of energy is currently still at the "unsolved problem" stage. We can't even write a project plan for how a viable reactor could be developed, because we simply don't know.
As such, there's no realistic scenario in which this situation somehow turns into viable fusion power before the end of the century, short of an unexpected major breakthrough. While such a breakthrough is conceivable, it's not something you can base a strategy on. The responsible position is to recognize that we can't rely on anything useful happening in fusion in this century.
One other point about the "think back to 1942" comment is that extrapolation only works when you have relevant data points to extrapolate. The technological advances since 1942 have been nothing like the advances needed to exploit fusion.
Fusion involves literally recreating the conditions in the heart of a star, but without the enormous mass of a star to provide the necessary pressure. There's no precedent for this in our technological history. Nuclear fission was trivial by comparison - nature does all the real work, all you have to do is arrange the fissile material appropriately. That's not the case for fusion.
It's an incredible achievement that we can produce a fusion reaction for 5 seconds, but there's no guarantee that we're going to be able to turn this into something that can be sustained day in, day out, and that generates more energy than it consumes.
Sure, if you acknowledge that fusion in twenty years is also a "maybe". If you can't, then you're not handling your probability distributions properly.
You could say it for 800 A.D. or 1200 A.D. too, and yet 880 A.D or 1280 A.D. wouldn't look too different. In fact, you could say it for 300 B.C. and 1300 A.D. - a whole millenium away - would still be quite similar technologically.
Not all periods have 1942-2022 amount of innovation. In fact 1942-2022 is itself bimodal and front-loaded. By 1980 already most major innovation had already happened. The rest is mostly efficiency improvements and diminishing returns, but much much much less major inventions.
in 1942 there were no nuclear (fission) power plants, yet today there is probably one within driving distance of your house.
additionally there are natural fission nuclear reactors that have been going for possibly billions of years... and one big fusion one within visual range :)
It's ridiculously arrogant to make predictions like this.