ET, I think the only available course is to economise, hard and fast.
The IPCC says we have ten years to halve emissions in order to avoid the worst. Well, ten years is about how long it takes to build a nuclear power station these days. The Experimental Breeder Reactor at Dounraey was indeed built in about three and a half years, though that excludes planning and commissioning. If we call it five years, that’s still half of our available time, leaving the entire CO2 saving to be made in the second half.
Next we have to consider scale. There are currently over 400 nuclear power stations worldwide, producing about 10% of all electricity. These have been built over the course of about four decades. OK, maybe we can build some power stations in five years, but globally we’d need over four thousand just to address current electricity generation. A lot of the construction is highly specialised, so we’d need to train people in the techniques – more delay.
Then we need to consider fuel. Both uranium mining and uranium enrichment would need to be expanded by a factor of ten. Enrichment in particular is even more specialised.
Is this still looking practical?
Enrichment produces tailings of depleted uranium, in the form of depleted uranium hexafluoride. It’s corrosive, highly soluble, highly toxic chemically, and mustn’t get into the groundwater. The photo on the left is just one of three DUF6 storage yards in the USA alone; note the older cylinders going black from corrosion:
Having done all this, we’d be set to exhaust conventional uranium supplies in around a decade, so concurrently we’d have to have built a similar quantity of renewables infrastructure as well, in any case.
I see great promise in the molten salt reactors, breeding fuel from thorium, and fast spectrum reactors that can cook down our existing spent fuel, releasing centuries’ worth of energy. But they aren’t even properly prototyped yet. There simply isn’t time to go this route.