But the premise was "solar is getting so cheap". Wind costs more than solar, and is still intermittent so you still need the backup.
> A simple cycle turbine power plant is maybe 1/20th the capex per W of a nuclear plant; a combined cycle plant maybe 1/10th. Combustion turbines are amazing.
That's just the capex for the plant itself. It doesn't include whatever you're using to produce and store the fuel. The capex to store a week's supply of hydrogen in particular would well exceed the capex for the plant itself. Plus the "operating costs" which largely still exist even if the plant is only operated 1% of the time. It still has to pay for full-time security and maintenance and have a full complement of staff on-call at all times in case the plant has to be brought into service.
And many of those costs go up rather than down as usage declines. Right now we have a network for the production and distribution of natural gas which gets funded by domestic use for heating and cooking and existing power plants can tie into those pipelines at minimal incremental cost. If we stop using fossil fuels for heating, now you have to justify the opex on all those pipelines just for the infrequently-used combustion power plants, or you have to build and maintain a different distribution system for the fuel.
The low cost thing to do would be to have have a production and distribution system that can replenish the week's worth of fuel over the course of three months, but now you can get a period of undersupply in November and the tanks won't be replenished until spring. Whereas if you want to be able to produce a week's supply of fuel the next week, add another billion+ dollars per GW of capacity for electrolyzers.
Even if this still costs less than a nuclear plant, this is the cost of the backup system by itself. You still have to add the cost of the renewable sources to be used the other 99% of the time. And the efficiency advantage if nuclear is used for heating via cogeneration.
> We will also see migration of thermal energy intensive industries to lower latitude places with more sun and less seasonality.
Industries are fine, the question is how to heat homes in Europe or the northern half of the US.
> these things cost nearly as much by themselves as the nuclear plants.
This is completely wrong. A simple cycle turbine power plant is maybe 1/20th the capex per W of a nuclear plant; a combined cycle plant maybe 1/10th. Combustion turbines are amazing.
We will also see migration of thermal energy intensive industries to lower latitude places with more sun and less seasonality. Diurnal storage of heat from PV is quite feasible. Sorry Europe, nuclear won't save you.
I will also note that 2/3rds of industrial heat demand is at < 300 C, which means heat pumps can be used (particularly if the process has a warm waste heat stream to feed into the heat pumps.)
https://www.nrel.gov/analysis/solar-industrial-process-heat....