[osmala]

Yes. And for renewables with battery backup would require 375 GwH of storage to get a reasonable two week minimum to be considered reliable backup for a equal electicity producion. You can calculate easily 24 * 14 * 1.117 Its not about just how much long sun is up, its about snowstorms and ice.

[jerven]

That is an unreasonable backup requirement. It is not a backup requirement for any single nuclear power plant either. Those can be taken out by snow storms too (single pylon failure on a line). Or just a refueling stop, which are quite common for nuclear power. One assumes grid resiliency through Geo-distribution and recovery. Any snow-storm taking all solar down for two whole weeks would take down any other power source as that assumes frozen rain followed by massive snow over a huge area. Such a snow storm would kill anything generating power as well as taking down massive parts of the grid. i.e. not any snow storm seen in the last 100 years would be sufficient for this.

In the end cleaning snow, while a pain in the ass, is not a real challenge. Much like desert sand cleaning, this is automatable (and small snow removing robots exist already) and would be worth the investment in northern (or extreme) southern climates.

We are basically talking about plant equivalence. You need 3x nameplate capacity and about 12 hours of battery storage for equivalency between a solar and a nuclear solution. Assuming night load is equal to day load. That assumption does not hold for a grid, night load is normally half the day load. So a system must only survive the evening peak plus the normal night load.

The quip about battery storage price basically showed that if you throw as much money at it as a single nuclear power plant then yes it is a solved problem today.