The trouble is they're both intermittent, even independent of time of day. For solar that's much less trouble because the demand is higher during the day, and aligns extremely well with air conditioning load in the summer.
But if you're relying on wind at night and then there isn't any, and you also have no solar because it's night, what's left?
> Generated fuels, like hydrogen, may work well for long-term storage and we'll them for other things.
At which point you have to add the cost of production, storage and generation facilities for some other generating technology.
> Third, we can overbuild solar and wind. It might be cheaper to make 3x or 5x than needed.
But how does that fix it? Sometimes it's calm for weeks, so your wind turbines are generating at 5% capacity for that long. Are you going to overbuild by 20x? Or build enough storage to power the entire grid for that long, even if you only use it for two weeks every three or four years?
> Finally, we are going to need extra energy for carbon capture and generating fuels.
This is a generic argument for building more of any kind of non-carbon generating capacity.
Overbuilding means that can use solar and wind more of the time. Partly because wind and solar are anti-correlated, partly because you build them in different spots, and partly because there is some energy being generated. 3x means that only need a couple weeks per year need backup. 5x means that couple days per year need storage.
The generated fuel production would already exist because need them as fuel. Repurposing natural gas storage should work for hydrogen and there is lots of that available. The power plants would need to sit around but should be able to convert natural gas to hydrogen.
> 5x means that couple days per year need storage.
But that's the problem. If you need the long-term storage at all because you have a period with ~0 output from renewables, you then need to maintain enough generating capacity to run the whole grid from something else during that period, even if you only use it once a decade. Which is enormously expensive.
> But night time usage is small, and things like charging cars can be delayed.
The night time usage is about half the peak daytime usage during the summer, and a significantly higher proportion in the winter. And it will get even higher if we switch to electric heating from fossil fuels.
You already want to be charging cars during the day anyway because solar is cheaper, but none of that is included in the existing numbers because the current number of electric cars isn't a major proportion of power consumption. We could perfectly well charge electric cars entirely from solar, but you still need to handle the existing nighttime load with something.