Overbuilt solar and wind generation should get us to 99%+ of scenarios.
The last 2 days in CA were edge cases that will need special attention. A vast high pressure system parked itself over the state creating both record setting temperatures and virtually no wind. On top of that, smoke reduced solar production markedly. Next, our neighboring states were also undergoing the same heat wave so had less power to export to us precisely when we needed it most. So in that scenario, within 2 hours of sunset spells trouble without additional resources.
Enough batteries could deal with this situation but at current and near future costs that looks unrealistic.
On the same last 2 days there was plenty of wind power to harvest in the midwest and Texas, but how to get it here. Massive DC interconnects would take care of this. That might be a silly waste of resources for the <1% edge case days alone, but DC interconnects would pay dividends with their facilitation of regional grid efficiency gains. Each regional grid would need fewer generational facilities as a result.
[I'm replying in the power-to-methane thread because this gets into why we need to plan for days, months, or even years of reduced solar. I thought it would be a little too off-topic for the Solar PV News thread.]
Today is interesting too, at least here in the SF Bay Area. So far today my rooftop solar has produced zilch, zero, nada — because a thick pall of smoke is blocking almost all sunlight. IMO we need to plan for edge cases like this, but lasting months, or even years. Consider 1816: big impacts on agriculture of course, so we have to consider food supply; but in a decarbonized economy we also need to consider what to do about energy during a long period of reduced solar.
Year Without a Summer - Wikipedia
I think power-to-methane will be part of the answer, but there's also a case to be made for power-to-h2.