That brings up a major undiscussed problem with renewables:
Installing batteries to supply during frequent source outages (i.e.: night) is obvious. That’s a given when installing a system.
The issue comes with statistical fluctuations: how many standard deviations of source outage do you want to cover ... when additional battery capacity, used with diminishing frequency, is no less expensive than the baseline.
16 hours of backup battery is a given (charged during a practical 8 hours of sun).
24 hours is a no-brainer while you’re at it.
48 hours is pretty much a given, covering cloudy days.
72 hours, yeah, storms may hang out for 3 days.
5 days of backup is getting pretty expensive.
A week of storage is expensive.
You probably won’t need more than a week of batteries ... but 1 hour past that and everything shuts down, so do you keep adding on the same blasted per-day battery cost for additional days of coverage which you probably won’t need but still may happen and will seem cheap when you do need it after all?
A homeowner running his own system can make the cogent decision support X days of battery backup, accepting cost of blackout vs cost of forestalling it.
A municipality faces a MUCH harder decision, what with 1 day of additional full-capacity battery storage costing enormous sums and yet being rarely used, vs increasing odds of prolonged blackout by not buying more storage even less used.
There’s also the issue of how long it takes to recharge those batteries on top of normal usage. Basically need to install an additional full day of solar panels just to get each day of battery storage filled in time for the next prolonged cloud cover.
In Florida, when there are clouds, air conditioning electricity demand is less.
Going 100% renewable would be All Out Crazy.
Batteries are very nasty too when they light off.