[ThatPlayer]

Peak is not usually midday; peak is in in the evening when people get home from work. So in places like California and Hawaii, you get the opposite problem where the solar drops off right before peak demand and you have to ramp up other generators to make up for it. It's called the duck curve problem: https://en.wikipedia.org/wiki/Duck_curve

The United States electric grid data is freely available and pretty neat: https://www.eia.gov/electricity/gridmonitor/dashboard/electr... Choose a grid or a state to get regional time and you can see that region's peak will usually be 4-7pm. You can even see that weekend peaks are a bit lower, and that there's a second peak at ~10am when people get to work.

[epistasis]

My sentence was "and without solar that peak is midday in most places." Remove the solar and California has a huge midday peak. Watch it shift over the years into the evening as more residential solar was added to the California grid:

https://www.caiso.com/Documents/CaliforniaISOPeakLoadHistory...

(Note also in your visualisation that all times are Eastern and should be adjusted for different localities. And if you go to a summer week rather than a winter week, you'll find the true peak, which is much higher, and which has a pretty standard curve with a peak that overlaps sunlight hours.)

[ThatPlayer]

That's why I said choose a region to get regional time. California's chart shows the time in Pacific time: https://www.eia.gov/electricity/gridmonitor/dashboard/electr...

They have this summer's data too, though no way to link directly, and it still peaks at ~7pm: https://i.imgur.com/16mssuH.png . Using the 16th as an example, a peak demand of 44,008 megawatthours @ 20H PDT. Comparing that to their generation graphs, which you can separate into sources, like solar. On the 16th, peak solar generation is at 11 @ 13,201 megawatthours. By 6PM, it's down to 853 megawatthours. By peak time, it's nothing. My own residential solar matches that curve on that date.