This curve (data recorded every 15 minutes) shows typical summer and fall performance. The late afternoon "wing" is presumably just from skylight on the panels since the sun has moved so far to the northwest by late afternoon that it is "behind" the panels. Solar noon at our longitude is almost exactly 12:00 – shifted to 13:00 during daylight savings time. The midsummer curve peaks right at 13:00. The Oct 1 curve shows a little less peak power, as expected. More significant for the daily kWh production is that the curve is narrower as the days get shorter. Interestingly, the peak has shifted to about 12:00 because the panels face about 20° east of due south, causing the sun to be most perpendicular to the panels a little before true solar noon. At midsummer, with the sun very high, this was irrelevant. As more data become available, I'll display a typical curve for each quarter of the year. This chart shows monthly kWh generated by the panels, monthly kWh used in the house, and the difference (the green bar). A positive difference (green bar) means we generated more electric energy than we used; a negative difference that we used more than we generated. Our goal is to be electrically neutral over the course of a year, having enough positive net production during 8 or more months a year to offset the heavier use and decreased generation of winter months. Note: These are PG&E "billing months," which are offset from actual months by 3 or 4 days and don't correspond exactly to the length of the real month. Thus the values shown as "kWh generated" don't exactly match the values on the archived monthly data. December saw a huge increase in useage, primarily from electric baseboard heater used for supplemental heat. The goal of being electrically neutral for the year is in doubt