SanDust said:
My guess is that production drops in AZ during the summer because of increased heat. I cleaned my PV panels last year on the theory that they were dirty. Production jumped during the cleaning as the water cooled the panels but then dropped down to its baseline. >>>>>>>>>
AFAIK microinverters aren't more efficient than a central inverter. It's just that as a distributed system they're more robust. Claims of greater efficiency usually refer to the life of the system and situations where there are shading or other issues with one panel in a string, which brings down the production of the entire string.
Temperature makes a substantial difference. I like to hose down my panels at least once/week during the peak production of the day, to take maximum advantage of the temporary cooling you mention. Today before/after my hose-down, production went from 2465 -> 2740 watts, a gain of 11%. Within 20 min or so production returns almost to its previous level, showing that there was only a modest amount of dust degradation.
I have had my array operational 7 months, including one Winter and one Summer solstice. The maximum production of 17 KWhrs has occurred 4-5 times in that interval, always on a cooler, clear day. Once such day was in July, the rest in May.
Shading loss is more significant for me. I am unwilling to cut down the trees whose shade starts to reach some of the panels after 3:30 PM this time of year. Each string of a solar array has all of its panels wired in series. I have a single-string system, so full shade on even one cell of one panel can shutdown the full array. However, the by-pass diodes in my panels usually mitigate this problem by passing the current around the shaded sections. I chose a single-string system to take maximum advantage of this effect. My panels are divided into 3 sections, with one diode by-passing each section. Small changes in shade pattern can make a big difference.
You can see whether your by-pass diodes are effective under a particular partial shade condition by looking at your inverter's output as the shade moves in. If the the total voltage goes up while the current drops precipitously, the panels still in sunlight are under-loaded and the diodes are not activated. If the total voltage drops while the current level is maintained, this means that the diodes are active, passing the current from the illuminated panels around the shaded panels. In this case the inverter has been able to track the voltage down to a new local optimum.
Since I am driving the Leaf more than I projected, I am considering adding a few more panels. Shade in the new location is again an issue, so I may go with micro-inverters this time.