tbleakne said:
(1) So would your panels underperform in clear cold mountain weather? (22-19 %)/(.38-.30 %/C) = delta 37 C, 67.5F. STC is 25 C, so the temperature would have to fall to 25 - 37 = -12 C before your panel efficiency would drop down to LG panels.
Actually, the resources I've found seem to indicate that PV efficiency continues to improve as the temperature drops below STC. For example: http://www.sciencedirect.com/science/article/pii/S2090447913000403
In that article, there is a graph that shows significantly greater power at 0 C compared to 25 C. Other resources suggest the same thing. Reportedly, PV panels in Antarctica have done quite well in terms of efficiency, when the sun is visible that is.
However, as our panels have a smaller power temperature coefficient (-.30%/C) than yours, I'd expect that their efficiency advantage at low temperatures may not be as great. That said, particularly as panels can become much warmer than the ambient air temperature, I admit that this should be only a relatively trivial concern for us - the smaller coefficient should do us more good than harm.
tbleakne said:
2. Yes, if air can circulate underneath your roof panels, your roof will be cooler. I am expecting this benefit as well for my new installation, which sits over my attic and my office and bedroom. You might also consider a small mini-split system if you decide to add active cooling powered by your solar production.
Yes, there is some room for air to circulate under our panels, and on warmer days with intense sunlight, I think we are already noticing a cooling effect upstairs.
Thanks for the tip to consider a mini-split system, something I hadn't previously investigated. As these systems utilize heat pumps and can perform both heating and cooling, this might fit into my longer term goal of reducing and ultimately eliminating the use of natural gas for home heating (and cooking).
tbleakne said:
3. Higher production during partial clouds: this is a little surprising. In clear weather the trees shading your roof will cast strong shadows with moving patterns of shade and full illumination. Partial clouds can reduce the contrast between illumination and shade.
Most panels are organized into 3 subpanels each with one bypass diode. Dark shade blocking any one of the 20 cells in one subpanel will shutdown all output of the other cells in series in that subpanel. Modern string inverters are good at adjusting the string voltage such that a single cell will shutdown only one subpanel.
I believe your Sunpower panels have an integrated micro inverter in each panel, with 3 subpanels in series. It is possible that the micro inverters do not operate well if 2 of the 3 subpanels have shade because the panel voltage is too low.
With respect to clouds, the best scenario occurs when the sun is not blocked at all and clouds elsewhere in sky reflect the sun's light onto our roof. But even a completely overcast afternoon (starting at 2:30 PDT or so when the roof is mostly shaded) seems to beat a sunny afternoon in terms of our production.
Yes, our panels do have integrated micro-inverters. I may at some point try taking photos of our panels in partial shade and comparing with our panel-level power production, to help me better understand what's going on.
Finally, something else to note is that, according to our installer (LA Solar Group), we may be able to install additional panels by utilizing our north-facing roof and employing a "reverse tilt" to the south. For a single row of panels not too far below the ridge (we'd need a three-foot offset below the ridge for code compliance), it seems to me that this could work well. Our roof is scarcely visible from the street, so aesthetics aren't too big a concern up there. Depending how things go, this could be an option to consider next year.