Originally Posted by Iceni John
Would it therefore be better to have a greater quantity of lower-power panels, instead of fewer high-power panels? This way, if one panel stops producing power due to shadow / bird poop / Act of God, will there be less overall net loss of output from the entire system? Is this sensible, or just crazy-talk? (Yes, I know it may be cheaper to have fewer larger panels instead of many smaller ones.)
The paragraph you quoted under my name was all a direct quote from Handybob. I only added the bold. As I mentioned in Tesla's thread, all panels are made up of individual cells added together in series or series-parallel, and the shading problem comes from shutting down one or more of those cells.
I don't know if there is a one-size-fits-all answer to your question. The idea is to shut down none or as few individual cells as possible, and if they must get shade, try to get all the shaded cells on the same panel so you don't lose the whole system. The answer probably depends on the real estate for mounting the panels, and any fixed obstructions.
I have had some thoughts on the subject. If a single panel, not in parallel with others, due to its design having no bypass route shuts down when any cell is shaded , there is not much that can be done to recover the remaining power captured in that panel. If the panel merely bypasses the shaded cell(s) and drops the output voltage to a lower value, protection diodes might cause shut-down due to "back-pressure" from fully-lit parallel panels at higher voltage. For the second case, I have toyed with a few untested
1. Get a high-voltage input MPPT controller, connect all the panels in series, and have the controller adjust to the available power in the same way it would in the morning and evening when the sun is not full. This would not get maximum efficiency from the controller compared to lower voltage parallel operation, but if it prevents shutting down one or more whole panels, it may make up for the conversion loss.
2. Have each panel run its own MPPT controller, let the inputs vary as they self-adjust to the available panel power, and combine the equalized parallel outputs to charge the battery bank. I would not try this unless I had controllers that communicated with each other about battery condition and temperature. You do not want them mistaking a parallel controller for a fully-charged battery, and throttling back the power too soon. There was one manufacturer whose design for home systems was/is to have a mini DC to 120 VAC inverter on each panel, and combine all the AC outputs in phase. The idea was to eliminate the big DC cables with their ampacity losses. That is what gave me the (expensive) parallel controller idea.
I can't give better answers until I have some spare equipment to play with. There is one solution that is proven to work, though.
Move the panels, the obstructions, or the bus so all the panels stay in sunlight. Duh!