LOL.. I just lost a post I worked on and off for two hours but apparently timed out writing..
Please pardon my brevity as I rewrite the salient points...
Solar panels use Schottky bypass diodes with a forward voltage drop of .15 to .45 volts.. I would conjecture the losses aren't too significant in many
installations -- In this example panel inexpensive grid tie manufactured panel (of which i use two 240watt versions) Vikram Solar Pvt. Ltd. | 60P Prime Series | Solar Panel Datasheet | ENF Panel Directory
two panels in series, 1 completely shaded, would result in a voltage drop of (absolute max) 1.35 volts through its 3 bypass diodes. At a NOCT(normal operating cell temperature) maximum production of 180.24 watts, the 1.35 drop from the bypassed panel would net 171.48 watts, a loss of only 5%.
In my installation
I switched from parallel to series because
at elevated cell temperatures the maximum power point voltage would fall below my LiFeP04's set float voltage of 27.2v. I would still charge, but lower and lower current the warmer the cells were.
I wound up netting more energy per day in my situation, primarily due to several factors --
1, due to the higher operating voltages the charge controller would wake up and produce usable power much earlier in the day and would go to sleep much later. A couple hours worth of extra production at less than max insolation..
2, the higher voltage produced more usable power in cloudy days, where the paralleled panels' max power point would quickly drop below my charge profile voltages.
3, my panels are oriented so the mast
could only shade a maximum of 2 out of the 3 sections on one panel. So in my particular case, in a partial shade situation I am producing at least 30% more power in series, even accounting for bypass diode voltage drop, than in a parallel array, where partial shading would immediately drop the entire partially shaded panel below my charge profile minimum and render it net zero production.
Added benefit: The fact my MPPT controller is switching lower currrent along the long run (read: antenna) to the solar panels creates MUCH less RF noise
in and around my boat. My VHF
doesn't break squelch randomly like it used to. I am an extra class amateur radio
operator and care much about these things..
Devil's advocate may state that the higher your array voltage from your battery
voltage, the less efficiently your mppt charger
will operate -- but -- we're talking percentage points here.
And of course, everything is percentage points here...
I am not saying series is the answer or parallel is the answer -- truely, I think that a "general consensus" is a bad idea here. We all have different operating temperatures, shading considerations, battery system charge profiles, panel specifications, installation
considerations, etc. I just want to offer up additional considerations to the subject that many people are probably not aware of -- it is important to read the data sheets
on your solar panels and carefully consider both temperature effects and shading effects on arrays and net energy production before we commit one way or another. YMMV.
And anyways, doing some empirical testing on both methods isn't that hard as long as you take that into consideration during the installation.