> The drop off after cleaning could well be because some of the initial boost could come from the water cooling off the panels: solar panels become less efficient at higher temperatures so the drop of after they're all cleaned could be them warming up again after the water has evaporated.

This was almost certainly the case. AIUI, that cooling still makes a large difference.

Makes me wonder if it is worth it having a small esp32 or similar control a sprinkler system for the panels. Keep them cool during high ambient temps by spraying a litre of water on each panel every 30m.

An idea I had a while back (when I was considering panels) was running my mains water through the back of each panel (before the mains got to the hose) in a sort of radiator pattern.

People make the same claims about cleaning their air conditioner. Yeah it’s blowing cooler inside: you’ve just chilled the condenser and left some residual evaporative cooling.

Yeah, came here to say this. There are a lot of things that will affect efficiency and how clean your panels are is only one of them. How dirty your panels get is also highly sensitive to both location and installation. All of this gets horribly confounded by string inverters (especially if they're clipping the peak, which they often are).

I easily see 5% increases in efficiency after cleaning (with micro inverters), but also see ~3% increase in output on a hot day from just spraying the panels with water to cool them off, which goes away in 15-20min.

If you're going to clean once a year, do it after the "last" rain of the year before summer peak solar output. If you live somewhere rural with summer thunderstorms, with high angle panels, and without a lot of pollen (or dust storms), you may not need to clean them much at all.

If you're going to climb on a roof, do it safely! A 10' fall can kill or maim you.