Current surges mainly, and of course that means power surges as well. Let's say on a normal day with full sun, X array makes about 34 amps. On cloudy days with the sun in and out behind clouds. As the sun comes out behind the clouds you can see currents in the mid to upper 40's as you hit full sun once it emerges.. So yes it is significant. On very cold clear days, current may approach 40 so this is still a bit more than just 'cold panel current'.

I was originally wondering if it had to do with temperature of the panels, so looked at past records of average temps .vs output seen and there does not seem to be much correlation. in other words a panel bathing in the sun all day might hit 120 degrees or so, a cloudy day they'd be more say 80 degrees because the thermal radiation is not hitting them constantly. So I was looking at, what happens on a full sunny day when it's say 30 degrees out, so the panel temp is only around the 80's, and also looking at the manufacturers derating charts for temperature derating, and it still does not add up to that much extra power. (amps)

Perhaps im missing something but it seems that maybe the light filtering through the fringe of the clouds it does something to it to make it more 'reactive' to the cell? I don't know if polarizing would be the best word to use, but filtering maybe? I am not sure.

If it is something to do with the light, that peaked my curiosity of, well what if we put a thin film filter on the panel to 'simulate' that kind of 'whatever' is going on when it ducks around clouds, and if that is possible, would it still yield enough additional energy to make it worth while, since you'd be passing the light through an additional medium, of course some losses would occur.

If this is something that happens and you guys working with plastic, perhaps as an interim step, make a film that you put over conventional cells, or ultimately design your plastic cells with this 'built in' for maximum efficiency. Then again, i could just be talking out my backside too :D

Aaron

We may have come up with a way to minimize the waste through the validation process. It may take a month or two to implement, but we’ll keep you posted on how it is going.

Every improvent is welcome :)

Something light for the weekend:
http://xkcd.com/676/

There has been an interesting, ongoing discussion on the pros/cons of the 12h time limit for WCG calcs.
http://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,30547