Production Curve No Longer Smooth Since October

[skepticcyclist]

Since around October my solar production curve started looking more like a turkey tail than a nice smooth normalized curve on perfectly clear days with no shading at all. Prior the curve was alway smooth.

I have enphase iq7 microinverters. When I check production for each microinverter it appears to be random. Different microinverters dropping production by a few tens of watts for a few minutes to a few seconds once (sometimes) twice a day. This only happens when production is ramping up or maxed out. It never happens as production is decreasing later in the day.

Here are the last 3 perfectly clear cloud free days. Ignore the portion after 1:00pm. That’s from my palm trees to the SW shading half the panels which only occurs during the 3 lowest sun angle months. It’s the feather like appearance near peak.

It seems to match a software upgrade on the Tesla gateway. Maybe just correlation.

 

[skepticcyclist]

For comparison, here’s a sunny day from December of last year.

 

[skepticcyclist]

Interesting is that my total daily production as well as peak production has increased when I compare similar perfectly clear days from last year to this year.

 

[Southpasfan]

There have been a couple of days in the last week where there have been some really sparse clouds in SoCal, just enough to cause that pattern for me.

I would have guessed some shading but it sounds like you ruled that out.

 

[skepticcyclist]

There have been a couple of days in the last week where there have been some really sparse clouds in SoCal, just enough to cause that pattern for me.

I would have guessed some shading but it sounds like you ruled that out.

It is not shading at all. Trust me. Not only am I observing it directly as it is occurring, it goes all the way back to October, and I went back and looked at production from each microinverter.

During the minor drops there is not a single cloud in the sky, not even a light high cloud. Plus with clouding you will occasionally get an edge of cloud effect which will produce a little peak higher than the rest of the curve.

And if it was shading from clouds then every microinverter would would drop production at the same time. Most often it is just 1 or 2 that drop production and each of those little drops occurs from a different 1 or 2 miroinverters.

 

[skepticcyclist]

For example, today does have high clouds. I drew an arrow to the peak prior to the clouds moving in.