Morning power dip?

Questions
#1

Hi All,

Have you ever seen this behavior on your grid?

image
image940×410 37.2 KB

Checking the individual inverters, they all drop at the same time and stay off for 3ish minutes before coming back. This, to me, says that the MAINS had a voltage sag and turned off my grid. I dont see any evidence of a sag when I am in the house in the morning. I am tempted to put on a data logger on the line to see what I can see, but its a bit complicated to get it going.

This happens every weekday about the same time (within a minute or two).

Has anyone ever seen behavior like this? is this some sort of thing the power companies do as a diagnostic?

Cheers,
-Ben

#2

Could be the utility switching a capacitor bank in or out which could momentarily through your voltage out of range, either over or under.

#3

You have a Voltage measurement line, which doesn’t seem to twitch.
Could it be something else causing the solar generation to droop - shadow of a flagpole? regular plane flying over? (kidding, that would be too fast, but you get the idea)

#4

Keep in mind that PVOutput resolution is 5 minute data. A voltage dip or surge could last for only a few cycles to a few seconds and would be enough to cause the inverters to drop but still not be visible on PVOutput.

#5

Its fairly early in the morning - what is the altitude of the sun at that time? is it low enough that the sun is shining through or past a far-away tree or building, perhaps on another property?
If its a terrestrial shadow, it should gradually move earlier or later by one to four minutes a day - over a month you should see it gradually shift along the timescale.

#6

Thanks for everyones responses! Some notes:

-There isnt any shading to the panels.
-I run the Raspberry pi gateway (Enecsys Solar 4.1.0), which logs data every minute. The output:

image
image1796×702 55.4 KB

Each inverter has the dropout (at the 7:53am mark) at the same time).

At this point, I might just log the powerline voltage for the morning. Since it repeats about the same time every day, I wont have too much data to parse.

Would anyone happen to know what the turn-off voltage is of the Enecsys inverters? Periodically I need to start a large motor, causing the voltage to sag and my inverters to shut off for a few minutes. The above plot is not from starting my motor.

#7

I used to get a dip like this in the early morning on my ground mount array.
Caused by a very tall eucalyptus tree to the East in the alley two houses away.
Was a fairly significant reduction in generation for some parts of the year.
Cost $1600 to cut the tree down and donate the firewood to the neighbors.
Worth it.

#8

Its a doozy of a puzzle, to be sure. 7:53am isn’t even on a nice hour boundary to suspect something happening in software at 0:00UTC, and then being timezone-shifted. Best of luck.

#9

Is the Raspberry Pi configured to run some complex job at or before 07:53? Saturating its I/O.

DHCP lease renewal? Router restart?

#10

I had a further thought. Is the time of the Raspberry Pi correct? If the Pi’s time drifts strange things might happen. I am running ntpd on mine.

#11

I use Enecsys inverters, and also run the PI with the same software (older version of Enecsys Solar but the same no less) .
I sometimes see the same issue you describe. Mine happens around 10-10:30 AM, but not every day, mostly in the summer months. The issue is the line frequency going out of range. The inverter(s) will shut down, then restart after about 5 minutes time. Not all of my inverters see it but when it happens there are a few that do.
If you want to talk directly, email me at vskinner@rgv.rr.com From there, we can exchange phone numbers and discuss if you like.