question

sesshoumaru avatar image
sesshoumaru asked

Loads runaway

Hi


I noticed that suddenly my Pylontech Battery went down to 10% and stopped discharge - as configured. However there was no load on the system which could justfiy the discharge from 100% to 10%.


Luckily the home assistant recorded all happenings during the night. Out of nowhere suddenly the loads (AC + critical) went up to 10.000 Watts, decliing slowly to 2200W in the morning. This causes the battery to discharge. There was no real load. All the dichargd energy went into the grid (15 kwh export over the night).


When I checked it was still claiming 2000W load. Just because I was out of ideas, I restarted the CerboGX, and TADA ... the loads took about 1 minute but then went down the expected value and now everything works again in balance.


Question: This smells like some firmware bug (Cerbo was running for about 8 weeks). Is this a known issue? Is it expected to reboot Cerbo GX after a certain period of time?

MultiPlus Quattro Inverter Chargercerbo gx
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

1 Answer
Alexandra avatar image
Alexandra answered ·

@Sesshoumaru

Something in the system caused the batteries to discharge into the grid?

It is unusual but may have been related to a grid set point being changed? Since it is integrated I would suspect that something happened in the interaction of the integration between home assistant and the Cerbo.

And the Cerbo should have recorded feedback to the grid.

2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

sesshoumaru avatar image sesshoumaru commented ·
Thats the strange thing. My setup has a EM24 to act a grid-meter.

Still the miscalculation of a high load on the system caused the battery to blow out its energy, the EM24 must have shown that to the Cerbo, still is continued until the battery went down to 10%.


This makes me think how about the balacing works. Grid set-point was at 50W at this time (it changes from 0W when PV is on, and 50W when PV is out, for reason to trigger discharge of the AC coupled Varta Element 6).


The calculaton of how much energy the battery puts out is probably not based on the grid-meter alone, and the second part on the formula must have been caused the runaway. I mean 10000W loads where shown, which never where there and went away after Cerbo restart.

0 Likes 0 ·
Alexandra avatar image Alexandra ♦ sesshoumaru commented ·
The energy had to go somewhere. If there was a fake load and the Pylontechs did not actually discharge, they would still be showing full [or whatever you they would normally].

The only way to feed from battery to grid is to set a negative grid set point as far as I know.

Does the Varta part of the setup discharge to grid ever?

Is it possible the energy meter had a malfunction of some kind? The Cerbo only reports the information is is fed, it performed the low SOC stop discharge function correctly, which leads me to believe one of the devices that feed it information sent a bad signal or command. In this case the suspect is the EM24

0 Likes 0 ·