Battery jumps from 76% to 100% and stops charging - solar diverted to grid

Good Afternoon folks.

Yesterday I had my battery percentage go from 75% to 100% and stop charging. Excess Solar was diverted to the grid. Interestingly, givTCP reported the battery at a constant 76% whilst the app and the dashboard reported 100%.

Later in the day the battery percentage started to fall from 100% gradually, just as you would expect once solar had stopped generating. As a household our idling consumption is between 200w and 500w and the battery seemed to go down in line. This morning the battery was at 60% when solar started up and the battery started charging again when generation was high enough.

I thought this was a one off and maybe down to a corruption at data somewhere.

However, it's just gone and done the same today. A good solar morning had the battery up from 60% to 76% by 11:15am but then it seems to be oscillating between charging and export and slowly creeping up - very slowly. At least according to the dashboard and app. Both the app and givTCP report export to grid as before.

Anyone know whats going on? Is the battery suddenly telling the inverter it's full at 76%?

I'm a bit new to the fold, sp to speak so am still getting my head around how it all works but this does seem odd to me. Up until now the system appears to have behaved pretty much as I would have expected.

Coming in to the Autumn/Winter period I would like to make sure I am able to get the maximum benefit of the battery capacity.

The setup is AC 3.0, 9.5kwh battery retrofitted to existing 4kw solar array with power one inverter. The givEnergy installation was less than a month ago.

I would post images of the dashboard graphs but can't work out how to do it.

Thanks

cluelesscris
Was your battery calibrated when it was installed?

On GivTCP there is a SoC reported by your battery. Another reported by the Inverter.
Do these tally?

Yes the Battery was calibrated when installed. It was also re-calibrated by GE support after we had a new smart meter installed and the engineer screwwed up the shutdown/startup order. The numbers reported though givTCP all seem to make sense with each other.

The only other thing I could think of was that this happened after the first time the battery was deleted to 5% and went idle for a while and dropped to 3% before solar kicked in to get it up to the 76%.

I have since calculated and checked the battery usage against the drop from the '100%' to 60% overnight and it seems to match what I would expect.

The slow creep has now got the battry to 79% so I'm thinking it might all right itself.

Watch this space . . .

This appears to be a data issue. Delving into the details it seems tat even though the portal said the batery was not charging it seems to have been. Yesterday, after a period of slow drift normal charging resumed (at about 3 in the afternoon) and the battery charged to full and showed so in all channels.

My batteries were installed and went live from 11/07/22. Using the data graphs available from the portal, it's easy to spot instances of a sudden SoC snap to the alleged capacity (2no. 8.2 100% DoD).

22/07/22 13:25 67% - 100%
29/07/22 12:50 79% - 100%
01/08/22 12:00 74% - 100%
19/08/22 13:35 58% - 100%
(This one prompted a call to GE who initiated a recalibration)
26/08/22 11:40 88% - 100%
06/09/22 13:15 64% - 100%
10/09/22 13:55 79% - 100%
13/09/22 13:40 87% - 100%
17/09/22 13:45 80% - 100%
20/09/22 14:05 88% - 100%
28/09/22 14:40 75% - 100%
06/10/22 13:20 46% - 100%
08/10/22 14:40 87% - 100%
10/10/22 13:50 87% - 100%

There are many instances of a sudden snap from 95% to 100% but I have not included these in the list above. For most of the remaining dates, the sunlight was insufficient to get the battery to a full charge, which to me suggests a fundamental flaw in the SoC calcs for my batteries (when sufficient light for a full charge) or a physical issue (or both).

GE suggested a 3010 firmware was due, and somewhere here there is a sizeable post with people having the reverse issue (sudden drop in SoC), with a possible a small beta test of the firmware.