Empty battery gained charge with no input.

Hello,
I've recently had a GIV-HY3.6 inverter and a GE 9.5 kWh battery installed along side a 4kW Viridian array and after some faff and hassle (partly down to the installers and partly down to firmware/settings) it all seems to be working properly. I've only a week or so of data to go on so don't know if this is normal or not, hoping someone can let me know.
With the poor weather here over the last serveral days my battery was empty and the house was running from the grid. I went to bed and expected the battery to remian empty until the sun came up sufficiently to start charging the battery from the PV. When I looked at the charts the next day it looks like the battery suddenly found some extra charge (down the back of the sofa perhaps?) and ran the house for a little bit in the night, twice.
I had seen this happen the night before as well when I was up and I don't think it's a display issue or sensing issue with the GE kit because the smart meter display suddenly went from around 400W to 0W for a period so it does seem genuine. Is this down to battery recovery when not under load?
This is the battery chart:

This is the grid chart which shows it wasn't charging the batt. from the grid:

I'm interested to know if this is normal or weird!
Cheers, Mike....

Sunrise is a bit after 4am at the moment. By 8am the panels will be charging nicely. You don't need clear skies, just brightness. My 5.2kWp array was charging at over 4kW the other day in wall-to-wall cloud cover.

Yes, but my concern are the two instances (one at midnight, the other around 02:00) when there's nothing coming from the PV and nothing showing coming from the grid but the % charged of the battery increased and power was sent to the house during those two periods.
Cheers, Mike...

I think your battery will have cooled down overnight (have a look at the temperature in the battery data column) then the calibration curve will say it wasn’t as empty as it thought it was.
I’m not 100% sure of this but if it’s only ever at night then it’s the most likely.

if your battery is on firmware version 3012 then this may well be an OCV adjustment. Basically the battery stopped when it got to 4% but after an idle period it checked the voltages in the battery and decided that the true SOC should be 10% (for example), this then means it is happy to supply more energy to the house until it hits 4% again.

Sounds like a temperature/SOC calculation thing where the battery suddenly finds some electricty after a rest to cool down. I get the same on my EV. After parking for a while after a trip, the SOC and range increases a bit.

Ah, right, thank you. That would certainly explain it. Although configured for 4% min. mine always stops at 5%, dunno why. I am on firmware 3012. I'm glad it does reconsider and not just stop 'till charged, shaves another little bit off my power bill.
Cheers, Mike...