Went from 27% SoC down to 0% over the course of a 5 minute snapshot this evening with demand being around 300w.
I've had this happen before but not since the last firmware update which I thought corrected it.
Bloody annoying as its then pulled grid energy to get the SoC back up to 4% as well as not supporting demand until it's scheduled charge tonight either.
THALL nope it's 4.1kwh usage for an 8.2kwh battery.
Only charged up to 75% overnight, so I make that about 6kwh available. Was around 50% at 1620 and 27% at 1930 before dropping down to 0.
Household load for the day was as expecting and fairly low as we were out all afternoon with limited PV due to the weather.
Should have had more than enough to cover the days usage as I make 27% around 2kwh still. Discharge figures for the day match the SoC prior to the cliff edge drop off too.
It's not too bad today but suddenly loosing 25% SoC on a heavier load day or in December is not ideal.
This happened to our batteries last night too. Suddenly dropped from 23% to 4% in between 5 minute updates, when the house load was only drawing about 200 watts.
hoggy last firmware update was mid December.
I've had some days since then where it's I've run out of battery, Christmas day for example and it's run down and discharged as you would expect based on load.
Interesting I wasn't the only one yesterday looking at this thread and it seem like similar drop offs of around 25% SoC down to 0 in a single 5 minute snapshot period.
Going to raise this with the hardware support team side as something a bit odd is going on of its more than just me. I seem to recall that last time it did this the battery lights didn't agree with the portal info but did reflect roughly where I thought the SoC should be.
Random question but what time did your drop offs occur?
Mine was between 1941 and 1946, if they are all the same time window it might help with the diagnosis.
Had an answer back from @Adam-Reynolds many thanks for the rapid response hopefully what I've written below is accurate!
The condition of the battery cells drift over time versus the last battery calibration and the calibration is used to achieve SoC % values. The system uses cell Voltage values as an actual measure of the state of the battery instead of the % SoC value.
As these values drift apart, ie last calibration %SoC versus actual battery condition as cell voltage it's not noticeable in day to day usage but it become obvious when the system corrects for the variation resulting in a large %SoC drop-off.
@Adam-Reynolds also stated that this is normal behaviour and that no further action is required post the SoC % jump as this is the system completing its correction. He also said it doesn't always happen with every system nor is there a frequency etc.
We have a 8.2Kw battery, with a 5.3 added a few weeks ago. I know the system sees 2 batteries as 1. I noticed over the day, the 8.2 green battery leds drop down as it discharges, but the 5.3 battery always stays full, as though at some point, when the bigger battery is empty, it should then indicate it's charge reduction as it discharges ( I've not actually noticed the leds on the 5.2 change later on in the evening ).
I wondered if there's a 'fault' when this changeover happens? It might be a simplistic observation, but I'm pretty p*ssed off I'm now on grid all evening since my battery dumped to zero around 1900 this evening? I logged a bug with Giv, and surprised others are suffering this problem too. Also, my 5.3Kw battery still show 2/4 green leds on it, which isn't 4% that the system shows.
Giv - please help!
I wonder if there is a correlation between those who are seeing this behaviour and not charging to 100% or emptying the battery very often. I would expect that if you charge to 100% every day, or regularly fully discharge, then the battery voltage would regularly reach it's high and low extremes and hence be able to recalibrate the percentage to 100 or zero (4) with very little opportunity to drift. As a data point, I have not seen this behaviour, but I also charge to 100% multiple times a week. I have the AC coupled 8.2 battery.
chrisdev I've been charging to 100% overnight up until middle of last week where I dropped it down to 75%.
It has been getting fully charged via PV instead though bar a couple of odd days so plenty of opportunities for the BMS to sort cell balancing etc.
Worth noting that as mentioned above this is expected occasional behaviour.
As for the battery lights I think this is due to the way the battery status lights are driven, if the battery lights are sequenced via the initial calibration to show SoC% as the drift occurs versus actual battery cell voltage they won't change.