State of Charge SoC drops of nearly 50%!!!

Like many others, my Gen1 9.5kWh battery has had SoC drops of between a few percent right up to 48%. This drops in around thirty minutes meaning that the SoC readings are completely meaningless. This seems to occur when the battery does not discharge completely before charging up again (ie typical in the summer).
I was using firmware 3017 and have had beta 3018, both have SoC issues and the beta seems even worse with the SoC drops happening every day!!!!
GivEnergy appear to have had SoC issues for years and I see no sign that they are close to fixing the issue. All you get offered is further delays while you wait for a new firmware version that does now work. This problem has been visible since May 2024 and I see no solution in sight.
How can GivEnergy sell a product that does not perform to specification?

Would be glad to hear of any solutions from anybody.

IanCoopergu47 Morning Ian. The SOC really does seem unreliable if the battery isn't 'exercised ' to both ends of the scale. Chemistry was never my strongest subject and it appears to be a problem for the designers too, together with the circuitry in the battery.
However, I have a 9.5kw battery same as you, and as I'm on Flux during the summer, I do empty it most days between 4-7pm leaving me just enough to see me through till the sun comes up in the morning.
Sometimes overnight in the wee small hours, there is a 5% odd increase in the SOC when the battery is not under any main load and the battery cells settle, the BMS re- adjusts itself.
Likewise, at the top end, at 96 or 97% it will sometimes jump to 100%.
If you are able to export to empty the battery even just twice a week, you may find that the SOC gives you a more accurate reading.

Givenergy say exercise your battery
https://youtu.be/s7lAUEvJpwU?si=Qg9jBDFKO937jF-s

TimHutchings Sometimes overnight in the wee small hours, there is a 5% odd increase in the SOC when the battery is not under any main load and the battery cells settle, the BMS re- adjusts itself.

I see the same: I choose to idle the battery every night at around 10% SoC for an hour. Every 3 or 4 days, say, it drifts up a few percent. In fact, looks like it's done it on each of the last 3 days, so that's interesting...

Went up from 8% to 14% before AC charging started at 00:30. (Battery idles from 23:30 to 00:30, but it never seems to start drifting up until after midnight.)

TimHutchings However, I have a 9.5kw battery same as you, and as I'm on Flux during the summer, I do empty it most days between 4-7pm leaving me just enough to see me through till the sun comes up in the morning.

A quick check of the flux rates in my region shows peak-rate export of 22.9p vs day-rate import of 22.49p. So there seems to be a very small benefit in over-exporting between 4-7, even if it leaves you slightly short later, since importing to make up the shortfall is at-worst cost-neutral.

DD Export rate 4-7pm in my area is 23.63p. So I export to around 30% at 4-6.10pm which usually sees me through the night till the sun comes up. If we use a lot of power in the evening then very occasionally I'll top up between 2-5am at 13.85p for half an hour, but it rarely happens.

TimHutchings Export rate 4-7pm in my area is 23.63p

What's the daytime import ?

Out of interest, do you (usually) run from grid between 2-5am, or continue to run from battery ?

DD Daytime import rate is 23.08p for Flux in my area, but I never use it. I charge the battery fully through the day using sunlight and the battery is normally full between 11am and 1pm with excess exported at 15.51p.
I don't bother using the grid between 2-5am except, very occasionally to add 15 minutes charge, if the Inverter app tells me I'll run out before 7am. Obviously, in spring and autumn as the days shorten, the nights lengthen, I'll adjust the charge import between 2-5am and shorten the export between 4-7pm. When the A2A heating starts kicking in mornings and evenings I'll need more and when the balance is right, I'll probably swap over to Cosy again around November to Feb. All pretty straight forward really!

IanCoopergu47 Well GE is just bad at what they do. I was suffering from what you are describing when my system was provisioned almost daily. Months later they released 3012 and 3013 firmware fixed if for me.
A couple of months ago I was foolish enough to update my battery firmware to 3015 and to 3017 (beta at the time) and the basically reverted my battery to almost daily SoC drops. Eventually after many pointless calibrations I asked GE support to downgrade my battery to version 3013.
I do not agree with other peoples statement that you have to "exercised " your battery in order to get reliable SoC (you shouldn't have to). On firmware 3013 that is not required at least on my battery - I can stay above 60% SoC for weeks and still be able to discharge to single digits % without drops.
Well I learned my lesson and I have no plans to upgrade any of the firmware any time soon.