A strange new battery charging problem

I have 18 x 435 kW panels, a 5kW inverter, and a 9.5 kWh battery installed in February 2024. It has been working perfectly; the panels charged the battery to 100% by noon and excess PV power was exported to the grid for the rest of the day. After sundown the battery had sufficient charge to run the house through to sunrise.

After a grid outage on 2nd June a problem appeared. The battery charges normally at the start of the day at a rate of 3600W, but at some apparently random trigger point each day the charging rate falls to a flat and constant 290W. This trigger point can happen when the battery charge is anywhere between about 60% and 90% full. The battery trickle charges at 290W until it reaches 100%, which can be quite late in the day (and would be after sundown in the winter). I cannot find any parameter with a value of 290W, and the battery charge and discharge rates are set to the default max of 3600W. The system is set to ECO mode. See the attached image.
Why is this happening, and where does to 290W value come from?

You could try a reset to defaults in the portal, inverter settings. It could be that an inverter register was corrupted in the outage. If that fails, you could try a complete shutdown of your system, then a power up. If that fails, I would suggest calling support to ask them to have a look at your system remotely.

Brassy one possibility is that the BMS has lost track of the amount of energy in the battery. The charge rate tends to slow down when the voltages approach their max allowed.

How low does the battery level tend to get overnight. If it rarely gets below, say, 30%, it might be worth running it all the way down, which might allow the BMS to realign it's state. (The voltage curve of the battery is very flat over a wide range of charge levels, and it's only at the extremes that it can realign.)

Someone more knowledgeable than me could probably look at your battery voltage curve and take a more informed view.

(I generally run battery down to under 10% most evenings, ready for a cheap overnight refill. But whether that's beneficial financially will depend on your tariff.)

Brassy ; the panels charged the battery to 100% by noon and excess PV power was exported to the grid for the rest of the day.

With that setup, you might experience solar clipping. Not harmful, but potentially means you miss out on some solar goodness. (Just a small obsession of mine...)

TimHutchings IIRC someone else had the same problem and its the battery. Needed replacement after many attempts from GE to "calibrate/slow calibrate" the battery. Replacement was the same too but the replacement for the replacement was/is OK 🙂

The fact it happened after a grid outage probably isn't good news - seen at least one similar where the whole house was on EPS and the installer hadn't fitted the right trip for the battery.

Call/email support would be my advice.

I had similar to this, and it was when cells got too higher voltage compared to others.
On mine cell 5 and 13 went too high (or too low on discharge) and BMS going into safety mode
It was caused by a loose bus bar between pack of cells 1 and 2 & 3 and 4 (the 9.5 battery is 4 cells tp a pack, 4 packs to a battery)
Engineer came out and 15 mins later it was fixed.

Thanks for all of your thoughts given above.
@TimHutchings; I tried a rest to defaults, and there was no improvement. I then did a complete shutdown and power up and once again there was no change. The installer is talking to Giv Energy about this, no response yet.
DD; at this time of year the battery is not falling below 30% ish, I think I'll try your suggestion of forcing a full discharge to the 4% baseline in the early hours. I guess that any solar generation during the discharge will add in with the battery discharge up to the inverter max of 5kW if I limit the discharge rate to, say 2kW, and export excess to the grid after supplying the home. I do get solar clipping sometimes if the total power through the inverter exceeds the 5kW rating, but not a lot, and I'm happy with this because the system will perform better in the shorter duller days of winter.
Vestas; my own suspicion is that I have a faulty battery cell.
wrighar ; Interesting.

The trickle charging of the battery begins at a different set of system readings every day. I cannot pin it to battery charge level, battery temperature, inverter temperature, time of day, PV excess output, or home load. The only constant every day is the trickle charge power at 290W. This lasts until the battery eventually reaches 100%.
See below
![
](https://)

Brassy my own suspicion is that I have a faulty battery cell.

If you use Android, can I suggest https://givenergymonitor.wixsite.com/tutorials which (among other things) gives a view onto individual cell voltages. Which might allow you to investigate that.

There are other ways to get hold of the info, but that's possibly the simplest.

DD Unfortunately we have no Android devices in the house, only Windows and IOS.
Thanks anyway.

Just had this from Giv Energy via my installer:

"This is a known issue with battery version 3015, there is a fix in the works and it is being beta tested so should be being publicly released as 3017 soon".

Good that they know and they're working on it; but why is it not listed under "known issues" on the cloud portal?
Why is nothing listed under "known issues" on the cloud portal?

Brassy There is an iOS app too. It's different, but can probably access the same info. Actually, it's available for many platforms, inc. windows.

See https://community.givenergy.cloud/d/3611-all-in-one-bms-release-8/266

(I've never tried it myself.)

Brassy Good that they know and they're working on it; but why is it not listed under "known issues" on the cloud portal?

Yeah, very frustrating. Last interaction I had with support was much the same - "it's a known issue" that's somehow not listed on the list of known issues. Would have saved me a lot of time composing a detailed description of the symptom, and saved them having to answer my email, if they'd just written a brief note that it was something they were working on.

There was a recent Q&A video where they said they were working on some new and shiny known-issues interface with lots of bells and whistles. Which I'm sure will be nice, but only useful if they actually take the time to admit that there are known issues.

Even in the past, when I've seen that list populated, the known issues didn't seem to get mentioned until after they'd been fixed. (Perhaps they were more useful before my time, though.)