Battery damaged?

Hi - hope someone can help.

GivHY-3.6 (Gen1.5) inverter and 5.2kWh installed 22nd Nov, so 20 days ago.

I've been using GivTCP to monitor it. and can see it repeatedly drops below the 4% minimum charge level for extended periods. In fact this morning, it's dropped to complete zero (39V) and won't do anything.

During the 20 days since commissioning I've had a timed charge each night for 00:30 - 04:30. Of those 20 nights it's ignored the timed charge on 17 of them. Sometimes it would randomly charge at other times... 3am, 6am, 9am... And I've now realised on all those occasions the battery had just tripped below 4%, so maybe that was forcing a charge to prevent undervolt. But now, it doesn't even do that.

GivEnergy support have kept telling me my timed charge settings needed resetting, but I don't think they've realised the significance of the battery repeatedly dropping below 4%. Which tells me it's not a timed charge issue. It's just a charging issue overall.

Now it's in fault mode.

Will these extended periods of deep-discharge at <40V, and below the 4% protection threshold, have damaged the battery long-term?

Many thanks.

See this thread

https://community.givenergy.cloud/d/713-firmware-3007-has-not-fixed-sudden-soc-drop-to-0-problem/1464

In short
there is a problem calculating the percentage. A fix is due soon.
4% isn't really 4% there is some extra buffer for safety
Can't comment on whether the crash is bad for batteries, mixed opinions on that I think.

As for the fault status, not sure what to advise re that, maybe contact your installer plus hopefully someone else will be along to offer advice!

Riscy suggest you use the web portal to do an inverter restart (top of inverter tab). Once it’s back reset to defaults (inverter tab, setting at the bottom). That should put you to a known position. Check ECO mode is enabled. Set your timed charge to whatever you want. Also check that the Charge to % is where you want. Check the battery settings in the same inverter settings tab has charge and discharge power set to max. That should have the inverter set to properly charge on your schedule. You may want to give it a ten minute boost to check it will charge otherwise it’s a call to Givenergy again.

Thanks both. :-)

I don't think this is the same "sudden drop to zero" problem. It gradually gets to zero by going through the threshold floor quite gently. I'm predominantly using the voltage reading from GivTCP as my measure (because I know zero isn't really zero)

It's now at 37V, when the undervolt warning protection is triggered at 43V. So way below the protection level.

I've already done a reset-to-defaults on the Inverter several times over the last 20 days, during various fault-finding requests from the support team. Each time it's been set back to Eco and then a timed charge. It still doesn't charge when told.

And now it won't take any charge at all. Not even if I reset everything (battery off, AC off, PV off, restart.) It still won't charge from AC, regardless of the time window. Even with force charge set, it's sitting there continuing to empty and the voltage continues to fall...

Will this be damagng the battery? Am I better totally switching it off, so we don't get further charge leaking away?

Is the battery warm enough to take a charge? I seem to remember seeing that it will not charge below 0C.
If it is not this, then it sounds like there may be an issue with the install or the hardware.

I think the inverter powers it self using the battery ( if there’s no mains e.g. power cut) and it will go below 4% in that scenario

So I’d be suspicious about the cables and switches between the consumer unit and the inverter, including the RCBO in the consumer unit.

Maybe the energy meter too. Make sure that’s got power and is cycling through the values and that those values appear to change e.g. total import is increasing. But I’d lean more towards above.

TX200 Ah, now that's interesting...

I had serious problems with the LoRa connection when it was first installed. (the installer forgot to put the antenna on the supply sensor... This led to umm "erratic" inverter behaviour for several days until I figured out it was getting duff demand readings.)

Having said that, I've been monitoring the inverter's demand readings from GivTCP ever since then and comparing them to my Glow MQTT feed, and they do seem in-sync.

Do you mean I should look at the EM115 display and check import? If there's any good documentation for it then do let me know.

Cheers folks, very impressed with the community here!

Try here meters uk

Another common problem is installers putting the Ct clamp on the incoming cables the wrong way around. If they did that then it might look like you are exporting when you are actually importing.

Riscy As @TX200 suggested, check that the inverter is connected to AC. What does the portal inverter tab say about mains voltage and frequency? If the AC connection between the grid and the inverter is disconnected, the inverter will power itself from battery but I don’t know if it will report the internal (eg EPS) voltage and frequency. Either way there will be an AC isolator by the inverter (you’ve already checked this). There may also be a fuse or MCB nearby as well as the device in you house consumer unit. Check they are all on?

Tim Hi, thanks. Yeah I've checked all the MCBs and they were on, and the AC isolator too. Also I've since totally isolated the battery (i.e. turned off the 2x MCBs that connect it to the Inverter) and the Inverter is still running (in the dark..!), which I think is further confidence there's no fault with the AC connection to the main distribution board. So yes it shows the current mains voltage and frequency, 259V and 49Hz as I type currently.

I might post a few screenshots from the portal showing the strange charge behaviour. It's only observed the 00:30 start time on one occasion. All other times it's charged much later in the night; usually once the storage has dropped to 1% and then it seems to have panic-charged...

But now, it's totally dead. The status light is a solid red, and the capacity light flashing green. So I've now isolated it totally to prevent it discharging even further.

Bunch of warnings:

BMS warnings

All the following images are from a baseline of resetting to defaults and then a timed charge from 00:30. But, it looks like it only charges at random times, and normally as a result of the charge level dropping way too low. Most other nights it didn't charge at all.

2022-12-01
2022-12-03
2022-12-08
2022-12-11

EDIT: Just seen your Grid voltage 259V. Provided that's not a typo then that is above the statutory limit of 253V.
It may not be allowing the charger to kick in as the voltage is too high.

Original post below:

It'll have dropped below the level at which the BMS will allow it to take a charge. There is a way around this by essentially putting a shorting link across the BMS and force a charge direct to the cells HOWEVER this is not something you should attempt yourself.
One for your installer & GE to battle i'm afraid. It should be self recovering at all times provided the inverter AC is still live it should keep topping backup to 4%. Yours not being able to do this suggests all is not well.

hoggy Thanks hoggy. Good advice. I will definitely not be shorting across anything! 😆

I'm a bit annoyed because I did repeatedly tell GivEnergy support that the battery was repeatedly dropping below the lower bounds and wasn't recovering.

hoggy Possibly I misunderstood there... Sorry for the red herring. The Grid voltage is indeed more normal; it's the Inverter voltage that swings between 220V and 260V... Presumably that's correct because that's the mechanism for pushing current back the other way?

Grid voltage
Inverter voltage

Riscy

If you seeing hugely different Grid and inverter input voltages, Id also have your installer check the cable. Grid measured at the EM115 and the Inverter Input should be pretty close if now the same, unless full load/charge where you may get 1v difference (Volt drop).

Either way if the inverter sees voltage over 253.6 it will go into protect mode, till it reduces below the upper limit.

Riscy there is something amiss here for sure. The grid voltage should be 230V +/-10%. If you’re seeing voltages ranging as you say, it implies to me that either your grid voltage is varying way out of spec, or the inverter isn’t connected to the grid. I think you need a visit back from your installer to check it out. Givenergy told me that the inverter will increase its voltage by as much as 3V for maximum export to grid but by much less to supply house load. In fact it needs to be only very slightly higher than grid voltage otherwise it will be exporting.

TheDragon (GivEnergy) voltage over 253.6 it will go into protect mode

The operating range of the hybrid inverter is 180 to 280V so I wouldn’t expect it to go into protect mode. It would make sense for it to do that but I’ve seen high voltage trip on my Zappi at 258.6V while the inverters are happily exporting. I’ve complained to the DNO but they’re not interested as I think it’s solar installations on the local distribution system that’s driving the voltage up.

Tim Each inverter is different. I know my 2 Solax shut off at 256 and throw an alarm.
Not seen since batteries went in

Check the size of cable back to the main consumer unit. I know it's a 3.6k hybrid inverter - but most 'Solar' installers have no clue when it comes to Ohms law. Check the cable - voltage drop is deltaV per meter per amp. 2.5mm cable for short runs <10m is okay - anything longer should be 4mm and for a 5kW inverter 4mm for 'short' runs and 6 or 10mm for longer runs. If the cable gets warm at full export power then it ain't thick enough ...

my inverter to CU is about 4m and a 6mm2 cable was used.
This is like feeding a cooker

Any chance of some more help - fairly urgent question...

Installation engineer visited today. Sure enough, battery well below voltage and wouldn't charge from Inverter. Engineer charged it manually using a bench power supply. It took ages to take any charge, but eventually got the voltage up from 36V more towards 40V, and then the battery BMS woke up again and started a recovery charge from the mains.

But obviously, a recovery charge only lifts it to 4%. And that's what got me into this mess. The battery was at 4% and then discharged well below (to 36V) and then died.

Do I risk leaving it at this low charge again?

Or should I set the charge window to 00:00 - 23:59 right away, and get some more charge (=heat) into the battery while it's briefly alive?

Sorry to ask for help - just I don't want another 2 weeks of chaos going back and forth with support :-(

Cheers

Personally I would be looking for a warranty replacement. In the interim I would charge the living bejaysus out of it as soon as possible.

Yes but perhaps lower the charging rate, so it charges slower and thus keeps warm but not hot for longer?

I would get it chaeged, at say 0.1C. So for a 5.2 single 10A or 550w
Once you can see it charging ok, Say at 20%, then up it to 1000W per pack till 50%.
Ten normal full power

I like the idea of monitoring the temperature and charge level and adapting the charge rate accordingly. Once this all settles down - and I can convince Support that it's not due to me changing settings (which it's not) - I'll certainly investigate.

For tonight, I'm graphing just about all measurements I possibly can do. Once the Voltage drops below about 45V, or the cell temp drops to less than 5 degrees, I'll initiate a charge. Unless the timed charge actually works and does that for me, by some miracle.

graphs

Worked last night.

Failed again tonight.

Battery temp was a little over 5 degrees when charging should have commenced. Voltage 50V.

Incredibly sick of this now. I've been incredibly patient but increasingly feel fobbed-off by support. I've done exactly what they asked, including getting the installer back (who was at a total loss.)

Any ideas? How does the Live Chat work on the website - I can never see the option, even during office hours. Is there a screenshot somewhere of how to find it?

Riscy During office hours there contact us page has a dark blue circle in bottom right with a white quote bubble. Click on that and you should be good to go. Sorry screen shot uploads aren't available as I'm away ATM.

Riscy
Did you try a manual forced charge?
Results?

TheDragon (GivEnergy) Last N times this has happened it's refused to force-charge. I.e. even when I've set it to 00:00 - 23:59 charge period it still wouldn't do anything.

Last night I stayed up til 1am (again), realised it still wasn't charging, and decided to put a portable oil-filled radiator next to it in case the cell temp (6 degrees) was causing the failure to charge.

Within literally 2 minutes of turning the oil-filled radiator on, the battery started charging at 2.7kW. Now, I think this is suspiciously quick... So either the warm air convected over the battery sensor (even though the cells were still at 6 degrees)... Or there's something weird where the sudden high-load on the house supply triggers the inverter to wake up. More datapoints required?

Riscy This is definitely now something to do with load.

Same again last night:

1. 15 minutes into the charge period, nothing was happening. Cell and battery temp all > 6 degrees. Two data sources (Inverter and Glow IHD) confirmed only baseline grid import.
2. Switched on a high-load (oil radiator).
3. Immediately, not even with a minute's delay, the Inverter started flowing charge at 2.7kW to the battery.

Does anyone know what might be causing this?

If it's notable, my electrical configuration has two Consumer Units with the Inverter at the far end:

Meter ----> House CU --[32A breaker]--> Garage CU --[32A breaker]--> GivEnergy CU

Riscy
Where is your grid CT located? Does the grid reading in your GivEnergy portal match the iHD?
Shut battery and solar off.

Then take a reading, say 500w import. Has to be impirt as no other sources. IHD should also be 500w
Now turn the kettle on. IHD goes upto 2500w, so should GivEnergy.

If they dont, the CT location or direction is wrong. Hence your odd symptoms

TheDragon (GivEnergy) Saga continues... Thanks for your continued interest too.

Did as you suggested. Dead of night, battery off. Turned on a large house demand, and the Giv portal shows this:

GRID DATA:
[Time] . . . . . . . . [V] . . . [Fac] . . [W]
2022-12-19 00:47:17 . . 242.5 . . 50.01 . . -1999

Note the Minus sign - verbatim.

Is that right? Or should it be a positive? I've compared to the Giv demo portal and it looks right.

Sadly I'm a software guy and not an electrician, so I can't validate exactly where the CT clamp is within the House CU... But it's definitely in the House CU, because I already diagnosed (and fixed) a problem with the LoRa. So it's:

Meter ----> House CU/CT/LoRa Tx --[32A breaker]--> Garage CU --[32A breaker]--> Giv CU

Riscy that's correct.

Import reads on my grid data as negative.
So CT Maxwell be in the right place.
If its inside the CU then you won't see it.
Normally there put on the meter Tails, incase the main house CU isnt the only one and its split at the meter in a Henley block.

Seen this a few times, where the CT is only monitoring some of he house and not all.

TheDragon (GivEnergy) The meter box definitely only has 1 set of tails (i.e. 2 cables, Live + Neutral in a 100A breaker)

Whether that's then split in the House CU, I've no idea, without the bravery to remove the CU housing.

Managed to get online chat with Giv Support just now. They say it's an Inverter firmware bug (odd this hasn't been mentioned in the month of support calls?) and will get me an upgrade.

Current firmware is D0.904-A0.904 and the unit is a "Gen1.5" i.e. Gen2 internals.

They bumped the firmware one minor revision to D0.905-A0.905 but the problem persists.

I am at the end of my tether.

Again, every night, I can "kick" the inverter by applying a huge load to the house demand (2kW) and within 60 seconds it starts charging the battery. I then remove the load and it carries on charging to 100%

But if I don't give it that "kick", then it sits idle and the battery keeps falling below the floor charge level.

GE at a total loss. Installer at a total loss. Savings from overnight tariff: also a total loss.

Any more electrical insight? I promise I'll help with GivTCP but probably not if I end up ripping the whole installation out...

[unknown] This is definitely odd! You said at the start that you had one AC-connected inverter and battery; you don't have a separate inverter or PV?

The EM115 seems to be reporting grid import/export correctly but the AC3 does nothing during a timed charge period unless it sees a large import, when it starts charging, but not before? It's almost as if the inverter sees the large grid import as export and starts charging.

TBH, if it was me, I'd get a fresh pair of eyes on the installation to confirm that everything is connected as it should be. Exactly why the inverter won't charge the battery during a timed charge period is very strange. I can understand that very low temperatures recently may have clouded the issue, but if the inverter charges the battery to 100% when given a "kick" but not otherwise, there's something amiss.

My daughter had an installation which was problematic and they couldn't commission it. It was eventually traced to the neutral connection between the consumer unit and the inverter that hadn't been connected (at all). All good now but it took three different sets of eyes to work out what the issue was. That's obviously not the issue with your installation because it does charge when "kicked" but there must be something amiss if the inverter firmware works fine with other inverters of the same type but not yours. Sounds like hardware; either the installation connections or the inverter or the Modbus connections. Sorry I can't be more helpful.