Expected battery efficiency?

Ongoing saga... Too many problems to bore people with in entirety. Any thoughts on this latest one? 5.2 battery.

Taking a typical day, numbers reported by the Portal:

• Grid to Battery: from 4% to 100%, using 4.67kWh
• Battery to Home: ran down from 100% to 4%, delivering 3.57kWh
• Battery temps are good, >10 degrees all day long.

I have two observations - any opinions welcome.

• Charging from 4% to 100% should take more like 5kWh, so it's accepting 10% less charge than it should be.
• The efficiency loss is a whopping 1kWh on a 5.2kWh battery, i.e. more than 20% which doesn't seem reasonable to me.

Note this started from 4%, went up to 100%, and back down to 4%. It's a brand new installation, albeit one where a fault in the firmware let the battery drop down to 35V for a day or two, which I wonder if has frigged it?

I think the number of overlapped problems on this installation is bewildering the support team.

35V is 10V below the BMS lower limit (you sure it's not 45V?) so I would honestly say something is very much wrong with your system if it even allowed it to get to that point. I would not be surprised if you have a bad/weak cell somewhere (beware that the 5.2 batteries actually have 32 cells inside not 16 so although GE band about "all cells being monitored" it's actually 16 PAIRS of cells being monitored so may be masked voltage wise from an adjacent cell)

hoggy It was down to 35V, yes. Not a typo! That was at a point where the Inverter was solidly idle and wouldn't put any charge into the battery. I sat watching it drop below the 0% mark, all the while being reassured by support that "0% isn't really 0%" and being ignored when I pointed out that the voltage was well below the lower threshold. I had to get the installer back out to use a bench-supply to re-charge it. It's never managed to do a recovery charge, ever.

Hence after all that, and now the problem with the battery capacity and in/out charge, I wonder if it's been permanently damaged.

Quite possibly a fatal mistake them not acting sooner.
All they can ultimately do is re-run a calibration which will empty it & (hopefully) refill it back up which should then allow you to get an accurate reading.

(as an aside there is a whole thing about the 5.2s DOD - 80% or 90%. Nobody has yet been able to find one with 90% written on the side of it. Some users claim that theirs do run to 90% DOD but it's hard to get hands on actual data.
GE have now said the 90% thing was a typo all along (which somehow remained on datasheets for months & months with "Premium/Gold" or whatever tier thing they use installers stating it as fact.
Then you have the long lived SOC bug making it quite difficult to know if the % you see is the true value...
Long story short - I'm not sure what goalpost you should be aiming at for what constitutes a "normal" 5.2 battery.

It's all a mess really!!!

Riscy sounds like something is amiss - either the clamps doing the monitoring or the actual battery. I have 2x5.2kwh batteries, and the numbers generally look OK to me. For example last night it charged from 25% to 100% - i.e. 75% charge. That's about 7.8kWh (OK I've ignored any buffer). My separate EmonCMS monitoring system (uses current clamps) reports 7.46kWh of charge was recorded - so close enough for me.

I do find that the emoncms vs the givenergy figures for charging the battery are often out by about 500-600W, which I think probably reflects the inaccuracies of the clamp based current monitoring. The clamps are on the same feed

GrahamRead Do you see the GivEnergy reading higher or lower than the current clamp power reading?

GrahamRead Daft question; where are the battery current CT clamps located?

Assuming the current monitor is correct, I'm losing 1kWh per charge cycle of ₄.5kWh in. That's a 25% system loss. Almost enough to render it pointless doing Eco7 charging, given the penalty of more expensive daytime rates.

1kWh of lost energy would obviously be dissipating as heat, and given that this all happens in a 10hr period (the battery is flat by midday ish) it would have to be generating an average of 100W of heat from the Inverter/battery continuously... Doesn't feel right.

NB I'm charging at 1.3kW to reduce the I2R losses already.

I am losing about 1.5kWh on every full charge/full discharge cycle. I have 2x 5.2kWh batteries. The losses could be due to the BMS or the inverter. Holding a constant charge over several days (to eliminate the inverter losses) I am seeing a decay in the battery % which the inverter tops up every few hours. This works out to be a battery drain of 50-70W, far higher than typical BMS power consumption which I read online to be around 2W per BMS (so 4W in my case).
I have a ticket open with GivEnergy for several months but they just do a test and answer as though the issue is a state of charge/capacity problem, which is a separate issue altogether. They don’t understand the question. I am now trying to seek warranty redress.

To add, I also had an issue after installation where the installer incorrectly wired the system and the batteries went seriously low for several weeks. I didn’t measure the voltage but I suspect damage has been done.
Can you do a similar test to mine and see what you get.