Added 2 additional HV2600 batteries - BMS now all over the place

[andrewmoore]

Hi all,

On 22nd November, I added two additional HV2600 V2 batteries to my stack, taking the total to 7. With the H1-5.0E inverter. I followed the following process:

1. Powered everything off and turned AC and DC isolators to off.
2. Measured existing and new battery voltages. They were within 0.2V of each other.
3. Connected new batteries to the stack per FoxESS diagrams.
4. Set DIP switch on BMS to 5 (2 less than 7 in the stack).
5. Powered everything back on, power up checks all okay. BMS reported 17.02 kWh available, when existing batteries are at 95% SoH this is bang on (2560W * 7 * 0.95).
6. Charged stack to 100% at 5A, then held at 100% for ~2 hours.
7. Discharged stack to 10% at 5A, then held at 10% for ~2 hours.
8. Charged stack to 100% at 5A again. Then continued to use battery as usual.

Now... the issue. I'm only seeing about ~9.1 kWh usable capacity. If I discharge from 100% to 10%, the inverter reports ~9.1 kWh discharged that day. This is about the same capacity that I saw with 5 batteries, so significantly less than I'd expect from 7.

Also, when charging and discharging, the SoC jumps the last ~20% or so. See this graph where the SoC jumped from 30% to 9% when discharging and 75% to 100% when charging:


Do I just need to be patient and wait for the stack to balance, or is there definitely something wrong? It's now been 4 days, and I'm still seeing about the same reduced capacity.

Thanks in advance.

[Will]

With the temps being lower this time of year it can take 10-20 cycles to balance everything.
Reading what you did everything is correct so just monitor it over the next week or two. I'm sure things will balance and improve.

[andrewmoore]

With the temps being lower this time of year it can take 10-20 cycles to balance everything.
Reading what you did everything is correct so just monitor it over the next week or two. I'm sure things will balance and improve.

Thanks Will, I appreciate the reassurance!

Current cell low is 14.6C and cell high is 19.7C, so they are on the colder side.

Is the actual cell voltage imbalance the only thing that matters, or is there more to it than that in terms of the BMS "learning"?



I have a Home Assistant template sensor that's just:
((cell_high - cell_low) | abs) / (cell_high + cell_low) * 200

It's a bit all over the place, settling around 0.6-1.2% when steadily discharging but rocketing up under heavy load or charge.

[Will]

the BMS recalcuates with every full cycle to update it's internal estimate. Cell imbalance and temp are going to be the only things causing what you're see.
I don't see any other issues but I know it's concerning as in the summer expanding battery doesn't take more than 2-3 cycles but I've seen it with my own eyes take weeks when cooler.

[andrewmoore]

I'll report back in another week, fingers crossed!

[Will]

keen to hear your update, thanks