DelongTop LiFePo Packs SOC unbalanced

I live in New Zealand and have purchased three 10kWh LiFePo battery packs (total 30kWh, 3 x DL-LFP-51200’s) from Delong Battery Technology (DelongTop) from Shenzhen, China. I have been having a problem with the three packs getting progressively further apart regarding their individual SOC % (State of Charge). It got serious a month ago when we had several days of wet weather, and two of the packs were showing about 45% SOC while one pack was only showing about 25% SOC.
I noticed that when the packs were being charged their BMS sent the highest SOC value of the 3 packs to the Growatt Inverter/Charger, (displayed on the inverter screen), but when they were discharging the BMS sent the lowest SOC value. This meant that I was not going to be able to use the extra storage that was in the two higher SOC packs when the inverter inevitably shutdown due to the minimum SOC alarm setpoint being reached, (set to 20% SOC).
It is impossible to get any sensible answers from the Delong Support people because none of the engineers speak English. Has anyone else encountered this SOC difference problem from the Delong battery packs before? I am also keen to find out how many of these new Delong 10kWh pack have been sold into NZ and Australia.

So, basically, your batteries are not being top-balanced during charging due to the battery with the highest SOC controlling the charging. Very strange

Most professional battery systems have a master battery or master BMS that is able to balance all individual batteries (and cells) and provide a universal SOC to the inverter. This is one reason why they cost more.

Is the BMS internal or external? Can you provide photos of the setup? I’ve never heard of Delong

Hi Svarky, thanks for your response.
The battery packs are individually stand-alone, that is, you could use a single one on its own, or stack up to five vertically, all interconnected with parallel loops, and a single pair of wires running to the inverter/charger. See this below link to DelongTop.

On the Home page, mine is the one on the bottom right-hand corner which shows 5 stacked, but I only have a 3 stack, (30kWh).

Each pack has its own internal BMS balancing the 16 x 200Ah cells, and its doing that job really well, (apart from a major problem when they first arrived, but I won’t address that here). Each pack must have its own CT’s and it continuously adjusts its own SOC% based on what current is passing in and out of its own pack. They each have different CAN node addresses, mine are set to 1, 2, and 3, and they have a serial loop connecting them all together. Only Pack 1 is connected to the inverter/charger via a Cat.5 cable presumably running a CAN protocol. So Pack 1 acts as the master of the 3 packs as far as the inverter/charger is concerned, and as I said in my first post, I have watched the charger display panel reading the highest pack SOC% during charging, and the lowest of the 3 packs when discharging.
Because they are all wired in parallel, when there is no current in or out they have exactly the same voltage reading, (can’t be different of course). So as far as I can see, if any one pack hits its maximum allowed charging voltage (set to 57.6v), they should all reset whatever SOC% they are currently reading to 100%. This would make them “self-calibrating” and it would occur at every full charge, (ie, on most days).
I have tried to get this message through to the engineers at Delong, but since they only understand Chinese and I just use Google translate for my emails, I am getting no proper responses from them. I have been programming PLC’s for the last 32 years and I reckon I could do a much better job than they have to ensure that the SOC% don’t steadily diverge from each other.

By the way, that Delong graphic and the wordage was automatically added by the post system, not by me. At this stage I personally wouldn’t entirely agree you the words that have appeared there.