XUBA 280ah in 48v w/ "dumb" inverter/charger

[thegangler]

Looking to connect 16 of the 280ah 3.2v XUBAs in series to my Schneider Conext XW6048 and hope to get some expert opinions. I can control min/max charging profile, but unsure on how/if I can use an external BMS considering it uses a proprietary protocol for inverter/charger control. My understanding I could connect the batteries in parallel to even them out, then in series to the inverter and run 10% discharge/90% max charge and be fine, only issue would be not fully maximizing the potential. My use case is ac-coupled, grid tied as a battery backup for my house, even at that profile it's 3 days of backup power.

 

[nosys70]

the problem is you will not manage each cell, so one or more cell can go wacky on top or bottom and you will not notice it before it is too late.
i got only 8 Xuba cells in serie (they were all perfectly showing the same voltage) but at first charge without BMS, i got 2 of them showing 0.5V difference when charging (3.2 for 6 and 3.7 for 2)
So you should get a 16 cells BMS.
the only thing you need to check is the controller is cutting before the the BMS (like 3.1v for the controller if the BMS is set at 3.0V)

 

[nwillitts]

nosys70 is completely right.i tried what you are suggesting and monitored each cell by hand and with out a BMS,cells will get whacky.

 

[thegangler]

the problem is you will not manage each cell, so one or more cell can go wacky on top or bottom and you will not notice it before it is too late.
i got only 8 Xuba cells in serie (they were all perfectly showing the same voltage) but at first charge without BMS, i got 2 of them showing 0.5V difference when charging (3.2 for 6 and 3.7 for 2)
So you should get a 16 cells BMS.
the only thing you need to check is the controller is cutting before the the BMS (like 3.1v for the controller if the BMS is set at 3.0V)

Right but the Schneider inverter/charger can't integrate with a BMS so all I'd get would be cell monitoring - assume no solar in my setup, I'll be charging from the grid.

 

[Denis]

I have two banks of cells in parallel 16s2p, burnt two bms on the first day of use, I have top balanced every cell individually, with a drok dc-dc buck converter 10-75v input to 0-60 v output 0-12amps, and I have been using them without any bms for 90 days now and all my cells stays balanced up to a certain point. I've set my controller to charge up to 56.8v and it stays there for about 5 minutes, and then goes on float at 54.4v. I don't suggest to go any higher than 56.8v, I'm still experiencing with them myself and at 56 v would be even safer to prevent a cell runner. The lowest I've seen the banks go is 52.5 v during the night time. I've got my inverter to cut cells bank out at 50.8 v.

 

[nwillitts]

how did you kill your BMS?
what BMS brand was it?

 

[nosys70]

you probably didn't kill the BMS, it was simply that you have such discrepancy on some cells that the BMS disconnected.
Some BMS are just here to monitor and protect, they do not try to balance.
I have a 200A Daly BMs that is a blackbox. there is absolutely no indication of whatsoever happens.
for less than 50$ you can get a battery monitor that shows voltage for each cell.
that could be enough

 

[Dzl]

Right but the Schneider inverter/charger can't integrate with a BMS so all I'd get would be cell monitoring - assume no solar in my setup, I'll be charging from the grid.

The BMS doesn't need to integrate with the inverter (unless I am missing something or misunderstanding you). Its a plus if it does, but for most people the BMS is an entirely separate layer of protection from any component level management/protection done by the inverter or chargers.

 

[thegangler]

The BMS doesn't need to integrate with the inverter (unless I am missing something or misunderstanding you). Its a plus if it does, but for most people the BMS is an entirely separate layer of protection from any component level management/protection done by the inverter or chargers.

The BMS can't stop the charger from pushing charge to the batteries without some form of integration, right?

 

[Denis]

how did you kill your BMS?
what BMS brand was it?

The BMS were the active balancer bms brand new on the market , look at the thread under bms you will see it all explained and with pictures, these were defectIive. cells had no discrepancies as I was monitoring them with the meter as well to make sure . The balanced feature was excellent,

 

[Dzl]

The BMS can't stop the charger from pushing charge to the batteries without some form of integration, right?

Not necessarily. A FET based BMS (the BMS is the switch) or a BMS that uses external relays can cut charging to the battery regardless of what the charger does and requires no integration.

 

[nosys70]

there are pretty powerwul BMS. you can program some to do anything you want based on conditions like cell voltage.
obviously putting hard rules on BMS on a battery pack that has very different cells behavior is not a good idea since this condition will
likely trigger often a BMS disconnection.
In the case you want to go in "paranoid" mode, the first thing to do is bin sorting the cells to make sure they are all the same, then you can
implement a BMS with hard rule to report any discrepancy.
another way to go is to just monitor the cells and find a low and bottom limit that keeps all the cells happy even if you expect one cell to fail sooner than the others.
My opinion is if on 16 cells you got only 2 problematic ones, i would replace them.
you can probably use these cells for other purpose.
I got only 8 cells and 2 acting different, so i will keep them.
i will try to find a max voltage that does not push the high ones to high, but still charging to low ones enough.
on the low side, i got no problem, the cells tends to equalize when they drain.

it is also possible you get the reverse problem (some cell going low faster than the others), but the solution is in the same ballpark

 

[YellowBuddy]

Looking to connect 16 of the 280ah 3.2v XUBAs in series to my Schneider Conext XW6048 and hope to get some expert opinions.

Not an expert, but here's my novice opinion.

1) Schneider uses their Xanbus for communication on the XW6048.



If you don't trust your BMS, it seems they make a Conext Bridge to help it communicate.

https://solar.schneider-electric.com/wp-content/uploads/2017/05/Conext-Bridge-DS20170405_ENG.pdf

At that point it seems it'll talk ethernet, xanbus, modbus, canbus. Greyhound would be jealous..