diy solar

diy solar

Safely balancing batteries with different SoC

kolek

Inventor of the Electron
Joined
Sep 29, 2021
Messages
510
Say I have 4 large 280AH batteries with a high state of charge, and 4 more with a low SoC.
I assume if I use big busbars and parallel these 8 batteries there could be drama like sparks and possibly even damage to the cells.
I have some 18awg wire, could I safely get around this by just using the thin 18awg wire to parallel the cells, so that the transfer of power is slower, and just wait longer for the 8 cells to come into balance, or could there still be sparks or damage?

Edit:
Adding background information:
I started out charging a 16S pack of 280AH batteries in groups of 4 cells. The problem with that is, after charging 4 separate groups of 4, each group will be at slightly different SoC, so I will need to parallel them at the end anyway.
So I decided rather than charge them in groups of 4, I will just parallel all the cells and charge them as a group.
To do that, I need to lower the SoC of the first group of 4 cells which are at a higher SoC, or combine all the cells "as-is" and them them shock each other into an equal SoC, and I want to do that safely and slowly.
 
Last edited:
Andy at Off-Grid-Garage on YouTube did a test with paralleling low SOC with high SOC, and the result was very uneventful. I have paralleled 10% SOC 48v 280ah batteries with 100% SOC of the same, and it peaked at like 150a or something. Nothing that could do damage. The voltage differential even from 10-100% is not very large.

If you want to be safe, why not charge up the low ones, or discharge the high ones a bit?
 
@Lt.Dan Thanks for your reply. So using 18AWG wires should work then to safely parallel the batteries and bring them to an equal SoC?
 
@Lt.Dan Thanks for your reply. So using 18AWG wires should work then to safely parallel the batteries and bring them to an equal SoC?
I would not use 18AWG, as they will not support the current! Use proper gauged wires when paralleling batteries.

I should have been more clear here:
it peaked at like 150a or something. Nothing that could do damage to the cells or BMS.

I assume if you paralleled them with 18AWG, the wires would melt/burn fairly quickly while the batteries attempt to pass 150a through them.
 
That's my concern exactly, the part about melting/burning because the wires aren't thick enough, but I actually think maybe it's not a valid concern. Let's see what others have to say. Thanks!
 
18 awg would be unsafe. Use large wires and as long as they are within a volt of each other resting you probably won't trip the BMS. Better if you can get them closer to .5-.1 volts close first.
 
18 awg would be unsafe. Use large wires and as long as they are within a volt of each other resting you probably won't trip the BMS. Better if you can get them closer to .5-.1 volts close first.
No BMS is connected. When you say "within a volt" do you mean 1 full volt?
 
It sounds like you are talking about individual cells so I would take one charged and one discharged and parallel them with standard bus bars, yeah there's going to be a bit of a spark but they'll soon settle down.
Or you could use something like a 12 volt 21w car brake light bulb that will help to slow the current down and produce less spark.
 
Yeah, I think there is confusion since it sounds like the term "battery" by the original poster is being used incorrectly here. It sounds like he has 8 280Ah cells which he wants to use as a single 24V (8 cells) battery?
 
I started out charging a 16S pack of 280AH batteries in groups of 4 cells. The problem with that is, after charging 4 separate groups of 4, each group will be at different SoC, so I will need to parallel them at the end anyway.
So I decided rather than charge them in groups of 4, I will just parallel all the cells and charge them as a group.
To do that, I need to deal with the first group of 4 cells which are at a higher SoC.
 
I started out charging a 16S pack of 280AH batteries in groups of 4 cells. The problem with that is, after charging 4 separate groups of 4, each group will be at different SoC, so I will need to parallel them at the end anyway.
So I decided rather than charge them in groups of 4, I will just parallel all the cells and charge them as a group.
To do that, I need to deal with the first group of 4 cells which are at a higher SoC.
Oh that's right, this one. I think you have to stick the course and get your other 4 or 8 or 12 up to the first group now before you parallel them. If you have 12 you can do all 12 at once. It won't save time to try and take 4 backwards.
 
Ah yes, using the term "battery" in place of "cell" caused me to assume they were assembled packs you were trying to parallel.
 
Do you have a BMS for the 48V battery then? What is the voltage of the highest and lowest cell? A JK BMS for instance can have a 2A active balancer built-in which depending on how far off the cells are could balance them relatively quickly. Otherwise I've let my JK BMS balance cells out over a couple weeks when I wasn't in a hurry and the cells were pretty out of balance.
 
Having some cells sitting for a day or two while you get the other cells charged up should not be a problem.
 
Oh that's right, this one. I think you have to stick the course and get your other 4 or 8 or 12 up to the first group now before you parallel them. If you have 12 you can do all 12 at once. It won't save time to try and take 4 backwards.
@hwy17 Thanks a lot for your response.
That would be good but I don't want to have a fully charged battery because it could be a couple weeks before I can manage to get everything put together. So ideally I want to just drain down these 4 cells or spread the charge out among all 16 cells.
 
@hwy17 Thanks a lot for your response.
That would be good but I don't want to have a fully charged battery because it could be a couple weeks before I can manage to get everything put together. So ideally I want to just drain down these 4 cells or spread the charge out among all 16 cells.
What are the voltages of the higher and lower SOC cells now?
 
If I had a proper load tester I would just use that to drop the charge on these 4 cells. On the advice of someone on this forum I tried using a 200W bulb, but it's only consuming about 8W (obviously because it's designed for 100V, not 12V)
 
I'm seeing the 18 awg method theory now, let us know the voltages and the length of 18 awg and let's do some math and then if you monitor it for fire maybe we can learn something.
 
@hwy17 Thanks for your help I'll get those voltages for you in a second.

Just wanted to mention another option I have handy is Nichrome wire. Meaning, using that as a way to lower the SoC.

Nichrome 80 24g
1.61 Ohms/ft Resistance

Voltage / Current = Resistance

13.3V / 10A = 1.33 ohms
13.3V / 5A = 2.66 ohms

1.61x = 2.66

X = 1.65 feet would generate 5A of current, half that would generate 10A, if my math is correct.
No idea how many watts it would generate or how hot the wire would get.
 
Back
Top