Variances in packs capacity

I don't know how to title this question...

Say you have plenty of cells of 2200, and then a surplus of 1500-1900 ones up to be able to make two banks (14s100-110p). What would be more correct/safer to do?

• Mixing all the cells and make two banks of mixed cells.
• Build first one bank of 2200 (200Ah packs)and then build another bank with the rest. Without increasing the pack size to match. Is it possible to parallel different capacity banks?
• Build the first bank and increase the pack size of the second to match 200Ah. The cells on those packs will range from 1500 to 1900.

The reason is, that I sorted things with energycells for the first bank, but I've also been contacted with someone relatively local, willing to send me laptop packson cheap. The only problem is the sample 300kgcells that I've checked so far, range this capacity.

What you will do with such different kind of cells?

SeRiusRod said:

What you will do with such different kind of cells?

Click to expand...

It depends on what the SoH (State of Health) of the 1500-1900mAh cells is. For that matter of fact also what the SOH of the 2200mAh cells is.

Anything below 80% SoH is not really worthy of a build.
Let say you have a 2200mAh rated cell and tests out at 1580 mAh. That's 72% SOH I wouldn't use that cell. The IR (Internal Resistance) would most likely also be too high.
The same goes for a 3000mAh rated cell. Even if it tests out at 2230mAh its SoH is 74%. Again not useful in a powerwall. That is if you want it to last.
Also the chemistry makes a difference.
The best thing you can do is to match IR and mAh of all cells as close as possible and you will have a balanced bank.

Wolf

The 2200 are all above 85% soh
Almost all 1500-1900 where originally 2200mAh.

I would recommend that all packs be the same ah(s) - regardless of how you mix cells. Also acceptable IR on each. This will give the best 'even handed' balance and wear over time.

As mentioned above - its not the original mah pec of the cell that matters, its how much capacity they currently have. You need totest them and write the mah on them and sort them by 100s. of mah capacity You might wind up with a box of2000mah, a box of 1900mah, a box of 1800mah, and so on down to 1500mah. Also as mentioned above, youmay want to exclude cells that test <80% of the original capacity.

Then you evenly distribute the cells among your packs - till they add up to the target ah that you have for your pack design. Let's say you want 150ah packs. Then if you distribute them like this...
- 18cells * 2000mah = 36ah
- 20cells * 1900mah = 38ah
- 15cells* 1800mah = 27ah
- 20cells * 1700mah = 24ah
- 5cells * 1600mah = 8ah
- 12cells * 1500mah = 18ah
Then you have a total pack of 36 + 38 + 27 + 24 + 8 + 18 = 151ah. 147-155 range is close enough to 150ah in my experience.
You'll notice that the method above tends to distribute the cell typesthru the packs - so the packs are pretty similar in makeup... which is good as well if you can get each'battery' (7s or 14s etc) to be similar packs as far as cell types.

In my case, I have a physical pack design to accommodate up to 112cells260ah/pack. Some of my packs have 112cells and some have 87cells - but all the packs add up to 260ah (+/- 6ah) and this works well for me. I try to make sure each battery (14s87p-113p) has all the same cell manufacturer/type - but that's not strictly needed and I do it because I know I want to ahead of time and make purchases so I can.

**There are several approaches. The one I just outlined is just my own approach and not in any way the only approach The key thing I would emphasize is tohave all of your packs with similar ah / pack... so that your battery(s) (7s, 14s, ...) all have the same capacity if you hook them in parallel.

What a comprehensive answer, thank you very much. So all should be matched, between packs but also between batteries.