Using IR, C and C lost at 24S Li-pack

[occuplus]

As written in the title, I am preparing a 24S 5P LiFePO4 battery pack for my e-moped.
I bought 200pcs 32700 cylindrical cells to use 120 of them for this. I will not use 10 of them because of the capacity and IR. (The cells are very bad even though they are brand-new)
I measured the capacity(3,65>2,6@3A) of the batteries, Internal Resistance (AC 1Khz) and capacity loss in 1 month [full charge>1 month later, full charge=delta C](not SoH).
web based Repakr, which works online, uses a single variable (C).
Nemo@secondlifestorage prepared excel based RepakR. This document can deploy using both C and IR. (unfortunately I can't use C-lost here )

1-Is there a reason why the dev-from avg mAh is high at the first S?

2- Would it be of any practical use to use more powerful cells at the mostnegative in 24S?

3-Is there any problem in distribution?

4-Is there a better way to do it manually in excel?

Thanks for help. : )

Repackr.xlsm

drive.google.com

 

[Wolf]

Nemos repackr does not really sort by IR just capacity, I had Nemo add IR so that when the final build takes place you can rechecking V and IR before final cell placement.
What is the nominal capacity of these cells supposed to be?
Here's a thing I would try.

Since you are only using 5p there is a limit of how well the repackr can sort. I would limit my cells capacity difference to maybe between 5400 to 5600 or thereabouts. You can use excel to work out the best 120 to 130 cells that match the closest then put repackr to work on those. I'm sure you will have much better results
By removing the exceptionally high C cells and the closer the cells are to each other in capacity and IR the better the battery will perform.
Wolf