Making a pack from 9 year old Winston Thundersky cells

[Swing]

I have almost 200 Winston Thundersky cells of 60Ah , about 9 years old. I think there is still for 80% capacity in them, so at least 30kWh to be used.
However, some cells are probably not that healthy and will die quickly.
They were never paralleled.

What are acceptable internal resistance values for old Winston Thundersky cells?
They had less than 0.55mOhm when new.
I have an YR1035+ internal resistance tester. Some cells are now at double that, for example 1.2 mOhm.
I'm going to measure them all, and filter out ones with very high values (like 2mOhm and above)

But they will get an easy life with low C rating (far lower than 1C)
However, I want to parallel up to 12 of them (16s12p) and I don't want the bad ones bringing down the rest.

Would it be a good strategy to test them all for internal resistance, and filter out the bad ones (that have way more than double the original internal resistance).
Then, with the remaining good cells, just mix them (low with high resistance paralleled)
You would think it would make sense to group them by internal resistance, but then you would just create parallel blocks of weak ones, and parallel blocks of good ones, putting that in series will probably cause big imbalance in the pack.

I guess it would be good to capacity test each cell. But that will require lots of work and time.
Alternatively, I could just hook them all up parallel with a fuse on each cells.
That would be cost increasing, but I need some extra busbars anyway for connecting them parallel, might as well be fuses.

And of course I will put a strong active balancer on it, for example 10 Amps.

 

[12VoltInstalls]

Would it be a good strategy to test them all for internal resistance, and filter out the bad ones

Yes, if:

guess it would be good to capacity test each cell. But that will require lots of work and time

…you fully charged each of them, record resistances, and capacity test each of them.

I’m no expert, but if my understanding is correct there is really no other way to ensure that you are making an accurate assessment.

don't want the bad ones bringing down the rest.

Reflecting on the idea that often ‘new’ factory assembled cheap LiFePo batteries have suboptimal cells- perhaps used- that theoretically were factory tested and grouped for somewhat similar specs, and then ‘they’ label/assemble the battery for a predicted consistently attainable Ah, then essentially what you are attempting is the same same results for the cells you were fortunate source.
So if you do the work - due diligence- you will achieve similar or better results than a factory produced battery assembly.

30kWh is a pretty decent storage capacity for a wise “midsized” offgrid system imho, and even at today’s prices that’s approaching a $10k savings over delivered batteries. So it’s a) worth the time to test, and b) should give you maybe a ten-year window to bank the cash to replace them when they again age-out of their intended capacity.

 

[Swing]

Yes 30kWh (rough estimation) is significant.
But it might also be the case that I start using them and quite a lot of them die and it turns out to be more work.

I guess I need some kind of single cell tester that is either quick (higher Amps) or cheap (to do more parallel)
And I still would need to fully charge them, for which I think I cannot combine much of the cells.

 

[MattiFin]

I’d check the cells for self-discharge before building the pack.
Top balance and leave sitting for a month, discard everything that is not at full resting voltage.

 

[Warpspeed]

I would connect them up as twelve separate series strings of sixteen cells per string.
That way you can watch and follow the voltages of each individual cell.
Put the whole thing through several charge/discharge cycles and see what happens.
It should be fairly easy to spot any weak cells that very easily go under/over voltage with respect to its neighbors.

Then rearrange things so there are some known strong strings, and maybe a few cells that can be brought back, and others that end up in the wheelie bin.
I have a series string of thirty 60Ah Winston cells (100v 6Kwh) that have been running great for over six years.