Making up series/parallel arrays of batteries.

[opampsmoker]

Hi,
We have a  load of used lithium cells  that we wish to build up into series parallel arrays. We will just charge them all up, and put them in series parallel. Will there be any problems? We wont bother with any resistive "balancing" between the cells prior to paralleling them. Will we be OK?

[IanB]

How are you going to define "OK"?

[MosherIV]

Will there be any problems?

Yes!

Yes cannot just parallel cell willy nilly.
They must be matched in capacity, output rating (C), voltage and age (SOH).

Putting them in series - the chain will be limited by the weakest one in the chain

[sleemanj]

Parallel cells do not need to be matched for capacity, just ensure they are at same voltage when you parallel them, by nature of being parallel they will always be balanced as long as they are in parallel.  The discharge/charge rating of a set of parallel cells is of course no more than the lowest rated of the parallel cells multipled by number of cells. 

Series cells should be matched for capacity in order to utilise all the available capacity, if you do not then discharge or charge must stop too early for some cells in order that others are not over so.  Balancing cells is debatable value if the cells were well matched for capacity and state of charge (voltage) when put in series - balancing is rarely seen in commercial products, although for salvaged cells which may be of mixed heritage balancing may be of benefit.  However protectiom to monitor each cell voltage and cease charge/discharge when the likits are reached for any cell is highly desirable no matter the cells in use.  The charge/discharge rating of series cells is that of the lowest rated one.

It is worth considering failure modes, specifically if a cell goes short circuit.  Fuses or fusible links for each cell are common in large packs.

[MosherIV]

Parallel cells do not need to be matched for capacity 

So commercial pack manufacturers have been doing it wrong then ?
 

[voltsandjolts]

treez > zenerbjt > opampsmoker ?

[sleemanj]

!

Parallel cells do not need to be matched for capacity 

So commercial pack manufacturers have been doing it wrong then ?
 

At it's simplest, you would perhaps agree that 2 identical cells in parallel is the same as 1 cell double the capacity yes, so by your fear you would never have more than 2 cells in parallel, because that third cell would be only half the capacity of the battery it is being added to. 

That isn't to say that parallel cells do not ideally need to be matched in other ways, namely the internal resistance should be broadly equivalent in all parallel cells particularly if higher per-cell current draws are required. 

Capacity is not a way to measure that, although capacity degradation could be, if you have 2 salvaged cells that started new at 2Ah each and now they are 0.8Ah and 1.8Ah then they likely have developed markedly different internal resistance, but if you have 2 brand new unused cells, a 1Ah and 2Ah marked cells, then that says nothing of the internal resistance indeed they are likely similar being brand new they just have different amounts of material inside.

Pack manufacturers use the same capacity cells, because why would they not, it makes sense from a manufacturing standpoint, but it is not a technical requirement of it's own, in parallel.

NB: Obviously this is for same-chemistry cells.

[MosherIV]

The reason you should parallel cell with the same capacity is to preserve their life.

Lets say you pair cell A and cell B.
Cell A has half the capacity of cell B.

What happens when cell A is depleted ?
Now cell B has to supply all the current.
Why did we parallel cell in the first place? Because we wanted to increase the power ( current) output because a single cell did not have the power (C rating) that was required. However, now that cell A is flat, only cell B can supply current but on its own it does not have the rating needed - you are now using cell B beyond its rate C. This will degrade the life of cell B !

Wait there more ........
Since the cell are joined in parallel, and cell A is empty, its cell voltage will actually be lower than cell B (yes it will, it is just very difficult to measure). So cell B is actually passing a balancing current into cell A as well as supplying the output current. Further loading cell B 

In addition, if you did not over rate the bus bar joining the cells, the extra current could heat the bus bar because of the extra unaccounted for balancing current.

Yes, you CAN join miss-matched cells.
My original point - it will cause problems.
It is not worth doing it. If it was, you can gurantee that pack manufacturers will be doing it to save money!