Unequal numbers of cells in parallel groups?

[theabsurdman]

Imagine you had a 7s4p battery where one of the parallel groups was weaker (lower capacity) than the others. Is it allowable to add another cell to that parallel group (only) to boost it up? So that group would become 5p, which might now make it the strongest group in the pack. Are there any potential negatives to such an unmatched pack?

I've Googled the hell out of this to no avail.

I can't see how this would be any different to having matched (in number) parallel groups made up of different capacity cells, which is apparently OK - accepting that the resulting pack would only be as good as its weakest group.

 

[docw009]

I don't think it matters.

What if you connect a fully charged 21700 5AH cell in series with a 18650 3AH cell. Run an amp thru it for an hour. They will be at different voltages, but if we recharge them in series at 1A, will they both get back to 4.2V at the same time?. EVen if they don't, won't a balance BMS equalize?

 

[amberwolf]

Whatever you do, the lowest capacity group in a series string will limit the capacity of the entire pack to it's capacity. The highest internal resistance group will limit the capability (current delivery) to that capability. They are probably the same group.

You can add cells to any weaker groups to bolster them, just make sure the new cells are the same voltage as the ones you're adding them to to limit current flow across your interconnects as you install them, and keep in mind that any cell that is already causing you a problem is also only going to get worse, even with other cells to bolster it. If the cell has the wrong kind of problem, it could become a fire hazard. Unfortunatley one effect of bolstering a bad group with extra cell(s) is that it can mask a cell that's rapidly going south because now the BMS doesn't shut off from LVC on that group.

Also keep in mind that any pack not made of matched cells will discharge all the cells at different rates, which means they will end up at different voltages, and age at different rates too (the worse they are the faster they will age because they're being pushed harder, which makes them worse and ages them even faster).

Making matched groups (so each group is the same capacity and internal resistance, etc., as all the others) is better than nothing, if you can't do all matched cells.




A BMS with a balancing function (they don't all have one) can make all the cells (groups) the same voltage (balanced), but they are still different characteristics and will still discharge at different rates, and become different on every usage, and require rebalancing on every recharge.

All balancing is intended (or going) to do is keep the pack as usable as possible for as long as possible, by making whatever capacity is available in the weaker groups all be full (it doesn't matter if the other ones are full, as long as they have at least as much available filled capacity as the weakest group).

 

[harrisonpatm]

If you scroll to the picture at the bottom of this post: OffGridInTheCity Build

You'll see that this DIY powerwall builder is adding packs to his string based on capacity only. He wants to make sure that each series grouping has the same capacity. Number of cells per pack doesn't matter. Some packs have more cells than others, but each series group has the same capacity. Because as pointed out in this post and others, the lowest capacity group is always going to be the limiting factor.

So, to answer your question, if you were to add a single cell to the weakest group, then it would absolutely make it the not-weakest group. Electrically, there's no reason why this wouldn't work. But there's other considerations. Are you able to capacity test each series group individually, so that you can add a cell to the weakest group with the right capacity? If not, you're just shifting the weak spot to another series grouping. Do you have room in the pack for one cell unevenly? Is there an external reason why the weakest group is the weakest, that doesn't have to do with the cells themselves, like damage or shorted/damaged connections?

Anyway, its not to say that you shouldn't do it. Just keep in mind the other factors affecting the pack. For example, when one series group is identified as being significantly weaker than others, its somewhat common to remove that whole group entirely and replace it.

 

[Voltron]

What's your fire situation?
If you only ever charge it outside in a bbq under supervision kinda situation that's fine, but it's sure going to raise the fire risk of indoor unsupervised charging if you're in an apartment or big family house etc.