Parallel cells without BMS

[kdrgjhgdf]

There have been a few threads that have discussed and disabused the notion of running parallel cells without BMS.

I'm not arguing the point — it seems quite reasonable that, if things do get out of balance, then they can go south, and quickly.

But what intrigues me is: what is the difference between a large cell, and two cells half the size wired in parallel without BMS?

If we unpick the insides of the single large cell, then it's effectively wired parallel.

(I realise some batteries have their own BMS, and perhaps are some kind of composite internal many-cells-series/parallel+BMS architecture. But that composite battery architecture isn't what I'm referring to — I'm thinking of the physically large, singular cell.)

Is it manufacturing tolerances? Or perhaps it's simply the fact of a batch chemistry being relatively uniform? Or is there something about the electro chemical process that "smooths things out" in a way that is more effective at smoothing out discontinuities in internal SOC levels?

 

[sunshine_eggo]

There have been a few threads that have discussed and disabused the notion of running parallel cells without BMS.

I'm not arguing the point — it seems quite reasonable that, if things do get out of balance, then they can go south, and quickly.

But what intrigues me is: what is the difference between a large cell, and two cells half the size wired in parallel without BMS?

No.

If we unpick the insides of the single large cell, then it's effectively wired parallel.

No. It's just a bigger single cell.

EXTERNALLY, when you parallel cells, they become one cell for all intents and purposes, but they are physically different from a single larger cell.

Comparing 2P16S vs. 16S2P (two separate 1P16S batteries in parallel):

The cost of a second BMS is usually justified over a parallel cells arrangement.

Multiple lower current BMS can deliver more bank current, i.e., two 100A BMS for 2X 16S batteries in parallel may be not significantly more expensive than a single 200A BMS.

If a single cell fails in a parallel arrangement, the entire battery must be taken down, the cells in question separated and then individually evaluated to determine which one is actually bad. Your entire battery bank is offline in this case until you reconfigure it with good cells.

If a single cell in a 16S2P arrangement fails, you know EXACTLY which one, and you can continue to use 50% of your battery bank while addressing the issue with minimal system interruption.

Note that this is coming from a guy with a 21P14S Lithium NMC battery constructed from Panasonic PHEV cells. Had I not already had a lot of money invested in these materials, I would NOT have done it this way and stuck with a three 16S LFP batteries in parallel, each with their own BMS.

 

[aussie desert]

The only difference is the terminal connections between parallel cells.

I use powerplus batteries which are comprised of cylindrical cells in parallel, i think these give the easiest to understand layout for how parallel cells operate.

If you picture 10 cylindrical cells connected by two plates, the cell capacity is multiplied by the number of cells. If one cell fails short circuit, the other cells supply current and the shorted cell fuses. The capacity is reduced by the loss of the single cell.

If you have cylindrical 20ah cells, and want a 200ah battery, would you make 10 batteries comprised of single cells, or one comprised of parallel cells?

As the individual cell capacity gets larger, the cost penalty for extra BMS’ decreases and the physical size of the battery make it more attractive to use multiple single cell strings.

The actual chemistry and the way cells react in parallel doesn’t change. There is no chemistry or safety reason not to parallel large capacity cells, it is just often more practical to use multiple single cell strings.

 

[niktak11]

The only difference is the terminal connections between parallel cells.

I use powerplus batteries which are comprised of cylindrical cells in parallel, i think these give the easiest to understand layout for how parallel cells operate.

If you picture 10 cylindrical cells connected by two plates, the cell capacity is multiplied by the number of cells. If one cell fails short circuit, the other cells supply current and the shorted cell fuses. The capacity is reduced by the loss of the single cell.

If you have cylindrical 20ah cells, and want a 200ah battery, would you make 10 batteries comprised of single cells, or one comprised of parallel cells?

As the individual cell capacity gets larger, the cost penalty for extra BMS’ decreases and the physical size of the battery make it more attractive to use multiple single cell strings.

The actual chemistry and the way cells react in parallel doesn’t change. There is no chemistry or safety reason not to parallel large capacity cells, it is just often more practical to use multiple single cell strings.

I'd buy 200Ah cells

 

[aussie desert]

I'd buy 200Ah cells

Your call - but that’s no different to saying you wouldn’t use two 280ah batteries in parallel and would just buy 560ah cells.

Not many share your view.

 

[Skypower]

The only lithium cell you MIGHT not be in danger of peripheral property damage without a BMS is an is a LTO cell. Doesn’t mean that you won’t kill the cells from over/under charge without voltage protection. With our large LFP cells I don’t see the cost to advantage to running parallel cells because of the reasonable price of a BMS. I feel the redundancy of not losing the whole bank because of one cell or BMS a huge advantage. My first experience with solar was a 2P 16S with a single big Daly BMS and sure enough it failed. Totally down and out.

 

[niktak11]

Your call - but that’s no different to saying you wouldn’t use two 280ah batteries in parallel and would just buy 560ah cells.

Not many share your view.

If I wanted 2800Ah then I'd build 10 individual packs of 280Ah each with their own BMSs

 

[robbob2112]

Yes - this - ^^^^^^^^^

I would build 10 batteries from whichever cells made what I wanted and put a BMS on each pack then parallel them all with their own class T fuse each.

and I just realized I am currently charging 4 x 12v 314Ah batteries with only one of them having a BMS - the cells are brand new and all have the same Ir and the same voltage out of the box.... and for the last 10 hours of charging at 50amps they have stayed in lock-step with each other...

But now that I am getting close to fully charged I am thinkingI will DID - break them into separate 12v batteries and charge them separately so the BMS can cut things off when full or when the inverter/charger sees the full voltage.

Same time - safer - watts in are watts in....

And once they are fully charged I'll do capacity tests on them as 12v packs

 

[Hagger]

Sorry I was posting in the wrong subfoum. Will ask my questions in a new Thread.