Extend 130Ah with 190Ah?

[norbert66]

Currently I have 16pcs of 130Ah 3,2V LiFePo4 cells (BTS aliexpress) for my 48v bank.
I would like to increase the capacity. Sure I could buy another 16pcs 130Ah and build it up as shown in Wills video - first make single pairs and then serialize them together (assume it is named 2P16S)

Because of the better Ah/$ ratio I would like to buy 190Ah but I don't know if it is a good practice to mix different capacities.
... and maybe next year another 16pcs 190Ah (e.g. 3P16S).

Any suggestions?

 

[grizzzman]

You would need to keep them separate. If you parralled with the 130ah pac, you would only get 130 extra ah instead of 190 ah out of each added cell.

 

[norbert66]

Do I understand you correct:
Just break this down to a single cell. If I put 2 x 130Ah cells in parallel, I get 260Ah; doing this with 130Ah and 190Ah I still get 260Ah and not 320Ah?

 

[electric]

Grizzman is wrong. You can easily add cells in parallel of any capacity, no problem and you will get combined total capacity from all cells.

 

[gnubie]

The difference in amp rating will mean that each battery contributes a different amount of current so that the voltage on each battery falls in step with the other.

 

[grizzzman]

Oh ya?

 

[norbert66]

The difference in amp rating will mean that each battery contributes a different amount of current so that the voltage on each battery falls in step with the other.

Is your intention to say: if I pair a 100Ah cell with a 50Ah both will have the same voltage (at a certain time e. g. 3.2V). Charging with 9A will put 6A in the 100cell and 3A in the 50cell. Same for discharge.
Same if I put now 16 of those pairs in series to get 48V.

 

[electric]

Yes, that is exactly how it works.

 

[grizzzman]

Yes, that is exactly how it works.

Have you done this?

 

[grizzzman]

I would think that it's not likely that that type of system would have the same resistance? I guess if you kept out of the knees? I would think balancing would be problematic?

 

[electric]

I would think that it's not likely that that type of system would have the same resistance? I guess if you kept out of the knees? I would think balancing would be problematic?

I'm an EE with 30 years of experience, last 10 of which I work exclusively with large LFP batteries, designing battery systems for various industrial applications. So, yes, I have done this many times.
This relationship between cells is governed by Kirchhoff's circuit laws, if you study them you'd see that difference in resistance is exactly what makes it work. If resistance was the same across different AH cells, then it would not work. Difference in resistance is what makes the current split in fractions, allowing each cell to contribute it's own amount.
There are some nuances here, of course. Resistance changes over time during charge/discharge cycle, so current split can also change during the cycle. There may be periods when one cell takes a bit more or a bit less than ideal split, but on average they will always arrive to the same SOC, which means they contribute AH according to their size.

 

[electric]

I would think that it's not likely that that type of system would have the same resistance? I guess if you kept out of the knees? I would think balancing would be problematic?

Actually taking to upper and lower knees would improve the balance between parallel cells and it's not problematic at all. Same reasons as above, Kirchhoff's laws apply.

 

[Rider]

Is your intention to say: if I pair a 100Ah cell with a 50Ah both will have the same voltage (at a certain time e. g. 3.2V). Charging with 9A will put 6A in the 100cell and 3A in the 50cell. Same for discharge.
Same if I put now 16 of those pairs in series to get 48V.

Yes, that is exactly how it works.

I've always wondered about this. I figured for discharge you'd simply get battery A Ah plus battery B Ah for a combined total Ah, but wasn't sure you could charge them together without overcharging or undercharging one or the other.

 

[norbert66]

I've always wondered about this. I figured for discharge you'd simply get battery A Ah plus battery B Ah for a combined total Ah, but wasn't sure you could charge them together without overcharging or undercharging one or the other.

Just thinking: isn't the voltage which "handles" the over/undercharging - in other words over- and undervoltage? As the cells are parallel (not 2 individual battery banks in parallel ) they have the same voltage. Otherwise it would also not work with 2 cell pairs with same Ah (there is always a minor difference between each - otherwise we wouldn't need a bms in general)

 

[norbert66]

@electric:

So based on your experience it should be ok to make a cell pair of one 130Ah and 190Ah. And put then those pairs in seriell?

 

[electric]

@electric:

So based on your experience it should be ok to make a cell pair of one 130Ah and 190Ah. And put then those pairs in seriell?

Yes, we do it all the time. I don’t understand why this is such a controversial topic.

 

[norbert66]

Thanks for all the answers. You helped me so far.

 

[electric]

Just thinking: isn't the voltage which "handles" the over/undercharging - in other words over- and undervoltage? As the cells are parallel (not 2 individual battery banks in parallel ) they have the same voltage. Otherwise it would also not work with 2 cell pairs with same Ah (there is always a minor difference between each - otherwise we wouldn't need a bms in general)

Parallel cells are voltage locked forever, so it’s not possible to overcharge or over discharge one without the other. Their SOC will always be the same percentage, and they will contribute amps based on their relative capacity. This is why it’s best to parallel first, then go in series, for a single series string of multiple parallel blocks. Just watch for symmetry in connections. Make diagonal take off points, etc.

 

[grizzzman]

Thanks for the correction.