0
\$\begingroup\$

enter image description hereI have four 12 V, 50 amp hour lifepo4 batteries running in series, so 48 V. I need to increase the capacity of the battery bank. Can I get four 100 12 V amp hour batteries and create 4 sets of 150 amp hour battery banks by running a 50 amp hour and 100 amp hour in parallel. Then running those 4 banks in series to get 48? Will I have any issues in the long run? If so should I just buy 8 more 50 amp hour batteries instead?

Batteries are only 4 months old, very lightly used.

\$\endgroup\$
10
  • 3
    \$\begingroup\$ Please draw a diagram of how you plan to connect everything. From your description it should be fine. Make sure you match the voltage of the batteries before connecting them in parallel. \$\endgroup\$
    – winny
    Jun 11, 2021 at 18:59
  • 2
    \$\begingroup\$ How old and/or how many cycles on the existing batteries? In general, when constructing packs, it is best to use batteries of the same type and age and capacity. But what you propose is probably OK for LFP batteries as long as the existing ones are relatively new. If it were me, I might just buy more of the exact same batteries for consistency's sake. \$\endgroup\$
    – user57037
    Jun 11, 2021 at 19:06
  • 2
    \$\begingroup\$ Yes you can gang old with new but the net result must be each pair must match all 4 pairs in series within tight tolerances. For both Ah and ESR. You can test each 50Ah now to see how mismatched they are . If more than than the tolerance of the new, you must have a BMS to handle the charge discharge energy tolerance difference to absorb The differences in watts for each pair. \$\endgroup\$ Jun 11, 2021 at 19:22
  • 2
    \$\begingroup\$ When connecting different lithium batteries in parallel, be sure to have a dry fire extinguisher handy. Wear a flame-proof suit and welder's helmet. Use a very long stick to make the final connection. Do it outdoors, away from flammable objects. Other than that, no problem. \$\endgroup\$ Jun 11, 2021 at 21:43
  • \$\begingroup\$ @mkeith I have updated a picture \$\endgroup\$ Jun 12, 2021 at 13:29

2 Answers 2

0
\$\begingroup\$

Your circuit is pretty much legitimate, as long as connecting 12V LiFePO4 in series is legitimate in the first place.

For most 12V LiFePO4 batteries the manual explicitly states that series connection is not allowed. It is not that it will not work, it will need a pretty unusual type of battery management system and will be of lower reliability anyway.

Of course, the batteries should be of the same chemistry, connected as symmetrically as possible and should be as similar as possible, ideally the same brand and model.

\$\endgroup\$
-1
\$\begingroup\$

Imagine two capacitors in series initially charged with a load.

Which cap decays faster? The smaller one.

Which cap charges up past full voltage faster? The smaller one

BMS units are intended to shunt current on discharge with current on the stronger battery so it appears as weak as the weakest cell in order to prevent the weakest cell from getting damaged from undervoltage damage. Then during charging shunt the weakest cell to bypass current from exceeding the CV max while the stronger batteries catch up.

Normally 6cell batteries are matched <0.1% when new and the CV phase allows the current to decline while the stronger cells try to catch-up in series.

What happens without a BMS with external mismatched packs is the weaker packs accelerate faster in aging in a runaway condition from under and over charge, so you never want battery packs mismatched for Ah or ESR for longest life cycles.

This also means choosing to use less CV and cutout voltage reduces capacity but total life cycles increase dramatically. (E.g. 10x for 50% capacity utilization re Battery university, your mileage may vary)

It also means unless you have dozens of cells , when a string or module has 1 dead cell and the others are near end of life, it’s just a short matter of time before the next cell is dead and then the next etc. so batteries are like Caps and must be balanced as they perform as well as the weakest link.

To prove the point here, I modelled your 50Ah batteries to show how each battery gradually changes with respect to the others. It uses a comparator to decide when to charge 10A and discharge 10A with a simple battery model initialized to 12.0V then counts each cycle displayed on LEDs in binary. The voltage is sampled every second and the run time shows after about 50 cycles the capacity difference from Vmax has shifted by a spread of about 0.5V out of 14V. Eventually when one pack exceeds absolute max or min, degradation accelerates. this proves the point that a master BMS is needed to track all packs and need to be charged separately to same voltage after 50 cycles. Here is the simulation.

\$\endgroup\$
3
  • \$\begingroup\$ This is nothing new. BMS’s are essential for series strings solacity.com/how-to-keep-lifepo4-lithium-ion-batteries-happy and not just individual BMS \$\endgroup\$ Jun 12, 2021 at 17:21
  • \$\begingroup\$ While you are doing your drawing consider using 2 seperate 48V battery strings, a 50A and a 100A. Use seperate BMS for each string as each cell needs to be equalized without a tag-a-long. If you can get a BMS that will do both strings all the better. \$\endgroup\$
    – Gil
    Jun 12, 2021 at 17:29
  • 1
    \$\begingroup\$ Parallel packs won’t change the result of Vmax/min of the simulation when each of 4 pairs is put in series. FWIW @Gil. It just changes C by x3 and reduces ESR same roughly. So current sharing skews with ESR*C = T time constant \$\endgroup\$ Jun 12, 2021 at 17:33

This site is temporarily in read-only mode and not accepting new answers.

Not the answer you're looking for? Browse other questions tagged .