how to rescue lifepo4 battery pack with 1 dead cell

[racmaster]

situation:

- 4S lifepo battery 100ah each cell, connected in serie=13.6v battery

- 1 cell is semi/dead, holding voltage, but lost capacity to 20ah, probably overcharged before... it can be charged and discharged, but results in unbalanced battery pack and whole pack is therefore limited to 20ah...

possible solutions:

A: adding another 20ah cells in parallel with the broken-one, which means 5x20ah cell= 100ah and connecting this parallel pack to the serie with 3 remaining still good 100ah cells. of course, before paralleling them they need to be charged same SOC, voltage and balanced top or bottom... then vice versa with remaining 3 old cells...

B: removing the wrong cell and replacing with new cells connected in parallel, but as near as possible to actual real capacity of remaining 3 old cells. which means, lets say the 3 good old cells are aged and they are no more 100ah, but lets say 60ah of real capacity. so the 4th cell to be added to the serie pack should be built from smaller cells connected in parallel to reach the real 60ah capacity. of course again, first charged, then balanced and then connected together...

C: replacing the wrong cell with the bigger one, for example 110ah

pls consider these 3 possibilities and lets discuss which solution will be best technically, which economically and/or which practically pros and cons welcomed

[Ampster]

Option D. Build a new battery. Sorry, but without details about the cells, BMS, construction,, use case or form factor the three other options are hard to evaluate.

[azdave]

Some good general info found here that might not all apply to your case but it's a start.

https://www.orionbms.com/troubleshooting/replacing-cells/

[chrisski]

Assuming the other three cells are good, why not replace it with a like 100 ah cell?

I bought some 25 ah Topband cells. The first batch was from one company, maybe 24.4 ah capacity, the second batch was 25.5 ah capacity. I have 5 of the smaller ah and 3 of the larger ah cells in one battery.

I don’t think its a good idea to mix and match capacity cells, but its less of a bad idea if this battery pack is in a place that will not burn your house down if something goes wrong.

[Mike90250]

When your 100ah cell falls to 60ah capacity, it's very near end of useful life. Is it worth such intensive CPR on such cells ?

A better question, is how did the bad cell get damaged, why didn't the BMS protect it, and how to prevent damage in the future ? Before you buy new cells and do the same damage to them.

[racmaster]

I have 5 of the smaller ah and 3 of the larger ah cells in one battery.

I don’t think its a good idea to mix and match capacity cells, but its less of a bad idea if this battery pack is in a place that will not burn your house down if something goes wrong.

does the battery composed of different capacities, different cells, different internal resistance perform well? how about balancing issues?

[racmaster]

Sorry, but without details about the cells, BMS, construction,, use case or form factor the three other options are hard to evaluate.

there is no BMS. it is lifepo4, 100ah cell, use case is small loading about 20% DOD, load isnt more than 10% of max allowed cell current, or less. almost nothing...

do you need any other info to be able to evaluate the situation? what do you mean by form factor?

[chrisski]

does the battery composed of different capacities, different cells, different internal resistance perform well? how about balancing issues?

They were all 25 ah capacity Topband cells, but from different distributors. Its that when I got the Blue Shrink wrapped ones from BatteryPlus, they measured in at slightly less than 25 ah. When I got the green shrink rapped ones from someone else, they were slightly more than 25 ah. I don’t know the internal resistance. Kind of trying to say that even same cells from different distributors can be off.

I have a non-readable, non-programmable BMS. I don’t know about balance, but there is an algorithm built in to balance these cells.

Especially without a BMS, IMO its a bad idea to use the lithium cells, even if perfectly matched, and what you are planning is certainly not matched. The BMS has helped me a couple times. For me, these batteries are solely powered by solar, so when there was a bad connection and my solar panels died, it shut them off before they got too low. Also, the BMS shut the charging off in n over amperage situation. I don’t use this every day, but maybe once a week for 8 hours and I can’t monitor it for those times.

Maybe you have a situation you don’t need a BMS. That is a very narrow set of circumstances.

[racmaster]

chris, bms is another topic, lets not discuss that for this purpose.

i just want to know how your 24/25ah cells work together. how long do you have them, if there were some problems and so on.

what i think is, that even new battery has not the same cells. i tried many of them and all had serious differences in internal resistance, which of course must result in balancing issues during batt cycling which results in over charging and discharging issues. lets not mention bms, balancer, which will take care for it, but only for a while, sooner or later there will be the same problem. bms only masks this problem. and the longer it will be masked, the bigger it will became...

and here, we are getting to the point. lets talk about 5year old battery in a big appliance, for example car, or big offgrid. there will be pretty often situation, that one of those lets say 64 cells will be fault, or partially fault. what to do in such a situation? there must be better solution as buy new one... 4S solution may be buy new one, but 64 or more....???

what i want to say is, that i know the truth about best is connecting absolutely same cells, that's not big knowledge. everyone knows.

lets get another situation:

old 48v battery, 16P4S = 64 cells. any Ah, doesnt matter for this example... by cycling, some of them will get unbalanced, some of them will loose capacity naturally sooner, then others... but the battery still works. its not as new, but works. is smaller capacity, but works... and in this moment there are no equal cells anymore...

so lets say, the natural aged difference between majority of cells and the broken one is 10%. i think, that it must be better idea to take the 90% cell out and replace it with anything better than 90, even if it will be replaced by cell that is 95% if compared to the others in battery, the result should be better. and the cost will be 1/64 if compared to "solution" buy new battery...

now hopefully it is more clear what way im thinking.

[chrisski]

chris, bms is another topic, lets not discuss that for this purpose.

i just want to know how your 24/25ah cells work together. how long do you have them, if there were some problems and so on.

I have no problems to report from my two 25 ah 24 volt batteries that have the mismatched cells. Two battery packs are built into a 50 ah battery. I run this at a constant 8.3 amps draw, which I guess is around .17C for three to ten hours. More than I could run lead acid batteries. I had done the same work on a 25 ah battery, which is closer to a .3 C load for. Month, but added 25 more ah to take me through three where I could plug this in for three hours.

I have used these for maybe five months now, once a week. I have no real way to monitor them except the Ali Monitor and the Victron SCC app. I have no way to tell how far apart the cells drift. I have the cheap 15 amp charge, 30 amp discharge Daly BMS on each battery.

I solar charge these only with a 15 amp Victron SCC. The load is a crockpot that pulls around 220 watts from the batteries, which takes into account inverter. This stays on the entire time the Battery is used. I have 900 watts of panels, but the Victron SCC will only pull around 15, although the panels can push about 35 amps. Intent of this overpaneling is to use this on sunny days where the solar generator can pull 300 watts, while the same amount or more worth of energy is replaced by the panels. For me to do this and leave it unattended, I point 300 watts of panels to the west, 300 to the East, and 300 to the South, and this takes care of it.

[Ampster]

....... what do you mean by form factor?

Cylindrical, prismatic or pouch?
You have since answered my other questions and others have expressed opinions that I would not disagree with.

[BroncoLera]

AFAIK.....
The safe and proper way to rescue a battery pack like this with one bad cell...........

Is to replace that cell.

Trying to "rescue" / "resuscitate" a dead cell seems pointless if not counter productive.

But if for some reason you are undeterred, try an ACTIVE balancer as a last resort.

[bscott.raven]

AFAIK.....
The safe and proper way to rescue a battery pack like this with one bad cell...........

Is to replace that cell.

Trying to "rescue" / "resuscitate" a dead cell seems pointless if not counter productive.

But if for some reason you are undeterred, try an ACTIVE balancer as a last resort.

Before you replace the "bad" cell consider an active balancer. I have two 280Ah LiFePO4 batteries I built myself. I run them in parallel. I have a BMS for each battery, but even with a BMS one of the 8 cells became "weak": this one cell would reach both the charge and discharge cutoff voltage well before the other cells. I installed an active balancer and within a few days the "weak" cell has come into perfect balance and does not experience the over/under voltage problem. An active balancer is less expensive than replacing a bad cell, and avoids the problem of having cells of differing age (# of cycles).

[azdave]

I tried this and it worked. "Weak" cell is now staying in balance.

What worked? Replacing the bad cell or using an active balancer? Both are mentioned in the post you quoted and replied saying "It worked"

[bscott.raven]

What worked? Replacing the bad cell or using an active balancer? Both are mentioned in the post you quoted and replied saying "It worked"

Thanks for pointing out my ambiguity. I modified my original post.