2 cells reading ~.7 volts

[SlimyLemur56]

In feb 2020 I bought 4 120ah cells off aliexpress. Pretty sure at the time I tested each cell and the voltages were nominal. It came with cheesey m4 terminals which stripped immediatly (actually the screw holes were tapped badly and it stripped just trying to get the screws it came with to go in) and I never used the battery once, never put a single charge or discharge through it.
I finally got around to trying to do something about the stripped terminal and I made all the preparations, made some flat bottom drills and retapped the hole to m8 and put a stud in it. Then I rechecked all the cells and realized that while 2 of them read 3.x volts, 2 of them read .6xx volts.

If a 3.2v cell is reading .6/.7 volts, is it damaged beyond repair? What I'm really wondering is should I even bother to try to recharge it? Would it be dangerous to do so?
Should I just accept they are toast from sitting unused for 1.5 years and fork over the cash for a new battery, or is there some hope there's a way to recharge them and get some low duty use out of it? If there's any hope, how would I safely attempt to put a charge in them? The voltage difference between the cells is so high I'm not sure if I should trust a (daly) bms to balance them while charging.

 

[sunshine_eggo]

This guy says cells he had that had discharged to 0.6V recovered:

Somehow killed my pack

I have a 4 cell 12v 280 AH pack with the Electrodacus BMS, hooked up to 380 watts of solar. Been in service a few months without issue. We had not been using our RV for about a week and came back to a dead pack. The only load on the pack was our fan and our Dometic Danfoss fridge which the...

diysolarforum.com

If you do attempt recovery, I would charge the cells individually with a hobby charger or adjustable power supply and assess how they perform.

 

[John Frum]

This guy says cells he had that had discharged to 0.6V recovered:

Somehow killed my pack

I have a 4 cell 12v 280 AH pack with the Electrodacus BMS, hooked up to 380 watts of solar. Been in service a few months without issue. We had not been using our RV for about a week and came back to a dead pack. The only load on the pack was our fan and our Dometic Danfoss fridge which the...

diysolarforum.com

If you do attempt recovery, I would charge the cells individually with a hobby charger or adjustable power supply and assess how they perform.

I have head that its best to charge at a low c rate until you get above 2.5 volts.

 

[sunshine_eggo]

I have head that its best to charge at a low c rate until you get above 2.5 volts.

That's a good point. Even the smallest current (<0.1A) should provide a fairly significant voltage increase in a relatively short period of time.

 

[SlimyLemur56]

I don't have a charger I can set to something other than 12v. I have a bench power which can do arbitrary voltage and also constant current, although to my knowledge it cannot do constant current with also a max voltage.
Can I just set it to constant current (some small amount) and watch it with my eyes until it gets into the high 2v range?

 

[John Frum]

I don't have a charger I can set to something other than 12v. I have a bench power which can do arbitrary voltage and also constant current, although to my knowledge it cannot do constant current with also a max voltage.
Can I just set it to constant current (some small amount) and watch it with my eyes until it gets into the high 2v range?

Most bench power supplies do cc/cv.
Is yours peculiar?

 

[SlimyLemur56]

I've never used the feature, it's a combo soldering station/heat gun/power supply. I wanted to ask early but that meant before I get home to look. But looking up similar products on amazon I think it actually does not have CC mode at all. Maybe I can borrow something from someone.

 

[sunshine_eggo]

Can you set it to 3.0V and 0.1A?

Or you can do as you proposed. Set to 0.1A and check on it.

 

[time2roll]

I have a bench power which can do arbitrary voltage and also constant current, although to my knowledge it cannot do constant current with also a max voltage.
Can I just set it to constant current (some small amount) and watch it with my eyes until it gets into the high 2v range?

0.1 amp overnight should be fine.

Should not over volt until you have at least 100 amps in the cell. And if it does, the cell is shot anyway.

 

[RCinFLA]

If left for extended time at less then 1v the cell will grow dendrite shorts. By charging at lower current if they have dedrite shorts the voltage will not rise and you will not be putting in so much current the cell gets hots and bursts.

If cell voltage rises with low charge current then once it gets above 2.5v you can go to higher charge rate.

 

[SlimyLemur56]

Borrowed a bench power supply and learned about how cc/cv actually works, which from my above comment I clearly didn't know.
Two cells were at .6v and one was at 1.5v

I have been charging the following way:
0.5-2v => 0.06a
>2v => 0.2a
>2.6v => 0.8a

I stopped charging each around 2.7/2.8v to move to the next one as a first pass (kinda testing at the same time).
I hope that was a reasonable way to do it. I got all cells over 2.6v, so I guess that is promising.

Past 2.7v, my plan (unless told otherwise), is to hook the 3 in parallel and charge them together at 1.5a (power supply only goes to 2a and I'm charging off dinky 22awg wire) with voltage limited to 3.2 (the current voltage of the 4th cell), and leave them charging like that for 2-3 days or however long it takes. Once they are all at 3.2v I will hook all 4 cells to the bms and hook the battery to a standard lifepo4 battery charger. Then do a drain test through the bms and recharge while watching the balance. Sound good?

If I got them >2.5v does that mean there are not dendrite shorts? Is that permanent or will it break down?
If this all works and I get it behaving like a battery (maintaining balance within a level manageable by the bms), is there any gotchas I should be aware of for long term usage/maintenance?

 

[John Frum]

Borrowed a bench power supply and learned about how cc/cv actually works, which from my above comment I clearly didn't know.
Two cells were at .6v and one was at 1.5v

I have been charging the following way:
0.5-2v => 0.06a
>2v => 0.2a
>2.6v => 0.8a

I stopped charging each around 2.7/2.8v to move to the next one as a first pass (kinda testing at the same time).
I hope that was a reasonable way to do it. I got all cells over 2.6v, so I guess that is promising.

Past 2.7v, my plan (unless told otherwise), is to hook the 3 in parallel and charge them together at 1.5a (power supply only goes to 2a and I'm charging off dinky 22awg wire) with voltage limited to 3.2 (the current voltage of the 4th cell), and leave them charging like that for 2-3 days or however long it takes. Once they are all at 3.2v I will hook all 4 cells to the bms and hook the battery to a standard lifepo4 battery charger. Then do a drain test through the bms and recharge while watching the balance. Sound good?

If I got them >2.5v does that mean there are not dendrite shorts? Is that permanent or will it break down?
If this all works and I get it behaving like a battery (maintaining balance within a level manageable by the bms), is there any gotchas I should be aware of for long term usage/maintenance?

I suggest a self discharge test.
If you can spare the time charge them to 3.65 volts individually and leave them as individual cells for up to a month.
Even a week would probably give a good idea.
The cells should settle quite a bit in the first day and hardly at all over the rest of the month.
@RCinFLA would likely suggest you burn off the surface charge with a resistor.

 

[sunshine_eggo]

Personally, I would charge all cells individually to within 0.1V of each other, then parallel and charge all.

Impossible to definitively answer your future-prediction questions. It's all about how they perform.

As far as long-term usage/maintenance, just follow best practices with the caveat that some of your cells may have excessive self discharge. Long term storage may require additional precautions, e.g., knowing only what I know now, if I had to store those cells for > 90 days, I would discharge them to 40-70% SoC and then parallel them so the weak ones are supported by the stronger. Check voltage periodically. If voltage drops below 3.25V, put on a PS set to 3.25V.

 

[SlimyLemur56]

For a discharge test I charged them in parallel to 3.6 and have been watching their voltages for about a month and a half while working on other projects. I didn't think to number the cells until after charging them, when I stupidly lost track of which ones had over-drained, but looking at the data I'd assume it's cells 3 and 4.



What's the prognosis?
The difference in initial discharge rates seems pretty stark, but also seems encouraging that they are stable at 3.32v. The intended use for this battery would be using a daly bms, sustaining < .1 C (<10A (probably more like 6A)) for 8-10 hours or until the low voltage cutoff. Would the self drain differences between these cells be a problem for the balancer?

 

[time2roll]

I believe the full prognosis will be known when put in service and the capacity is confirmed after a few cycles. Need to set them in series with a BMS and see which cell(s) triggers the low voltage cut-off and what the voltage delta is at that point.

Self drain shown will not be any trouble to the balancer.... however if the cells really have capacity issues the balancer will not help much.

 

[John Frum]

For a discharge test I charged them in parallel to 3.6 and have been watching their voltages for about a month and a half while working on other projects. I didn't think to number the cells until after charging them, when I stupidly lost track of which ones had over-drained, but looking at the data I'd assume it's cells 3 and 4.

View attachment 77326
View attachment 77327
What's the prognosis?
The difference in initial discharge rates seems pretty stark, but also seems encouraging that they are stable at 3.32v. The intended use for this battery would be using a daly bms, sustaining < .1 C (<10A (probably more like 6A)) for 8-10 hours or until the low voltage cutoff. Would the self drain differences between these cells be a problem for the balancer?

Looks like all of those cells are bad.
The full resting votlage for a lifepo4 cell should be ~3.45 volts or higher at 30 days.

 

[RCinFLA]

In two weeks you should not have dropped below 3.400v on any of the cells. Ambient temp of cells matters if stored in warm environment.

What did you measure voltage with and how much confidence in its accuracy? It is very suspicious they all leveled off at 3.32v from day 29 to 49. That would mean they all lost about 25% of their capacity. Not likely in a month.

Cell self discharge rate spec at room temp is 3%/ month. It typically is less than this between 1.5% and 2%. It can go as high as 5-6% if stored at high temp.

Anything above 3.45 vdc is surface charge left over from full charging. You can bleed it off with 1 to 5 ohm resistor in less than 60 seconds. That should give you a more even starting point of about 3.45v per cell.

Unloaded test is only about leakage current and not correlated to battery capacity or internal series impedance.

I have four cells I did a 3 minute load test on two months ago. I left them at about 3.310v and just measured them. They are reading maybe a mV or two less than I left them two months ago sitting between 75 degs F and 78 degs F for the last two months.

 

[sunshine_eggo]

When referring to the large ~280Ah cells, yes, self-discharges below 3.40V may be a concern; however, I would not necessarily apply that to 120Ah cells.

I have some 40Ah CALB cells that will plummet to below 3.40V VERY quickly, and they rapidly settle to 3.3XX. After sitting for a year fully charged, they still delivered ~95%+ of their prior capacity while measuring 3.300V prior to discharge.

IMHO, this is a function of the cell's internal resistance. Smaller cells tend to be higher due to the decreased surface area, and their voltages tend to settle more rapidly.

They MAY be bad. They MAY be fine. I would run at least a few charge/discharge tests on the completed battery and look for outlier behavior.