Strange cell behavior

[jtvt]

I've got a 280ah cell with some weird behavior.

- When charging, it's voltage leads the other cells. I have absorb set to ~56.1 and balancing set at 3.51v. It reaches this before the other cells which do range with lowest at ~3.46v at that point.
- When it hits the absorb/balancing voltage area, it's voltage drops to ~3.44v while the others balance out at 3.51v. I don't know exactly when this happens. I don't know if the balancer triggers this exaclty when it comes on. The voltage will not rise even with extended absorb.
- After charge all the cells settle to normal voltages
- When discharging, the cell's voltage is the same as all the other cells.

I've taken the busbars off, cleaned, etc and reinstalled. I've also tested with a voltmeter on the posts and it jives with what the BMS says.

The bank is 3 yrs old. In the past, that cell sometimes lagged a bit when charging, but would balance out with the rest of the pack.

Anybody seen something like this?

 

[jtvt]

Having written that down and thought about it more.... I've only tested with the voltmeter when the cell is in the 3.44v state. I'll need to test it when charging and the BMS says the cell is close to 3.51. I've replaced this blockmon before. I'm wondering if this is another bad blockmon and is reporting inaccurate voltage up until the balance comes on (which heats up the blockmon)....

 

[jtvt]

Further testing: I parked my voltmeter on the cell posts and BMS readings are accurate. The cell hits 3.52 and the balancer comes on. Target is 3.51. This coincides with absorb (or CV). Other cells hit the balancing voltage and the whole bank balances out to 3.48-3.51v. At this point, charge amperage is ~6a across 4 48v strings. So real low. While still in absorb, some cells are still in balance mode, but at low balance amperage. Then still in absorb, the cell slowly starts to drop voltage. I watched it drop to 3.46v.

So it seems like the cell isn't seeing voltage/amperage to maintain state. It should be sitting at ~3.51. I need to check those busbar connections. But I also wonder if there could be some sort of anomaly where input amperage is completely siphoned off by the balancer so a cell might not see enough juice to maintain state. Maybe compounded with this cell having higher resistance than the others.

 

[sunshine_eggo]

Is this a Batrium?

If not, why on earth would you not start balancing at 3.40? 3.51 is way too late.

If it is a Batrium, you should probably work to get the battery top balanced via autolevel.

Anyway, abrupt voltage swings with changes in current indicate increased resistance. Since you've cleaned the contacts and presumably re-torqued correctly, this pushes the concern to the cell itself. Increased internal resistance in a cell is a sign of degradation and a precursor to failure.

Have you swapped the cell with another to see if the issue moves with the cell?

 

[jtvt]

Is this a Batrium?

If not, why on earth would you not start balancing at 3.40? 3.51 is way too late.

If it is a Batrium, you should probably work to get the battery top balanced via autolevel.

Anyway, abrupt voltage swings with changes in current indicate increased resistance. Since you've cleaned the contacts and presumably re-torqued correctly, this pushes the concern to the cell itself. Increased internal resistance in a cell is a sign of degradation and a precursor to failure.

Have you swapped the cell with another to see if the issue moves with the cell?

It's Batrium WM4 with blockmons.

It's been up and running for almost 3 years and I haven't needed to balance anymore than what the 3.51 setting provides. Add to that, I'm leary of a potential bug in blockmons during balance. I've had a few go bad and it's always during balance. I currently have the heat/amp limits throttled back which seems to be the recommended strategy to avoid issues.

Never used autolevel because batrium docs say don't use it on lifepo.

It's on my list to move the cell.

 

[sunshine_eggo]

It's Batrium WM4 with blockmons.

It's been up and running for almost 3 years and I haven't needed to balance anymore than what the 3.51 setting provides. Add to that, I'm leary of a potential bug in blockmons during balance. I've had a few go bad and it's always during balance. I currently have the heat/amp limits throttled back which seems to be the recommended strategy to avoid issues.

Never used autolevel because batrium docs say don't use it on lifepo.

It's on my list to move the cell.

Batrium is mostly wrong.

Batrium has shoe-horned their system into LFP use. IMHO, it's actually not a very good BMS for LFP in general. LFP charges like lead-acid, and Batrium doesn't accommodate that.

There's no reason for a knowledgeable user to NOT use auto level to improve top balance. It's just much less effective than it is for 3.6-3.7V Lithium chemistries with high voltage to SoC correlation.

However, in your case, since you've confirmed voltages and connection quality, this sounds like cell degradation.

Recommend you:
discharge to ~80% SoC
Deliberately engage a medium-small constant load (.25C or so), I1 and hold for 2 minutes.
Record all cell voltages, V1
Engage a very high load (as high as you can), I2 and hold for 2 minutes.
Record all cell voltages, V2
Return to normal operation

You can then calculate the internal resistance of each cell as follows

R = (V1-V2)/(I2 - I1)

If the cells exhibiting "odd behavior" are also the cells with the highest IR, then you either have connection issues you missed, or you have degraded cells.

If you've been charging to 3.51V/cell daily and holding that as a "float" voltage for as long as you get charging, my vote is for degraded cells.

 

[jtvt]

Strangely, I've been trying to change the bypassvolt setting from 3.51 to 3.42. I get it to change in the software, use the passcode, and save but the WM4 still thinks it's 3.51. I restart the toolkit and it shows 3.42. So the setting sticks at least in the toolkit, but never makes it to the WM4. I've rebooted the WM4 and retried and still no dice.

 

[jtvt]

Strangely, I've been trying to change the bypassvolt setting from 3.51 to 3.42. I get it to change in the software, use the passcode, and save but the WM4 still thinks it's 3.51. I restart the toolkit and it shows 3.42. So the setting sticks at least in the toolkit, but never makes it to the WM4. I've rebooted the WM4 and retried and still no dice.

Had to resync.

 

[jtvt]

This problem went away. It seems to happen when the bank stays below the balancing voltage for a long period. In this case about two weeks of cloudy weather with the SoC bouncing around 30-80%. Now that there is more (sun) frequent balancing, the problem has completely disappeared. I have 4 strings at 48v and all cells balance out to 3.51 in 25 minutes of absorb.