Headway 10Ah cell problem

[TheThirdWheel]

Hello all,

I recently put together a 24S2P (72V, 20Ah) LiFePo4 pack made up of 10Ah Headway cells for an e-bike project. It is connected to a BMS that has performed flawlessly in the past.

After the pack (about 50% charged) was installed into the bike, the bike worked great. Then I charged it, and that's when the problem began.

The controller switched on and the Cycle Analyst switched on very briefly but then switched off again. I plugged in the charger and flicked on the power. The charger said the pack was charged, so I switched everything on again and this time the CA stayed on, however I suffered major voltage sag (as low as 60V) when applying the throttle on the UNLOADED wheel.

Knowing the problem was most likely a BMS issue, I accessed the pack and started testing the voltages at the balance leads. 2 banks of cells were low. One was ~2.5V, the other was 0.3V. Every other bank read a healthy ~3.65V. There was also a clear liquid around the low voltage cells, which smells a bit like exploded capacitor...Upon removing the 2 cell banks (2 banks of 2 parallel cells each), I separated them into the 4 separate cells and began testing voltages. Both the cells that were in the 0.3V bank read under 1V each. I also noticed under close observation that the silver cap/seal under the main positive terminal of one of them had lifted. It looks like the cell had blown, died, and taken down it's parallel counterpart with it.

Upon testing the other bank, one cell was fine, while the other was at 2.5V, however under the miniscule load of the multimeter, I could see the voltage plummeting down until it settled at 0.8 volts!
I tried charging the low 1 volt cells but my charger doesn't even recognize a connection (insufficient initial voltage). When I charged the 2.5V cells the charger immediately said the cell was charged, but once again I monitored it's voltage sagging right down under the load of just the multimeter. I assume it's dead since it can charge/discharge in less than a second.

I thought these Headway cells were pretty durable and quality, but for one to just blow its top on a 2C discharge makes me feel ripped off.

I know that a bad cell will drag down any other cell in parallel with it and the BMS can do nothing about it, but why did the next cell in SERIES also die?

My BMS is wired correctly, if it wasn't the e-bike wouldn't work.

Has anyone else had LiFePo4 cells just up and die (my Headways have less than 20 cycles on them and haven't been abused).

What could be the cause of this problem?

Thanks everyone.

 

[dnmun]

how do you know the BMS is wired up correctly? did you buy the BMS new and add it to this pack or did it come with the pack?

the cans get overpressured and blow the seals when they are discharged under load with no free ions to transport current. there are no free ions below about 2.1V

 

[TheThirdWheel]

Hello dnmun and thanks for the reply!

The BMS was purchased separately from the pack. This pack was used in an earlier e-bike project, also as a 24S2P, but a different shape. The BMS wiring however, remains the same. The balance leads connect to the positive of each group of cells and the very last one connects to the negative of the battery + terminal.

I know the BMS works because if I detach any of the balance wires to a single cell group, the BMS cuts power to the controller. In it's original shape, the pack always stayed balanced with cell groups within 0.01V of each other.

I always manually charge every group of cells to 3.65V with a single cell charger before assembling them into a pack, just to make sure the pack is balanced initially. Then the bulk charger and BMS do the rest. Time consuming, but safer.

If the cell died of manufacturing causes etc, would it simply not charge, or could the charging be what blew the cell? e.g. I know you should never try and charge dying LiPo batteries.

Thanks!

 

[dnmun]

these are not lipo cells. they are headway cans. if the BMS shuts off when you remove one of the sense wires then it should have shut off for the LVC that killed those cans. the balancing shunt transistor may be shorted on those channels which would drain them down but the BMS would be shut off too so the battery would not function.

 

[TheThirdWheel]

Yes the BMS did shut off when it detected low voltage on the bad cell group. After a few minutes charge it allowed current again but quickly cut off the main power shortly afterward when the throttle was applied.

However, I'm fairly sure I have worked this out.

I don't know why one Headway cell up and died, but I know how it caused the death of the others.

The cell in parallel was brought down by the faulty cell, and the BMS could do nothing about it. It's designed to save the majority of the pack, not a single cell.

The next cell in series was affected by shorting itself through the "insulating foam" I had packed the battery box with. When one cell ruptured, it spewed electrolytic fluid into the foam surrounding the pack. Electrolytic-contaminated foam conducts! My multimeter showed not only continuity between the battery and the foam/bike frame (with the lowest ohm reading being nearest the affected cells), but also gave a full voltage reading when I put one probe on the frame (or soaked foam) and anywhere on the battery pack. Somehow, the battery was conducting to the foam and frame. A small leak, but a leak nonetheless.

I picked up on this when a freshly charged cell that was then placed into the contaminated foam didn't show full charge within a minute after connecting my single cell charger. I figured that current was leaking from the cell somehow without it being connected to any other cell, thus almost charging infinitely.

After replacing the 3 cells, and replacing all the foam in the battery box, the multimeter shows zero electrical continuity between the cells, foam, and frame. The BMS is reinstalled, and the pack has balanced perfectly. Hopefully it will stay that way.

Thanks!

 

[dnmun]

so how do you know that these other cans are dead? if you did not push current through them when they were discharged below the LVC then they should be undamaged. you said you had a BMS that would cut off when cell voltage dropped so they would never have been below LVC under load.

not sure where you got your theories about how a BMS functions but it is designed always to protect the individual cells.

i found it not unusual at all for one pouch in a parallel group to puff up from being over discharged. that appears to be your problem, but not much info to go on.