xiaomi battery problem

[Blisk]

I have this xiaomi battery with 30 18650 cells connected together by 3 cells. I tested all and removed dead cells then replaced them with fully charged cells. BMS was disconnected and only cells were welded together, except the last 3 cells. When I try to weld it together it makes a short circuit and starts to heat up drastically (picture with white alligator). I remove that one cell and leave it for an hour after that I checked and some were also so hot I couldn't even handle it (5 of them where white alligator is). I removed those and one died others I charged again.
Don't know what is going on, because cells was only welded together and nowhere connected.
Any Idea?

[Phil1977]

I suppose you carefully checked the Ni-strips so nothing got shorted.

Beside that it´s always risky to connect cells of different type/age and especially SOC. Is it possible you paralleled one old 18650 with a fully charged new one? If you just connect to 18650 with very different state of charge (SOC) then there very high currents start to flow...

And, beside that again, please NEVER use solder on battery terminals. The right tool, the spot welder, is visible on your pictures. Please ONLY use that and don't heat up the metall of any cell to solder temperatures. Even if soldering may work for some time you definitely damage the cell and risk any kind of future failure.

[Blisk]

Yes strips are not shorted, checked many times, even when this battery looks bad I am working on it for few months now and it doesn't work.
SOC you mean battery capacity. I tested all batteries and check capacity and weld together similar like from 2000 mAh to 2200 mAh.
I was soldering just last part, but now all are welded.

[Phil1977]

No, battery capacity is usually called State Of Health (SOH). State of charge is just how full the battery is charged.

Cells with heavily different SOC have a different voltage. So it would be a good idea to check that the voltage difference between two cells is e.g. <50mV before connecting them to each other.

Frankly spoken I don't think that battery pack repairs like you are doing it are worth the time and effort. If the pack has once got out of balance you should renew all cells. I know oppinions there differ, and some fancy active balancing systems may help somehow. But anyhow, a LiIon-cell-pack with differently aged and partially even soldered cells stays what it is: An incendiary bomb.

[Blisk]

thank you I will check that. Soldering batteries are now out of it.
So fully charged 18650 must be the same voltage?

[Phil1977]

After charging the cells should all have slightly less than the final charging voltage, usually around 4.1 to 4.15V. If one of the cells looses its voltage quickly then it probably is already broken.

But then you only know that the cells are matched for 100% SOC. For the whole operating range its pure gambling 

[Blisk]

After charging the cells should all have slightly less than the final charging voltage, usually around 4.1 to 4.15V. If one of the cells looses its voltage quickly then it probably is already broken.

But then you only know that the cells are matched for 100% SOC. For the whole operating range its pure gambling 

thank you I do that all the time, charge batteries, and after a few days if the voltage is significantly lower than others I try to empty and charge them again and then again if it is the same it goes to recycle bin.

[Phil1977]

That´s quite a good start, but it doesn't guarantee your pack is well performing.

Let me explain one example:
You have 5 cells in parallel, two with a low ESR, lets say 10mOhm, and three older ones with 30mOhm. They both have the same end of charge voltage and roughly the same capacity of e.gn 2200mAh on the cell-tester.

Now discharge the pack with let´s assume 50A while it is exemplarily moving an ebike up a mountain. First, the cells with the low ESR will deliver all the power because they have the lower resistance. They deliver more than 20A each, wich is around 10C and to much for most 18650s. They heat up due to the high current - their ESR goes down even further.
Then they go empty and their voltage rapidly drops while the 3 cells with higher ESR are still quite full. Now they must deliver the biggest part of the current, and because power losses are proportional to the current squared they have much more than double of the losses compared to a system where all cells are matched.

In summary the pack heats up and its lifetime is heavily compromised.

And even if you measure ESR of the cells with a dedicated meter it doesn't guarantee that the discharge curves are the same... In the end its always the same, don't mix different cells in one battery pack.

[Blisk]

thank you for great explanation I appreciate that and will search more about it.
My charger/tester also measure internal resistance of battery, so I put together the ones with almost the same values.