Replacing cells in a Pylontech cell pack

[APB2025]

I've enjoyed reading the informative posts on this forum so time for my first posting:
As anyone tried replacing cells in a Pylontech cell pack?
I picked up 2 US2000c's destined to be scrapped due to swollen cells. When gently charged (52.7v @ 5A) I could see at least 1 cell position in each pack going OV and when taken out of the pack casings, I could see that they were the cells that were swollen. As an experiment I took a pair of "good" cells from a pack in the 2nd battery to replace the swollen cell and the other cell in that position (just in case it had been compromised) with the aim of creating a "good" pack. After bringing all the cells in the pack to the same voltage (using a Neey active balancer), I put the pack back in the battery and started charging. Within a few mins the replacement cells had OV'd. Removing the packs from the battery and using a bench PSU/Neey to do the charging in case its was a battery BMS problem gave the same OV result.
I wondered if the replacement cells had different charging characteristics (as they had different batch numbers from the removed cells and probably had a different usage history) so I replaced them with a pair of cells from another pack from the same battery as the cells being replace (i.e. all now had the same batch number and usage history) but got the same OV result.
I think I've eliminated BMS issues and cell variability - anyone have any theories as to why this didn't work as expected?

APB

[Krill]

Why not just run the 14 good cells in each battery with a 14s standalone BMS?

[Stinsy]

I've enjoyed reading the informative posts on this forum so time for my first posting:
As anyone tried replacing cells in a Pylontech cell pack?
I picked up 2 US2000c's destined to be scrapped due to swollen cells. When gently charged (52.7v @ 5A) I could see at least 1 cell position in each pack going OV and when taken out of the pack casings, I could see that they were the cells that were swollen. As an experiment I took a pair of "good" cells from a pack in the 2nd battery to replace the swollen cell and the other cell in that position (just in case it had been compromised) with the aim of creating a "good" pack. After bringing all the cells in the pack to the same voltage (using a Neey active balancer), I put the pack back in the battery and started charging. Within a few mins the replacement cells had OV'd. Removing the packs from the battery and using a bench PSU/Neey to do the charging in case its was a battery BMS problem gave the same OV result.
I wondered if the replacement cells had different charging characteristics (as they had different batch numbers from the removed cells and probably had a different usage history) so I replaced them with a pair of cells from another pack from the same battery as the cells being replace (i.e. all now had the same batch number and usage history) but got the same OV result.
I think I've eliminated BMS issues and cell variability - anyone have any theories as to why this didn't work as expected?

APB

Not sure I understood correctly because that really sounds like a BMS issue to me!

[APB2025]

Tx for the replies.

Why not just run the 14 good cells in each battery with a 14s standalone BMS?

I forgot to mention that the reason for trying to repair the battery was to able to add it to my existing stack of 2000's - so it needs to be 15s to fit in with the others.

Not sure I understood correctly because that really sounds like a BMS issue to me!

Unfortunately as the cells go OV when charged without using the battery BMS, its not the BMS causing the problem - though its not to say that I won't find problems with the BMS at a later stage (if I get that far!)

My current theory is the cells may have been thermally damaged when I removed them from the pack circuit board. As the battery tabs are folded over each other before being soldered together and to the circuit board during manufacture it took a lot of heat to remove them so its quite possible that I've overheated them. So far I can't see a way of mechanically removing them from the board without destroying the battery.

APB

[Countrypaul]

There was a post recently I think by Nowty concerning somewhere that repairs Pylontech batteries I believe, you might find them helpful if you cantact them. I will try and find a link.

Found it: https://camelot-forum.co.uk/phpBB3/view ... ech#p55461

[Marcus]

I too don't quite understand your description of what was going on:- you say you charged 'gently' at 52.7V and 5A? That's 3.51Vpc which is well charged for lifepo, so possibly you mean you set the max voltage to 52.7v before connecting to the battery? If so what was the battery's initial voltage? What was the lowest cell voltage?

5A may be a gentle charge for a cell operating In it's normal range: I.e. between 2.5 and 3.65v, >0°C and not swollen. But I'd consider it to be a bit aggressive if the cell is over discharged and swollen.

For reference:
I recovered some pylontechs that had discharged down to 17v; some cells were <1v and there was some swelling. I charged gently (0.05A) until all cells >2v and then at 0.1A until they were all >2.3v. Then put the lid back on (as the case 'clamps' the cells to prevent swelling - up to a point) and could charge more aggressively via the bms and external terminals. 0.05A doesn't sound like much, but at <2v the voltage rises fast.

I may have been wildly overcautious, but from what I'd read they don't take charge well once they're below their minimum voltage. I'm only guessing, but if you were charging overdischarged cells at 5A that might be a problem.

Otherwise your theory of excess heat might be the issue

[APB2025]

There was a post recently I think by Nowty concerning somewhere that repairs Pylontech batteries I believe, you might find them helpful if you cantact them. I will try and find a link.

Found it: https://camelot-forum.co.uk/phpBB3/view ... ech#p55461

Tx for the link. I have used Servtec in the past and as a satisfied customer can recommend them. The person I got the batteries from said that Servtec had already looked at them and said they were not repairable which was why they were getting rid of them.

so possibly you mean you set the max voltage to 52.7v before connecting to the battery?

Correct

If so what was the battery's initial voltage? What was the lowest cell voltage?

Initial values for the cells before charging were around 3.3-3.4V (I used the balancer to get them to the same voltage before charging). As far as I recall none of them were less than 3.2V before balancing

5A may be a gentle charge for a cell operating In it's normal range: I.e. between 2.5 and 3.65v, >0°C and not swollen. But I'd consider it to be a bit aggressive if the cell is over discharged and swollen.

Tx for the insight. As the Pylontech recommended charge is 25A (max 50A)I had thought 5A would be "gentle" in comparison :-)

I've had an idea on how to mechanically separate the cells so may try it at the weekend if I have the time.

APB