How many bad/dead/weak cells in your packs?

Hii everyone!

So after i got the feeling of thinking too much about my pack design which is not based on long term experience, i wanted to ask you,the community about your long-term experience witz faulty cells in your packs.
I cannot get rid of the idea to construct a maintainable pack where you can swap cells most easily.
Thanks to daromer for his great feedback, but istill like the idea


So the question is: How often/frequently did/do you experience a malfunction in your packs due to faulty cells?

Many thanks for your input!

Best,
Chris

Yes they will go bad. Create your system with n+1 so you can take a string out and work on packs. Have a spare pack. If you do the design i got it takes max 5 min to sort 1 bad cell if fuse is blowd. If you need to find the bad cell it take longer time. Then i just swap out the pack and all the cells gets charged up. Then i pick loose all the fuses... Then i check what cell is self discharging.

If thats not a problem they all go into retest bin. With the holders its not hard to get 1 cell out if you just take of the corners

I'm running my powerwall since March, started with 700 cells in 100p7s / 5kWh, then step by step added more cells to currently 1820 cells, 260p7s / 13kWh (last row with 7 packs/140 cells went in production 2 weeks ago).

Until now, I didn't have any issues despite the nuts on my busbar connections got a little loose over time, which caused a disbalance of 120-150mV from highest to lowest series (compared to a 'normal' disbalance of 30-40mV before). Fixed it by adding spring washers.

I know one day there'll definitely be bad cells, and in this case I'll locate the faulty pack, take it out and disassemble it to re-test all cells. Since I have only 20 cells per pack, this shouldn't be much of a problem. Already did that in production mode, no problems (although I made aspare pack, but I intentionally didn't use it).

so, maintenance and swapping out bad cells should be quite easy. The only thing I supect is that maybe my active balancers will "hide" bad cells for a while... since they actively transfer energy (max. 10A) between the seriesses, I suspect them trying to compensate a cell that acts like a "black hole". That's why I measure the transfer current between the seriesses from time to time in order to find out suspect high currents. I also thought about switches to turn off the balancers, would make it easier to identify a series with a bad pack in it.

Have sun!
Oliver

Yeah thats one thing you need to beware of in terms of balancers without monitoring of the current. Since you know about i dont fear for it but its worth mentioning