Another one bites the dust..

[Randy960]

I would love to know the specifics of how the BMS works, but I think you might be right about it only reducing power/locking out the battery based on whole pack voltage. I regularly see my weak cells in the mid 2v range under load while the rest of the pack is in the mid 3v range. About 3.2v should be the low end cutoff. So unless they left in the safety shutoff voltage from the 2014 LFP pack, it really doesn't seem to care about a few cells dropping that low.
I have DC charged the car up to about 90% several times while watching cell voltages, and it seems to derate to keep the highest individual cells under 4.150 volts, dropping from almost the full 50kw down to as low as like 8kw.

 

[FalconFour]

Indeed, I see the same behavior in quick charging as well - topping out at 4.15v for the cells that were previously "low". I had thought that was just coincidental, because the pack may still be relatively balanced at the top... just slightly underperforming for high current demands. However, seeing now two cases showing the same behavior (4.15v cap), I'm more inclined to believe DC charging may be safer than I thought.

I'm in the process of scanning/archiving the 2015 Spark EV service manual sets, so all can be told. That's estimated to be done in mid-April. But for the categories of these fault codes, the limits are absurdly high and low - a bit over 4.5v on the high side, and 1.75v on the low side. They even compensated for temperature (slightly different cutoffs) for extreme low temperatures, but they didn't think to program "hey, maybe we should limit and cut-off the battery based on these low cells, at more conservative levels, instead of making it a car-bricking fault".

 

[Randy960]

This i-miev has the same cell issues we do, except the BMS is smart enough to limit power until the cells balance out without locking out the battery.

 

[FalconFour]

Yup, sounds exactly the same! Well, our Spark batteries thankfully aren't *that* trashed (the oldest Spark is still newer than those i-MiEVs, though). Both those cars had some really trashed single cells that crippled the whole pack. But, by having a much more intelligent BMS design, the battery lasted much longer (at reduced performance, no doubt) until it became unusable, which is how EV batteries should behave - limit performance and constrain capacity to the highest/lowest cells. The Japanese got it right the first time (Nissan and Mitsu for example) - damn shame they gave up on EVs for basically the entire next decade.

I worry about that guy trying to piece through individual cells to try restoring it. All the cells in the "load test" are roughly the same now - for practical purposes. Even though some are 3.3v and some are 2.8v, they're all pretty g*-d* f*ed and it's just a spread of weakness. If he were able to get 90-or-so "new" cells, in good shape, then hey! That pack is practically prime for re-celling, since it's so modular. But to replace onesie-twosie, he's just gonna be chasing ghosts from 8 cells, to 16, to 32, to the whole pack soon enough, if he keeps chasing which one is the weak one. They're all trashed. haha.

As batteries age, it's just important to keep it *working*, operable, functional, as much as it can be, constrained by an intelligent BMS that knows its limits. Once the pack's constraints are no longer able to serve your needs, then it's time to either consider a new pack or a new car...