Lipos gone bad check out IR

I have a 20s setup with 2 parallel banks of 5 x 5000mah 4s lipo hard case batteries at 84v 10Ah.

I run the D166 BMS but I have a feeling that leaving the batteries fully charged for long periods may have contributed to some of the cells going bad.

The last few rides resulted in the BMS cutting out every time I would try to accelerate.

I'm thinking of replacing the red ones and considering the yellow one ? Your thoughts ?

I have set the charger to 84.5v since I'm not sure balancing is happening but maybe this may send some cells over 4.2v ?

Yeah, anything 20mOhm and higher I usually cull out. Maybe parallel a bunch of those together for occasional range extension or use 'em for lighting, mowing, etc. But, once they go above 20mOhm sag is just too great for my 30A eBike.

EDIT to add - the high 30's and that 40 will definitely be turning capacity into heat. You should also rig up a parallel dongle in order to watch actual voltages. My Bestetchpower BMS can allow cells to reach 4.25V OVD. Do that enough times it could be hurting things? Hard to know until you can parallel the balance channels and see what's happening. Personally I ignore the 84mA balance current - too slow for large Ah groups. I just touch up balance if/when needed via the balance wires.

we learned from that lecture on the tesla forum that time at full charge is the factor that is critical for cycle life.

time at temperature and time at full charge is like two strikes already.

can you hack the individual packs apart and then recombine all of the pouches into one pack and find a few replacement pouches to add to the end or wherever to keep it at 20S?

or you can just cull the bad ones out and rebuild the pack down at 18S or so with what you our BMS will handle it at 18S that you have left. you can adjust the charger voltage down to match.

if you have high resistance to charging then it may be that the pack is not fully charged when you think it is. the cells with high resistance cause the BMS to hit the HVC as that cell shows a higher voltage during charge because of the resistance.

in that case the BMS may be cutting out because it reaches the LVC immediately because the cell is depleted because it never filled and then high resistance adds even more delta V to drop you under the LVC.

The overcharge is a concern,
If it's possible to reprogram the D166 what settings would you implement if you could do it again ??

Over charge detection voltage 4.2V~4.35V(Adjustable)
Over charge release voltage. 4.0V~4.2V(Adjustable)
Over discharge detection voltage 2.3V~3.0V(Adjustable)
Over discharge release voltage 2.3V~3.0V(Adjustable)
Over current detection voltage 0.1V~0.2V(Adjustable)

I normally recharge once I get home, for one reason or another the bike might sit around for weeks before I use it again, maybe I can hook something up through the charge connector to discharge it back down if its not in use, any suggestions ?

It would be nice to have access to cell voltages in real time or even nicer if the bms can tell you what's going on..

use the standard set points. 2.8-4.25V for lipo. 4.20V balance Voltage.

lipo stays well balanced so you can just charge before you use them.

How did you measure those IR readings ?......from a charger ?,...from a discharge test?, ..IR meter ?
You can get seriously different results depending on the measurement method, state of charge, temperature, discharge rate, etc etc.
You might consider a recheck using different method/settings ?
I'm sure you were consistent, but the actual readings are hard to compare to other peoples records without reference to the technique used.
But something tells me IR is not the cause of your problem.
The difference in operation between cells at 17 mohm and 22 mohm is not significant.
I have several packs with cells at over 20 mohm that function perfectly fine.

Yeh sorry it was on a Hyperion charger discharging at 5 Amps.

You are right it did vary when I did an IR test while charging at 10A but not by much at all. I thought I would stick to discharge since this is what Im really worried about.

All batteries were done one after the other to minimise the temperature difference which was probably around 17 degrees.

Both left and right batteries at position 5 exhibited worst IR (around 40mohms) at Cell 4 which is the negative end of the power bank connected to the BMS/controller which is interesting !
I have done some reading on this and seems that anything above 20 causes too much voltage drop under load so best to change them out.

I tried to open up one of the hardcase packs but just ended up cutting hole into one of the cells and heard a lot of hydrogen gas venting... not good...

Obviously every cell will give some voltage sag, proportional to its IR.
So , yes the highest IR cells will give the most sag ..but IR is cumulative on series connected cells
This is more of a problem because you are running 20S,.. So for each bank of 20S you have a total IR of approx 0.35 ohm, which for every 10 amps load (per string) will give you a pack voltage sag of 3.5 volts,
Do you know what your max load is on the pack ?

Left and right 5 are definitely done, but the whole pack is heading that way. If two years old, that's about normal.

Leaving long periods warm and fully charged a big mistake. I did that to a 20s 10 ah set, got about 20 cycles out of them over 2 years. They weren't stored that hot, but they weren't stored cool in summer either. About 80F inside an old unused fridge outside.

Well I changed 4 of them out and its much better.
Still some noticeable sag about 3.5v when I take off with a bit of throttle and about 2-2.5v sag when I'm cruising at around 32km/hr.
Im guessing that Im running between 5 and 10 amps most of the time but to be honest I haven't measured it.
After a 17km ride this morning at 83v rest voltage down to 78.4 rest voltage. It was under 10 degrees C this morning so not ideal conditions for the battery.

Just thinking that I probably should have changed the lot... ?

They'll do noticeably better in warmer temps. I'd wait until it warms a little before replacing 'em all.

use a table saw to cut open the hardcase packs. cut around the top just below the top and only deep enuff that the kerf barely penetrates the inside. it will leave a ribbon of melted plastic around the inside where the saw kerf went through but it will not go all the way inside and contact the pouches so then you can pull the top off.

then cut along the seams on the side enuff to pull the sides apart and pull the pouches out.