when 18650s "die/let go" in parrallel

[jimmyhackers]

ive made a fair few golf trolley batteries for my dad and his mates out of used 18650s. usually 3s or 4s in a 4p or 6p config.

every 18650 gets tested for discharge/charge capacity, internal resistance, self discharge, and is installed with closely matching cells at the same voltage. the cells are mostly all of the same brand/type.

they use imax b6 balance chargers and have been told to always balance charge them before/when needed, and put on storage when not.

every now and then i do maintanance/check them and replace dud cells.

usually a bank of parrallel cells will be lower voltage than other banks, indicating dead cells.
i take out every cell and measure voltage individually.
The duff cell/cells in each parallel group will be at 1v or less while the others read an identicle normal non-broken voltage 3.6v and above.

my questions are....what causes this? is this preventable? am i doing something wrong?

 

[lnanek]

Sounds like some lithium batteries have a self protection circuit that triggers under 2.9v, others will go experience unwanted chemical reactions, heat, and even copper shunts forming in the battery that short circuit it:
https://electronics.stackexchange.com/questions/219222/is-draining-a-li-ion-to-2-5v-harmful-to-the-battery

Basically the chemistry of the battery is self destructive under 3V.

 

[spinningmagnets]

Probably dendrite formation from storing the cells at 4.2V each. The dendrites dont actually short-out the anode and cathode at first, they just get close enough to form a flow leak.

If they ever grow to enough length to short the cell, it suddenly gets very hot in that particular spot, and then the organic electrolyte starts converting from a liquid to a gas, and also the insulating layer starts to shrivel when it gets hot enough.

Charge to 4.1V per cell just before use, and store them at 4.0V or less.

 

[From-A-To-B]

Dendrite formation from storing the cells at 4.2V each. The dendrites dont actually short-out the anode and cathode at first, they just get close enough to form a flow leak.

If they ever grow to enough length to short the cell, it suddenly gets very hot in that particular spot, and then the organic electrolyte starts converting from a liquid to a gas, and also the insulating layer starts to shrivel when it gets hot enough.

Charge to 4.1V per cell just before use, and store them at 4.0V or less.

This post really helped my understanding. Thank you for sharing that, SpinningMagnets.

 

[spinningmagnets]

Go to electricbike.com and search

"How to make a lithium battery last...or kill it, if you like"

Also "Jeff Dahn Dalhousie University Tesla"

 

[jimmyhackers]

thanks

Their chargers only charge them to 4.1v per cell but i guess they are just not putting them on storage.

dont suppose there are any threads on that site about dendrite removal? :lol:

 

[docw009]

Sounds like you're not using a BMS on these batteries. The trolley might have an LVC cutoff of 9 volts? If not, golfers are going to run the batteries too low and they will get unblanced. It might be worth putting on a balance BMS, if they make them for 3S, and letting them charge with a regular charger. Sure would be a lot easier.

What's your cutoff on IR? When it gets up around 50 milliohms, you'll start seeing more voltage sag. That contributes to uneven wear.

I've not seen any deterioration like that in the few batteries I've built, but I mostly use new cells. If they're used cells, they're around 90% of original capacity and are ebike rated cells, like LG MH1/MJ1.

On the other hand, I've have two storebought battery packs fail on me, with same issue of one of the ten or thirteen cell groups not being able to hold a charge. I removed then, set the cells aside and later found that each group had a cell that would self discharge, Nothing wrong with capacity. Same as the others when tested, but it was just a leaker. I chalked it up to the batteries using cheap cells that only had a life of 2 years. I have yet to see a battery with Samsung,LG, Sanyo, etc, do that on me.

 

[jimmyhackers]

my cutoff ir is about 30mohms...(or heaters) or (slow leakers) basically when they go in the batts they seem normal enough.

all the cells are used mainly from laptop batts.....so ironically samsung and lg cells

they are usually in a capacity that can do way more than 18 holes of golf (8ah +). so its probably not overdischarge.... most of their trollleys are between 200w and 300w Max, but under testing current draw never reallly goes above 1amp..

most of their trolleys are "fancy" and have a LVC change option for pb or lipo. which i went through/changed when i sold them thier batts.

i dont like BMS's, the irony is/was half of "My customers" already had a store bought new fangled lipo/LiFe batt for their trolley which has failed.
the only failing item being the bms......which usually takes a few cells with it, these are the cases where i recover/replace what cells i can and attach a balance lead and give them a balance charger

 

[docw009]

If you can convince old golfers to use balance chargers, you're a good teacher. I have am Imax B6, and I imagine that it it catches those batteries when the cells drop under 2.5. and refuses to charge? Maybe they;re not letting them balance,

I've only used mine to charge one RC Lipo pack that is still good, but just starting to get a little wider. When do you get rid of a cell that starts to puff?

Maybe part of your future screens would be to charge a cell up, and let them sit for two 2eeks to see how much is lost.

 

[jimmyhackers]

When do you get rid of a cell that starts to puff?

ASAP!

your life or your house should be worth more than a lipo batt

 

[docw009]

New battery has been ordered. The old one was barely used, but I guess the 100A jolts add up. I have it discharging @ 1 amp right now.

I looked at a few youtube videos on the subject. They convinced me to bin it.

 

[jimmyhackers]

salvage of cells depends on your skill level/soldering/spud size.

they can be sperated out and re-used.... but you should probably test them before reuse....

also if one failed but the rest havnt, it might not be long before the other cells follow suit and die.