observable symptoms as cells degrade

Hi

I understand that capacity of cells degrades over the discharge recharge cycles, however I'm not sure how this manifests. Is it that the cells can not be charged up to 4.2V (falling back say to 4V as soon as removed from the charger) or is it that they sag faster during discharge?

Thanks

From this paper:
View attachment 285 degradation diagnostics for Li Ion cells.pdf

Sag is a sign of increased internal resistance, but not necessarily loss of lithium inventory, which is capacity loss.

pellicle said:

Hi

I understand that capacity of cells degrades over the discharge recharge cycles, however I'm not sure how this manifests. Is it that the cells can not be charged up to 4.2V (falling back say to 4V as soon as removed from the charger) or is it that they sag faster during discharge?

Thanks

Click to expand...

Resistance climbs, so as long as you have a **consistent** instrument and establish a strict protocol for comparability, just establish a baseline per cell and then compare as it rises, say every 50 cycles or so.

Yes max discharge rate will drop and voltage sag will be greater, but no need to objectively measure those.

IMO capacity drop is the surest indicator, how industry actually defines % State of Health nd thus EoL.

Use a CC load tester, again same setup used consistently, against a benchmark established after the breaking in period.

thanks folks ... great detailed answers (and that paper is fantastic)

(hat tip)

john61ct said:

Use a CC load tester, again same setup used consistently, against a benchmark established after the breaking in period.

Click to expand...

my problem is that the batteries are in a sealed pack (I assume with a BMS) so I can't really do that. I've noted that while absolute range of the scooter is slowly degrading (currently about 17km continuous riding flat ground, minimal headwinds, was about 20 4 months ago; approx 70% DoD once or twice a week usually only only discharging 20% of capacity) I see that the pack Voltage is now 52v where it was higher by a bit back in June. I anticipate this is about right for the (probably) low quality cells that are in the pack, as they've probably had about 80 times on the charger (in total) with most times being before the standing voltage drops below 47V (13 cells).

I'm just curious about the drop in voltage immediately off the charger and wondering if this means that I should probably dissect the pack (voiding my warranty) and look for physical damage to the nickle bus and as well measure cells (well bundles) and check for low or out of balanced ones.

PS: this is a communication I sent to the seller back on the 29th of July
https://www.youtube.com/watch?v=n8N2xySOA6Y

Thanks

Yes avoid such "drop-ins" in future, don't see how to gauge resistance there.

And whatever tests you use, set up benchmarks while the bank is at its peak, otherwise you don't know what you're comparing to.

You can still do a CC load capacity test on the battery as a whole though. Range itself is too imprecise I think.

Voltage on its own does not say much, 52 as 4.0Vpc after coming off the charger is still 100% SoC; retaining a surface charge is not a sign of health.

Hi

john61ct said:

Yes avoid such "drop-ins" in future, don't see how to gauge resistance there.

Click to expand...

not quite sure what you mean, but if you mean the pack type then that's pretty hard to avoid in eScooters. Referring back to your previous post, do you have a battery which allows disassembly to do cell level tests?

And whatever tests you use, set up benchmarks while the bank is at its peak, otherwise you don't know what you're comparing to.

Click to expand...

true

You can still do a CC load capacity test on the battery as a whole though. Range itself is too imprecise I think.

Click to expand...

agreed ... but on a consistent route in consistent conditions its not a bad indicator, certainly real world, just not transferable to other peoples circumstances.

Voltage on its own does not say much, 52 as 4.0Vpc after coming off the charger is still 100% SoC; retaining a surface charge is not a sign of health.

Click to expand...

... that's good to know

Thanks

I am going to disagree with the last quoted paragraph.

The difference in storage capacity, available wattage, between 4.2 per cell and 4.0 is not a "surface charge", there is significant wattage which should be available there. A small amount, such as declining from 54.6 to 54, or 53.8 or so, is not horrible, but getting a full charge at 54.6 and declining rapidly to 52, in a fairly new pack, is a bad sign and most definitely should not be happening. A good cell charged to 4.2 should stay at that charge level for a prolonged period. Not great for cell lifetime, but any level of significant self-discharge is a bad sign.

What is a very common occurance, especially in a low-dollar or low-quality pack, is that virtually all of that decline is present in just one or two cells, and not spread evenly over ALL the cells. In fact, a decline of that order is almost certain to be limited to just a few bad cells. Those cells will drag down and impose increased "wear" on their connected parallel string. Over time it will require replacement of the entire string, not just the bad cell.

Self-discharge on that order is not good at all, will get worse fairly quickly, and can become a dangerous situation.

Another possible cause, worthy of investigation, would be a poor connection, maybe more than one, which simply will not pass the full voltage from a particular cell or string, or is making intermittant contact, making for a wonderfully variable problem to diagnose. Again, this tends to be more common in "budget" packs rather than higher quality ones.

Ir tells the story.

My CA3 tells IR of the pack.
The hobby Powerlab charger I use has a IR graph.
The BMS I use looks at cell health, and rates it on a 1-100 scale.

There are three ways I check thae condition of my batteries.

My batteries are 5 plus years old and no real change in the past year. At least no real dataloggable.

Along with a great majority of the user base, I do not possess any device to detect or measure Internal Resistance.

I have four, red, LEDs.

I know how far I can go, on both medium and high speed settings, for each light, and I know both data points back from when the battery was new. That is all I need to know, that is all most users will know, and that is the only thing that is important.

I am not saying such tools do not have value, just that very few people will have them available. Work with what you have.

We can picture internal resistance as resistor in series with cell.
As it ages this resistor value increases.
This internal resistance causes voltage drop