Li-Ion: Self discharge and capacity

I'm not battery or electronics guru... actually quite the opposite. So i read this thread which was created several years ago about self discharge on a 18650 and how they are exceptional at maintaining voltage for over a year at a time. But i also read else where that Li-Ions perform in a manner where they maintain their voltage throughout use and finally lose voltage significantly as capacity is low. So i thought to myself how accurate is that test? The battery voltage may still be within 95%, but could the capacity be at only 60-70%?

Hello Anonymoususer,

With Li-Ion chemistry there is a pretty good correlation between open circuit resting voltage and state of charge. Here are some values for this.

4.2 volts 100%
4.1 about 90%
4.0 about 80%
3.9 about 60%
3.8 about 40%
3.7 about 20%
3.6 empty for practical purposes.

This means that if that cell had dropped to 60% capacity, the voltage would have dropped to below 3.9 volts.

Tom

Amazing really. Thought they would loose more.

Shouldnt be a big difference with protected cells - they only use a few µA.

SirJMD said:

Amazing really. Thought they would loose more.

Shouldnt be a big difference with protected cells - they only use a few µA.

Click to expand...

The good ones, yeah, but AIUI, this varies a lot by manufacturer -- I don't know solid numbers, but I think I've heard figures in the 100s of µA for some of the worst offenders. Since 1000 hours is roughly 6 weeks, that's an Ah per year or so.

EDIT: If anyone has data on this, please correct me.

SilverFox said:

Hello Anonymoususer,

With Li-Ion chemistry there is a pretty good correlation between open circuit resting voltage and state of charge. Here are some values for this.

4.2 volts 100%
4.1 about 90%
4.0 about 80%
3.9 about 60%
3.8 about 40%
3.7 about 20%
3.6 empty for practical purposes.

This means that if that cell had dropped to 60% capacity, the voltage would have dropped to below 3.9 volts.

Tom

Click to expand...

You are the man, Tom. :thumbsup:

how quick should regular, unprotected, not-in-flashlight cells lose their charge? I'm trying to weed out the baddies in my harvest.

Edit: sorry, I'm blind as a bat. Information is in the thread linked by anonymoususer. Thanks, AU!

Dude... did you read this thread and the link provided?????? Silverfox has provided all the information you need.

Will a flashlight (fenix tk11 in this case) die before the cells become over discharged and cannot be recharged?

High quality lithium-ion cells with have self-discharge rates approaching primary cells. In fact, there are several specialty applications where lithium-ion cells are effectively used as primary cells (typically high end medical devices and some military applications).

However, the self-discharge rate varies WIDELY with cell design, purity of materials and manufacturing quality. In fact, one very cheap and effective way to evaluate the quality of cells from multiple manufacturers is to charge them up to 100% SOC, then stick them in an oven between 40-50 degrees C and monitor the voltage with time. High quality cells will have the lowest voltage drop and the cell-to-cell variability will also be small. Poor quality cells will generally have a higher average voltage drop, and the cell-to-cell variability can be substantially larger.

Cheers,
Battery Guy

Battery Guy said:

In fact, one very cheap and effective way to evaluate the quality of cells from multiple manufacturers is to charge them up to 100% SOC, then stick them in an oven between 40-50 degrees C and monitor the voltage with time.

Click to expand...

This sounds like something rather fun to do with my cells on a Saturday afternoon. Unfortunately, I've not seen a household oven that goes down to 104-122F. Any tricks up your sleeve for that?

core said:

This sounds like something rather fun to do with my cells on a Saturday afternoon. Unfortunately, I've not seen a household oven that goes down to 104-122F. Any tricks up your sleeve for that?

Click to expand...

Ok, do NOT put your lithium-ion cells in your kitchen oven! Seriously, do not do it! Never, ever, ever!

The ovens I use are laboratory ovens that are calibrated for relatively low temperatures. They also have blast-proof doors and are vented in case a battery catches on fire.

So forget what I said about the whole oven thing. Besides, even at slightly elevated temperatures, it takes a few weeks to start to see the difference between cells. Also, this tends to be a destructive test, because storing lithium-ion cells at elevated temperatures for long periods of time will permanently degrade their performance.

Cheers,
Battery Guy

Yes, I was half joking. But only half.

I had a feeling you were talking about a lab oven. But you did say it was a "very cheap and effective way", and lab ovens do not come cheap.

core said:

Yes, I was half joking. But only half.

I had a feeling you were talking about a lab oven. But you did say it was a "very cheap and effective way", and lab ovens do not come cheap.

Click to expand...

Cheap and effective, provided you already have a lab oven and appropriate safety equipment.

I am over reacting a bit, but I really do not want people to go and throw their lithium-ion cells into the toaster oven.

SilverFox said:

Hello Anonymoususer,

With Li-Ion chemistry there is a pretty good correlation between open circuit resting voltage and state of charge. Here are some values for this.

4.2 volts 100%
4.1 about 90%
4.0 about 80%
3.9 about 60%
3.8 about 40%
3.7 about 20%
3.6 empty for practical purposes.

This means that if that cell had dropped to 60% capacity, the voltage would have dropped to below 3.9 volts.

Tom

Click to expand...

Ok, I'm confused. This scale rates 3.6v as depleted, but protection circuits usually kick in at 2.9-2.8 volts. Is that last ~0.8v basically useless? Sorry if this is a dumb question, but I'm a noob at using 18650s in anything but a laptop (obviously, lol).

HeyGuysWatchThis said:

Ok, I'm confused. This scale rates 3.6v as depleted, but protection circuits usually kick in at 2.9-2.8 volts. Is that last ~0.8v basically useless? Sorry if this is a dumb question, but I'm a noob at using 18650s in anything but a laptop (obviously, lol).

Click to expand...

Pretty much HGWT. I'd say that the remaining capacity of a LiCo cell, at 3.6 Volts, is probably less than 5%.

Dave

The chart lists "resting" voltages, meaning the cell has been inactive for hours, in high-power usage, a cell near the end of it's SOC is likely to sag quite a bit more, take the 3.6V resting cell, plug it into a 2 amp draw, and it may almost immediately fall to the 3.0 range, where the protection circuit kicks in to protect it from further falling.
Also the chart is just a "typical" chart, YMMV, the low end might slide a little lower, but if you take 2 identically rated fully charged cells, run them both at 2A for an hour, and then let them sit a day and measure them, if one is at 3.9, the other at 3.6, then you know you have a weak cell (the SOC is lower). So the higher cell might have another 1/2 hour of operation to fall to that 3.6V resting level also, at which point it probably STILL has a better SOC than the one that hit that level first (plug them both in again at 500mA, and the "better" cell will still run longer probably, the initial testing indicates it is better overall "internally" than the other, so even at the same resting voltages it's still likely performing better in usage).
Hope that helps?

Good point about rested OC voltage, Christexan. I assumed this point was already noted.

Dave

Thanks, guys! That all makes sense to me. I thought it was probably something like that, but I don't want to assume anything and end up with :fail:. I want to learn as few lessons the hard way as possible on this one. Killing a cell I can live with, but having a light grenade on me doesn't sound like any fun.

Think of a battery as a water tower. Large tank up in the air. When the tank is depleted, but the pipe leading up to it is still full, the water pressure is still pretty high down at the faucet, but the tank is empty (cell is dead), you can take a little more water out of it at this point, but the pressure will drop rapidly as the pipe empties. 4.2V is a full tank, 3.6V is an empty tank but a full pipe. The pipe holds a very small fraction of what the tank holds.


Eric