Maintaining Charge When Storing Li-ion Cells Long Term

I was wondering if I'm storing my 18650s at 3.7 volts for long term storage at what point do I need to charge them again? 3.65? 3.6?? 3.2? Also I'd it best to give them a few minutes top up just to the 3.7v or do I need to fully charge them and drain them down to 3.7v?

From what I've read it's quite a large margin. 3.5 to 3.7 seem to be reasonable targets to store Li Ion cells. So if you start at 3.7 it'll be while before it gets to 3.5 or lower. Have you got a DMM? Best use that and just keep them in the range.
It doesn't matter how you get there, charge or discharge, just that you do.

P

I'd top them up to 3.75v when they hit about 3.6v. If these are unprotected cells, you will probably only need to top them up every couple of years. If they're protected, you might need to top them up more often, due to the slight parasitic drain from the protection circuit. And, no, you don't need to fully charge them and drain them; just charge them up to 3.75v.

Sorry,
it is not possible to define voltages valid for all Li-Ion chemistries.
Usually you should target a charge of about 50% of capacity and allow a window from minimum 20 to maximum 80%.
The open loop voltage at 50% of charge depends upon the chemistry of the cell:
Panasonic NCR18650B: 50% at 3,6 to 3,65V
LG ICR18650D1: 50% at 3,80 to 3,85V
With the voltages (3,5V or 3,6V) specified by some posters here you may damage some batteries severely!
LG ICR18650D1 will be depleted completely at an open loop voltage of 3,5V!
Even at 3,6V this battery will sit on less than 5% of the capacity.

Practical hint:
Look at battery comparator from HKJ
chose your battery and display the curve for 0,2A only. With this curve you can estimate which voltage window will be safe for your battery.

Tobias Bossert said:

[...] Practical hint: Look at battery comparator from HKJ
chose your battery and display the curve for 0,2A only. With this curve you can estimate which voltage window will be safe for your battery.

Click to expand...

Better, see HKJ's Estimating remaning capacity in LiIon batteries. The rightmost 60 minute rest curves in the 1A charts will give a good approximation to resting voltage vs. capacity for the common chemistries listed there.

Sound advice, thanks to you both.
P

Gauss163 said:

Better, see HKJ's Estimating remaning capacity in LiIon batteries. The rightmost 60 minute rest curves in the 1A charts will give a good approximation to resting voltage vs. capacity for the common chemistries listed there.

Click to expand...

Thanks everyone......one is the aw 2600 cell on the link.....would the lower capacity 2200 aw I have that's the same chemistry be 50% at the same voltage as the 2600?

Nice: That's exactly the reason why I proposed to use the battery comparator, covering a wide varity of batteries. HKJ conducted measurements for a small count of batteries only to estimate remaning capacity in LiIon batteries. So, if you can find there your type exactly, use it. If not, go back to battery comparator and use this source.
AW 18650 2200mAh is quite different from AW 18650 2600mAh (2016):
There is a difference of up to 0.2V and the discharge curve of 2600mAh type is relatively flat. If you use the voltages from 2200mAh type you may damage your 2600mAh battery!
The other way round it is o.k., the 2200mAh type will have no problem with the voltages from the 2600mAh type.
AW 18650 2200mAh: 50% at about 3,63 to 3,68V => 2600mAh type will hold 5 to 10% at that voltage only
AW 18650 2600mAh: 50% at about 3,80 to 3,85V => 2200mAh type still has 75 to 80% at that voltage

Tobias Bossert said:

Nice: That's exactly the reason why I proposed to use the battery comparator, covering a wide varity of batteries [...]

Click to expand...

But the battery comparator is much less accurate for this purpose since it only shows voltages under load, so you need to attempt to extrapolate resting voltage from that - which can introduce large errors (esp. in regions where the curve is flat or low slope - so a small change in voltage corresponds to a large change in capacity). Accurately performing such extrapolation may not be so easy for many users since it involves knowledge that is often lacking (esp. about the effects of internal resistance).

Many cells used by hobbyists have chemistry the same (or close) to the handful in the link I gave, so the graphs there will typically yield much better capacity estimates. Only for cells of very different chemistry should one resort to less accurate methods such as extrapolation from voltages under load.

Update: I recalled that you earlier posted some interesting graphs illustrating the variation of IR over the course of discharge. This should help give readers an idea of the complexity involved in accurately performing such extrapolation.

I have the older aw 2600....is this the 2016 shown in the discharge graph by hjk? So the older 2600 will be similar to the 2200 and ok at 3.7v to store? Only the 2016 needs to be 3.8- 3.85?