18650 cells, capacity versus long term reliability

Good afternoon,

I'm building a pack for running my heated clothing on long dives, 21 18650's in a pressure tight canister.

Looking at samsung 2500mah 25R's, or something higher capacity like the panasonic NCR18650B (3400mah).

The draw on each cell is only about 0.7A and I plan to use a seiko s-8245a based BMS protection board (ebay)

The price is broadly similar but I'm getting more Wh per £ with the panasonic and I can get away with a very conservative charge level due to the good capacity.

Cycle life on the datasheets suggest the panasonics will have more capacity left after 300 cycles than the samsungs have new.



The question is, is there a reliability trade off with the larger mah cells?

There is likely someone who can say better, but here is my two cents.
At the current draw you are looking at with the Panasonics, that is well within operating range and a quick look at discharge charts shows it can hold voltage just fine at that level with no temperature issues, and you should get pretty close to the rated capacity for a full cycle. For your application I think you'll benefit from the capacity at your discharge current and need fewer cycles than the Samsung battery under the same conditions over time.

Where you would have a better use case for the Samsung cells would be in high drain applications such as 1-3C on that cell, where it will hold a higher voltage longer than the Panasonics with lower temperature from resistance in the cell, and likely more capacity as the Panasonic cell would struggle to hold anywhere near full capacity at 2-3C. A side note of safety, 3C discharge would be much too high for the Panasonic cell, and 2C is on the high side.

I would think twice about using this lithium chemistry in regards to your heated clothing application, especially a diy one at that.

If at all possible, change your chemistry to LiFeP04. While this isn't a magic wand to save you from any poor diy wiring infrastructure, consider that LiFeP04 is used inches away from motorcycle riders butts, and a few inches away from advanced wheelchair user's butts as well - for a good reason (see wheelchairdriver forum).

If you can afford the slightly larger physical cell size (due to the lower energy density than your usual 18650 chemistry), I'd highly recommend it. LiFeP04, aka LFP, is one greedy molecule that holds on to oxygen, rather than using it as a combustion source when things go wrong. That isn't to say that total abuse won't cook the electrolyte, or your wiring infrastructure can't go up, but taken solely on it's own, LiFeP04 is used for a reason when the application is something close to your body.

See if you can get by with perhaps 26650 LiFeP04's, (or maybe some Headway cells) and of course remember that these are nominal 3.2v / 3.6v max charge, so your bms would have to set for LiFeP04 charge / discharge voltages.

The cells are in a sealed canister worn on the rig, outside the drysuit with the entire north atlantic to keep them cool if it all goes tango uniform.

Since I don't know all the particulars of your application, all I can say is to do your due-dilegence in regards to safety no matter what the battery chemistry is.

Ie, I'm assuming you are going to have to tote this project to your site somehow. A pressure-tight canister is bad for obvious reasons during transport if something inside goes Tango Uniform, like the bms boards failing, cell goes into reversal and such, not only for yourself, but for others.

I'm not trying to be dramatic. Know your limitations and risks. Don't just dive into this project blindly.

Yes point taken, I realise that these batteries pack a lot of energy.
As it is, the canister is designed to withstand external water pressure but not internal pressure. I will take another look over this and make sure there's no chance of it building pressure. On the other hand, an extra PRV would be one more penetration into the canister, which has it's own drawback!

Some of the halcyon/salvo dive lights had problems a few years ago. At least one of which ended up going off with a bang due to this problem, I think it had partially flooded on the dive and the salt water had not particularly agreed with the cells.

This may well be an issue with another project which is a scooter/DPV, that pack will be something like 1kWhr. Scary thing!




p.s I had LiFePO4 battery on my motorbike as well, not so much of a problem with that 2" from your ***, you can always stand up! The 20l of petrol an inch from your nuts is a more pressing concern

yorkie_chris said:

I'm building a pack for running my heated clothing on long dives [...]

Click to expand...

Be aware that in apps like this one needs to pay close attention to how the cells behave in cold temperatures. Typically the IR increases at colder temps, which causes an intial voltage sag. This eventually alleviates a bit after a self-heating effect kicks in after a short time under load. For example, examining the graph below, we see that at -10°C the initial voltage dips down to 3.2V before rising from self-heating. This initial dip might descend low enough to trigger undervoltage protection on some devices. If this happened below then the battery capacity would be under 100mAh, vs. over 1000mAh capacity obtainable after battery self-heating kicks in. Often you can shrink these dips by adding a little heat (e.g. body heat) to jump-start the self-heating process.




Note: the above graph is excerpted from p. 14 of the book: Y. Barsukov and J. Qian, "Battery Power Management for Portable Devices, 2013. The authors are leading experts at Texas Instruments, responsible for the design of TI's impedance tracking fuel guage algorithm - used in most laptop batteries. It contains a wealth of useful information. Highly recommended reading.

See also related threads on Li-ion powered on hand warmers and foot warmers.