Testing "almost new cells" - Is it worth it

Mikethezipper

Member
Joined
Jul 7, 2018
Messages
121
Howdy,

A while back I purchased around 2,500 of the Panasonic NCR18650Bcells from a local company that had been making battery packs for drones but lost their govt contract. The cells are assembled into packs and although they were cycled - I was told they had barely used any of them.

So far I've taken apart about 500 cells worth of these (the packs are put together pretty darn well) and tested them, which has taken a ridiculous amount of time. Every single cell is at 3.2v or above. When I capacity test them, every cell is showing 3400-3600mah per cell, so they are testing at their rated capacity of 3400mah.Considering that I can only test ~24 of these a day, it's taking a long time. After that I have to check them for self-discharge so it's adding another month on top of that for each cell.

As I test these when I'm at work, I can't do a good job of checking for heaters. What I'm wondering is whether it's worth checking for SD cells, and if I should be worried about heaters. Is this a property that arises in heavily used and abused cells, or is this pretty much random and I should go ahead and do the full testing on all my cells?

I'm not sure if it was Wolf - but another thread on here showed a strong relationship between IR of a cell (assuming we're testing all the same cell here) and the likely-hood that it won't be worth the effort to capacity test etc as cells with unusually high IR for that type of cell also tend to either have low capacity, be heaters (duh), or be self-discharging cells.

What do ya'll think? The safest answer would be to just go ahead and check for SD on these cells, which adds quite a bit of time. But is it reasonable? If they were SD cells, shouldn't they have really low voltages by now? They have sat for about a year before I got to them, so I think them being at a voltage above 3 is a sign they are not SD cells. However, they are in 8s8p packs, so maybe an SD cell would be held up by other healthy cells. Is that reasonable?
 
test them in packs. If they are in 10p packs you can test them in such as well.

With that said you have to guestimate. Test xx random cells and estimate how many dead ones you might find.

/D
 
Mikethezipper said:
............but another thread on here showed a strong relationship between IR of a cell (assuming we're testing all the same cell here) and the likely-hood that it won't be worth the effort to capacity test etc as cells with unusually high IR for that type of cell also tend to either have low capacity, be heaters (duh), or be self-discharging cells.

My research has shown that theNCR18650B thrives in a m? environment of?45m? you can get away with up to 50m?.

The great thing is after recording over 7000 cellsI just sort my sheet byNCR18650B and wella I can give you a good educated guess supported by actual test results.

image_sqqhus.jpg


So as far as testing the cells..... a cell with anIR measured with a proper meter that is >50m? no go.
As far as testing all of them? What are you using these cells forand how easy is it to replace a bad one in the pack you build?

As far as an SD in a 8s8p pack the SD would bring down one of the8p in a 8s pack and the voltage would be lower thana healthypack.
3.2V X 8 is 25.6V so any pack bellow 25Vis suspect.
Conclusion :
If you want to skip SD make sure the cell has an IR of?45m? and you should be 99.999% OK
Also obvious is SOH.If the cell is coming back at >90% capacity there is a good chance it is a good cell and will not be a SD.

Wolf
 
Personally I would have did my preliminary testing them without doing any disassembly You can considering making a tool like this out of some balance leads and some alligator clips to make it easy to attached to the appropriate cell groups.

With the right RC balance charger, you would be able to charge, discharge, capacity test the pack, identify low capacity cells, and do an IR measurement without taking anything apart.



image_dyjzuv.jpg
 
Wolf: I am using these in a mini-powerwall setup for my fulltime live-in RV. I'm planning on a 26v (7s) 260AH pack for a total of around 7.2KWH. Based on my current setup and the extensive data I have for it, I don't suspect I will ever go below 80% SOC for my pack. If I were to max out my inverter, it would pull 1.5A per cell - whereas realistic typical usage would be < 200mA per cell.
Pack build will be as per the most common setup of the 4x5 cell holders ,
double bussbar, cell level fusing - so it should be doable to replace a single cell.

Thanks for the ideas guys! Although I won't be able to implement all of the ideas here, I'm realizing that I can at least charge each set of parallel cells together. I don't feel comfortable testing SD cells with them still in a pack since a single SD cell won't make a big enough difference over the course of a month.. I'd have to wait a couple months at which point I might as well pull them apart.
I could capacity test a parallel pack at a time - but now I'd have to buy new gear with unknown offsets and calibration issues that I'm not sure is a road I want to go down right now.

I'm concerned because I was getting quite a few cells with an IR over 120mOhms that had full capacity... but IR was measured by my Zanflare units, which are known to be inaccurate.. I will be retesting with the AC unit commonly used on this board.
 
Back
Top