Help with surplus prismatic cell specs

[Sparky_SC]

That's a huge number of cells! Testing them one by one will be tricky! I would simply charge them up to 3.65V and connect them 16 by 16 to built a 48V battery, with a BMS. Then running a load like an inverter, I would check if there is any runner on the pack and replace it.

It all depends what you want to do with them. If you plan to use them as 48V pack, then I would use one BMS per 16 cells, just to monitor cell by cell and not a bunch of cells in parallel...

That is pretty much the game plan. 16S1P with a bms in each battery pack. 4 packs total for the solar system. (48V inverter)
Reason is that seems like a decent plan and should keep the batts in the 25-85% SOC range during solar use which should extend their life considering their abusive past history.

Of course, its only a preliminary plan. One step at a time. First is testing in a 4s12v config to verify cells usefulness and sort out any bad ones. I have a cheap 12V4S BMS I got for test purposes only.

 

[ericgrand]

jump directly to a 16S pack...it will be quicker to highlight runners and bad cells...You maybe very surprised that none of cells will have been damaged...If none of them was at 0V, I would bet they are all in good condition!

 

[Sparky_SC]

jump directly to a 16S pack...it will be quicker to highlight runners and bad cells...You maybe very surprised that none of cells will have been damaged...If none of them was at 0V, I would bet they are all in good condition!

Problem is this is bench testing only at this point and no means to charge them in a 48V config or discharge/load test them either in that config. That is the reason for the 12V4S config during bench tests.

Once I have a single 48V/16S pack with a 16S bms, I will indeed haul the monster over to my solar inverter and give it a real world test. I have 4S2P 170AH agm's on the solar now.

Still real early in the testing, learning process. For this early stage I am quite pleased so far. The next few days will greatly increase the knowledge of the overall process and quality of what I have.

 

[RCinFLA]

1.1-1.5v are likely okay but better to initially charge at < 0.5 amp until they reach greater then 2.5v. When a LFP cell is left for extended time at less than 1v it will grow lithium metal dendrites. These grow like needles that can punch through the separator creating cell shorts. By starting out with low current you do two things. If there are leakage shorts, too much charging current will heat up cell and can cause them to burst and vent. If with low charging current the cell voltage does not rise you can be fairly sure it has some leakage shorts and should be discarded.

The fastest way to check cells is to check their terminal voltage slump after about two minutes of load. Charge them to mid state of charge so their rested (rested no load for about 5 minutes after charging) state voltage is about 3.30 vdc.

Get a iDST BattGO device that will measure 8 series connected cells simultaneously. Use two iDST monitors if doing a 16s configuration.

You will need to be able to apply a load current of between 0.2 to 0.4 CA, 32 to 64 amps for a 160 AH cell. You can use inverter to provide load.

With at least 50% state of charge, measure and record each open circuit rested voltage on each cell. Then turn on the 0.2-0.4 CA load. You are going to wait two to three minutes and record the loaded voltage on each cell. The slump has an exponential decay so it will drop quickly at first then slowly level off. When you start the load if a cell drops greater than 0.5v from OCV too quickly, just stop load and replace the cell that drops too quickly. That is a bad cell.

The more used a cell is, the greater the terminal voltage slump will be with load after two minutes of applied load when cell reaches equilibrium for given load current.

The chart below represents a new cell for an EVA 280 AH. The cells you have are similar thickness electrode design, just lower AH capacity. A used but still useable cell can be 3 to 5 times the slump voltage shown in this chart. Should match series connected cells that have same amount of terminal voltage slump for same amount of load current. The greater the voltage slump, the lower the cell's remaining AH capacity and poorer the condition of cell.

If you want to run AH discharge curves, First try a cell with worse terminal voltage slump. It will have the lowest AH capacity.



.

 

[Sparky_SC]

1.1-1.5v are likely okay but better to initially charge at < 0.5 amp until they reach greater then 2.5v. When a LFP cell is left for extended time at less than 1v it will grow lithium metal dendrites. These grow like needles that can punch through the separator creating cell shorts. By starting out with low current you do two things. If there are leakage shorts, too much charging current will heat up cell and can cause them to burst and vent. If with low charging current the cell voltage does not rise you can be fairly sure it has some leakage shorts and should be discarded.

The fastest way to check cells is to check their terminal voltage slump after about two minutes of load. Charge them to mid state of charge so their rested (rested no load for about 5 minutes after charging) state voltage is about 3.30 vdc.

Get a iDST BattGO device that will measure 8 series connected cells simultaneously. Use two iDST monitors if doing a 16s configuration.

You will need to be able to apply a load current of between 0.2 to 0.4 CA, 32 to 64 amps for a 160 AH cell. You can use inverter to provide load.

With at least 50% state of charge, measure and record each open circuit rested voltage on each cell. Then turn on the 0.2-0.4 CA load. You are going to wait two to three minutes and record the loaded voltage on each cell. The slump has an exponential decay so it will drop quickly at first then slowly level off. When you start the load if a cell drops greater than 0.5v from OCV too quickly, just stop load and replace the cell that drops too quickly. That is a bad cell.

The more used a cell is, the greater the terminal voltage slump will be with load after two minutes of applied load when cell reaches equilibrium for given load current.

The chart below represents a new cell for an EVA 280 AH. The cells you have are similar thickness electrode design, just lower AH capacity. A used but still useable cell can be 3 to 5 times the slump voltage shown in this chart. Should match series connected cells that have same amount of terminal voltage slump for same amount of load current. The greater the voltage slump, the lower the cell's remaining AH capacity and poorer the condition of cell.

If you want to run AH discharge curves, First try a cell with worse terminal voltage slump. It will have the lowest AH capacity.

View attachment 70113

.

WOW, priceless information ! Can't thank you enough !!!

This morning I picked up one of those load pile battery testers used for car batteries at harbor freight. It puts a 100A load on the battery and has a volt scale. My thinking was very close to the method you described. I figured it would show up any high impedance cells in the 4S test configuration. (checking the cell voltages individually with a meter)

 

[Sparky_SC]

Update !

Recovered a bunch of cells, assembled 2 12v packs in 4s config and charged. First pack load test aborted early (approx half way) because one cell was dropping faster than the others, not a bad cell but poorly matched to the others. Pack #2 balanced very well and load tested at 169.08 AH, so safe to say they are 170ah cells. Pack # 2 cells only varied between 6 and 12mv until the very end of the test when the spread became worse with 10ah or so left to go. The inverter tripped out on low voltage (10.5) and the bms low voltage never triggered. Pack was at very close to 11V at end of test and dropping quick.

I have another 4S BMS on the way so I can test/charge 2 packs at a time to speed things up, I have 180 cells on hand and more coming so test time is important. I have two power supplies for testing, one 12V 25A that is variable 10-15V, and a bigger one that is 0-20V at 45A. Working with two setups at once will speed up things a lot.

Now, after the first successful test I have a good handle on what to expect and the characteristics of the cells. Discharge/capacity tests were done at around 50A because the BMS is only rated for 60A.

 

[740GLE]

What’s the final goal of these cells? Will you be using all 180 cells?

Looking to unload any of them to willing parties?

 

[Sparky_SC]

What’s the final goal of these cells? Will you be using all 180 cells?

Looking to unload any of them to willing parties?

There are actually 300 of these cells ! I got 180 in hand and another 120 coming in a week or so !! Testing them efficiently will be a priority due to the number. Most if not all are already spoken for, not all sorted out yet. Personally I am preliminary planning on making 4 batteries in 16S config for my use (64 cells total plus keep a few spares on hand just in case)

 

[DJSmiley]

170Ah, 16S = little over 8kWh / pack
with 4 packs 32kWh

That's nice! And considering you have a whole lot more left and an incredible price (duh)... you're a lucky bastard

Definitely worth the effort to charge/test them

 

[kuranaga]

4.2V is for Li-Ion batteries, not LFP! beware, sometimes they sell LiIon as LFP on Alibaba. There are cells that look very close to the 280Ah blue ones but have completely different chemistry

 

[kuranaga]

for instance:

This product is no longer available.

This product is no longer available.

www.alibaba.com

 

[Sparky_SC]

4.2V is for Li-Ion batteries, not LFP! beware, sometimes they sell LiIon as LFP on Alibaba. There are cells that look very close to the 280Ah blue ones but have completely different chemistry

Yes but not a concern in this case. These are US made and I have the manufacturers spec sheet.

 

[Sparky_SC]

US made LFP?

International Battery Inc. Allentown PA Out of business now though