Uneven charge/discharge parallel lifepo4

[SCmountainman]

I have 2 factory built 48v Lifepo4 battery backs. 5.1kwh packs, 16cell prismatic. They are about 3 months old. They have built in BMS, breakers etc. Identical packs, wired in parallel per manufacturers specs. (Not divulging manufacturer’s company yet). One of the batteries always charges and discharges at twice the rate of the other battery. Always. Have been in contact with tech support numerous times. Tried all their solutions and tests. Verified amperages with fluke meter, tested each battery independently. Timed charges/discharges and measurements with constant input and or discharge on each battery. I think the one battery has something wrong with the cells, possibly resistance issue? Manufacturer says it’s BMS related and has been yanking me around for 2 months and promising to send new BMS, still have not received a replacement. I call them weekly. I have provided measurements and videos to tech support. They state the have provided xxxxxxx KWH of power with no issues. Can anyone offer any other suggestions or insights? Am I on the right track or completely misguided?

 

[rickst29]

I would add a coulomb counter meter (if you do not have one already), and test each battery for charge and discharge behavior while the alternate battery is fully disconnected. Verify that each delivers rated power before reaching low voltage (you don't need to pull them all the way down, I'd stop at 51.0V remaining - about 18% left unused) Within that test:

• Do both batteries discharge at the same rate, when separated and running a large load alone?
• Do they charge at the same rate?
• What is the internal resistance for each battery pack alone?

If their behavior separated is pretty much "correct", then either internal resistance OR external wiring is to blame for one being "preferred" over the other. The difference does not need to be very large - and ultimately, when the "faster" battery gets out of the flat part of the curve (exceeding 93% SOC while being charged, or 15% SOC while being discharged, the slow battery should take more of the charge or load as a simple matter of sharing the common voltage.

If it's all being caused by a difference in internal resistance, you can argue with the mfgr. whether then more resistive battery is defective in construction or internal components. If they do not offer to replace the "more resistive" battery or its internal components, then you can still rewire externally - creating more total wire length to reach the terminals of the currently "favored" battery, and less total wire length (i.e., unchanged shorter wire length) to reach the less favored battery. By increasing resistance on the path tho the "favored" battery, it will become less favored - and more even discharge/charge behavior should result.

 

[Steve_S]

There are a few possibilities at play here.
One test to do, is to change the location of where it is connected on the DC Bus Bars. Assuming a 4-bolt busbar and you packs connected to Bolts 2 & 3 and the main DC wire to SCC/Inverter on Bolt-1, swap the bolts between packs, or put Pack-1 on Bolt-1, SCC/Inverter on Bolt-2 and Pack-2 on Bolt-3. Do this on both + & -. Sounds silly but you'd be surprised. IF this makes no change then that potential problem is eliminated leaving ONLY the "pack" internals.

Pack Internals.
A BMS can be a problem, but most often it is a matter it works or it doesn't, very rarely anything else UNLESS there is bad software configuration but that would be fairly obvious quickly into the troubleshooting.
* You did not mention Type of BMS, Type of Cells, IF the BMS is a SmartBMS which can be accessed via software/bluetooth/wifi etc.

I think the one battery has something wrong with the cells, possibly resistance issue?

- Cells known as "Runners" are not uncommon with Grade-B, Bulk and lower graded cells. They can either reach a full SOC ahead of the other cells or reach Empty ahead of the others. In such an instance the BMS would trigger a Cell Protection cutoff and cease charging or discharging. Proper Grade-A Matched & Batched cells do not suffer such issues typically but these of course are most costly.

They are tough to figure out once inside a built pack. If the pack uses screws/bolts for the busbars or if welded makes a difference. A device like the Yaorea 1035+ battery tester can read the voltage & Internal Resistance of the cells. Each cell would have to be charged to the same voltage IE 3.400, allowed to saturate till taking less than 2A, allowed to settle for 1 Hour and test. Then discharge cells to 3.000 and retest again for IR. IF all good the cells should be within 2.mOhm

- Connection failures such as loose busbars, wires etc can also result in faulty readings which show up as capacity losses. Loose connections will Generate Heat ! This can be extremely hazardous leading to fire. Checking all the connections, busbar attachments, BMS Harness etc is important. The BMS Harness could also be faulty (hi resistance or loose) sending false signals resulting in BMS trips, it doesn't take much.

NOTE for SmartBMS.
If you have a smartBMS in these packs that can be monitored then you can watch both packs during the charge cycle. Normal operation will see near identical input voltage to both packs (it does float 0.0## volts between packs) and similarly with the Amps being taken which will also float a "slight bit" as the different cells take the charge and if there is a balancer at work as well. This is normal & expected behaviour but IF you see one pack taking less Amps or Voltage and you look at the cells and one or two are farther out of synch (especially noticeable above 3.375 or below 2.900 if discharging) that is an Indicator of a weak cell or likely "runner". Again a Loose or Dirty Busbar Connection at that cell may also be a culprit.

Loose & Dirty bus bars are the bane of systems, especially for the DIY scene, less so for professional builders. "Professional Builders" is subjective... BattleBorn, SimpliPhi, Relion, Victron, CATL, LG etc = Pro Builder's. Many "PreBuilds" can be "Dubious" if coming from who-knows-where-shops. Then you have companies like SOK, EG4 who are striving to take market share with good quality @ reasonable prices "pro".

Hope some of this helps.

 

[RCinFLA]

10-20% mismatch would be expected. 50% mismatch is excessive.

Try swapping two batteries at their connector to see if it follows battery.

Batteries of different age can cause this. Different BMS's with different BMS MOSFET switch resistance can cause this. Obviously a poor connection or connector could cause it.

If you want to get into battery, on each battery pack singularly, with a moderate load, check voltage drop across BMS. Check difference between no load rested cell voltage for a minute of rested no-load to 0.2-0.4 CA rate moderate load current after loaded for a minute on each battery terminals.

You have to allow 1-2 minutes to allow overpotential (terminal voltage change) to approach equilibrium when cell current is changed.

Older cells will have greater terminal voltage drop (and rise for charging), at one minute load, than newer cells. Watch this video. What he calls a grade B, I would call a grade D-. Neither of the two cells are new grade A cells. A new cell at that load should not have more than 75 mV terminal voltage drop for a 280 AH size cell. You can do test with inverter load if series cells are not too mismatched.

 

[SCmountainman]

Thanks for the helpful replies everyone!

These batteries are 3 months old. Clean/like new terminals. Brand new wiring. Inside a temp controlled battery room. Have never discharged below 50%. Usually only 30% discharge overnight. All inverter settings are programmed properly too per battery tech support.

The batteries are grade A prismatic. 16s1 with impactiv balancing. I cannot mess with them internally due to factory seals. Will mess up my warranty. I don’t have some of the fancy test equipment recommended. I need to step up my game. BMS has Bluetooth capability but only for MFG in their factory or lab. Per tech support. I currently have them buss barred on top terminals and inverter wires on the slow battery. Believe it or not this leads to the smallest differential. Criss crossed inverter cables leads to even larger differential.

Have tried factory buss bars on batteries, have put 4/0 cables on in place of buss bars then back to buss bars. Have tried all connections top and bottom terminals. Have switched inverter connections to different terminals numerous times. All under direction of tech support.

I don’t have any “big loads” to run batteries down. I used a 500watt heat gun and have tested the batteries for 1 hour independently with volt meter. The “fast” battery indeed draws down faster. Indicated readings matches BMS and on two different fluke meters.

Charging in full sun the fast battery will absorb close to 1000 +/- watts and the slow battery absorbs close to 500 +/- watts. Stays this way until the fast one tops out. Then the slow battery will jump to 1000 +/- watts until it tops off. Usually a 10% SOC difference when this occurs, this happens daily.

Every morning the fast battery is showing around 20% lower SOC than the slow battery. I.e. this morning fast battery was at 70% slow battery was at 85%.

I will take some pictures later today.
Thanks again everyone!

 

[mesia]

I have the same issue. What is your invertor/invertors and what is the model of battery packs?