Eve and Rept cells will not stay balanced when charging (24v x2)

[brandonboosted]

My setup is 8x 280ah Eve Cells and 8x 280ah Rept cells. I am using a JBD 24v Balancer and the 2 banks are connected in parallel with even lengths of wire connecting them. I am using a victron 250/100 controller and it can see up to 100 amps with my panels but averages 40-50a currently in these winter months. I have top balanced these packs 2 times now and over time they just get completely out of whack as it gets closer to full (27.40v-28.40v) As the battery discharges the cells all balance within .10mv without a problem and that is around 26.7v per bank. Here is one bank as it sits with the charger in absorption and the cell over voltage triggered-
1. 3.432v
2. 3.417v
3. 3.471v
4. 3.420v
5. 3.603v
6 .3.608v
7. 3.453v
8. 3.424v

The other pack is very similar. The entire time drawing down the batteries there is no problems only when topping up. I have ran the REPT cell bank by itself in the past before I received my eve cells and didn't have this problem. Only major differences since then are the fact I had half the charge amperage, now being parallel with another pack, and I was using a dally 200a bms on the rept cells. The setting in the bms is set to balance when charging & discharging.

Any help is appreciated!

 

[sunshine_eggo]

The working range of LFP is always very consistent even if cells are a few % out of balance. It's when you get into the upper and lower legs that the shitshow comes to town.

Triple check that all your BMS sense wires are properly crimped and attached.

Given that all the other cells are above 3.40V, your level of imbalance is probably VERY small.

Things to try:

1. Ensure that the BMS is set to balance ONLY during charge, above 3400mV/cell and 20mV difference. If you balance all the time, you can actually undo a top balance.
2. Hit each of the low cells with your 10A power supply INDIVIDUALLY while connected in series. You may find that it only takes 5-15 minutse to bring all the other cells up to 3.65V. This confirms you're dealing with a very small imbalance. You do not need to break the battery down to do this.
3. Consider reducing your absorption voltage. If you have surplus solar most days (charged before noon), you can get to 98%+ SoC @ 3.45V with a long absorption time. You may need to increase your FIXED absorption time and cut the tail current to 0.03C or so. Given your already low C rate, 40-50A vs. 560Ah capacity, you're likely not getting much capacity above 27.6V anyway.

 

[brandonboosted]

The working range of LFP is always very consistent even if cells are a few % out of balance. It's when you get into the upper and lower legs that the shitshow comes to town.

Triple check that all your BMS sense wires are properly crimped and attached.

Given that all the other cells are above 3.40V, your level of imbalance is probably VERY small.

Things to try:

1. Ensure that the BMS is set to balance ONLY during charge, above 3400mV/cell and 20mV difference. If you balance all the time, you can actually undo a top balance.
2. Hit each of the low cells with your 10A power supply INDIVIDUALLY while connected in series. You may find that it only takes 5-15 minutse to bring all the other cells up to 3.65V. This confirms you're dealing with a very small imbalance. You do not need to break the battery down to do this.
3. Consider reducing your absorption voltage. If you have surplus solar most days (charged before noon), you can get to 98%+ SoC @ 3.45V with a long absorption time. You may need to increase your FIXED absorption time and cut the tail current to 0.03C or so. Given your already low C rate, 40-50A vs. 560Ah capacity, you're likely not getting much capacity above 27.6V anyway.

Wow thank you for the very informational reply. I will try your suggestions and post back. I changed my Absorption voltage to 27.6 as you suggested how long do you suggest setting my fixed absorption time and float voltage?

 

[sunshine_eggo]

Wow thank you for the very informational reply. I will try your suggestions and post back. I changed my Absorption voltage to 27.6 as you suggested how long do you suggest setting my fixed absorption time and float voltage?

Float at 3.4V/cell

Honestly, I typed that absorption time bit before I thought about your low C rate (I added the C rate sentence to that paragraph after some noodling). Try 2 hours and see what your charge current is at termination. Make sure it's on Fixed and not Adaptive or Adaptive+BatterySafe. It will cut absorption short if not on Fixed.

If you're using a battery monitor, you may need to tweak your "charged" parameters to get it to reset to 100% consistently.

Another possible benefit of 3.45V/cell charging: daily charging to 100% is the most stressful portion of the cycle for LFP. Reducing the voltage to 3.45V/cell vs. the typical 3.55-3.65V/cell may offer improved cycle life.

 

[brandonboosted]

Okay, re-bulk offset at .20V, tail current disabled or .03A?

 

[sunshine_eggo]

I like to have my re-bulk delayed to allow for transient loads and brief dips in solar - 0.8V.

Tail current of 0.03C = 0.03 * 560A = 16.8A.

If you have constant background loads, you may want to add that current to the tail current, e.g., my background loads are about 80W. 80W/27.6 = 2.9A. In that case, tail current should be about 16.8+2.9 = 19.7A.

Generally speaking most LFP are full @ 3.65V/cell and 0.05C. In your case, that would be at 28A @ 29.2V. Your winter peak is around 0.1C, so almost all of your capacity gain is happening at lower voltages. Since we're setting a lower absorption voltage, we're also lowering the tail current.

If you find that you're not getting charged fast enough or float is triggering too soon:

1) lower tail current if tail current is triggering float
2) increase absorption time if absorption is triggering float
3) increase absorption voltage in 0.1V increments to speed charging. 28.4V is a practical maximum. Almost nothing to be gained above that.

Consider also that you will still charge to 95% full at 3.4V/cell given enough time.

The way I look at it is if I can get an LFP battery to very near 100% by end of solar day, I'm charging fast enough, reducing cell stress, improving cycle life and maximizing balance time. Poor solar, particularly in the afternoon and over-utilization during peak solar can also contribute to insufficient charge on a given day.

 

[brandonboosted]

A little update, I have done-

1. Ensure that the BMS is set to balance ONLY during charge, above 3400mV/cell and 20mV difference. If you balance all the time, you can actually undo a top balance.

3. Consider reducing your absorption voltage. If you have surplus solar most days (charged before noon), you can get to 98%+ SoC @ 3.45V with a long absorption time. You may need to increase your FIXED absorption time and cut the tail current to 0.03C or so. Given your already low C rate, 40-50A vs. 560Ah capacity, you're likely not getting much capacity above 27.6V anyway. (absorption set to 2 hours float at 27.20V)

Triple check that all your BMS sense wires are properly crimped and attached.

As the batteries were taken down to 26.0V throughout the night they stayed within 20mv all looked healthy without balancing during draw. I haven't pulled below 26.0V yet but have some cloudy weather coming and I am hopeful I can test the bottom side of the batteries as well. As the batteries currently sit right now in absorption they are out of balance still but only with a delta of 48mv compared to the almost 200mv yesterday. I have not performed the top balance yet and I'm assuming since the voltage was only taken to 27.6 instead of my previous 28.4 it kept the over charged/balanced cells from hitting the max voltage level and or they are slowly leveling together with the other cells. What are your thoughts on this? Also my max tail current allowed is 10a what do you recommend there?

Thank you

 

[sunshine_eggo]

A little update, I have done-

1. Ensure that the BMS is set to balance ONLY during charge, above 3400mV/cell and 20mV difference. If you balance all the time, you can actually undo a top balance.

Good. This should help and might be sufficient to get you back on track without additional single cell charging. If it was set to balance all the time, it likely un-did your top balance.

3. Consider reducing your absorption voltage. If you have surplus solar most days (charged before noon), you can get to 98%+ SoC @ 3.45V with a long absorption time. You may need to increase your FIXED absorption time and cut the tail current to 0.03C or so. Given your already low C rate, 40-50A vs. 560Ah capacity, you're likely not getting much capacity above 27.6V anyway. (absorption set to 2 hours float at 27.20V)

Just make sure you're happy with the state of charge at the end of the day.

Triple check that all your BMS sense wires are properly crimped and attached.

Good.

As the batteries were taken down to 26.0V throughout the night they stayed within 20mv all looked healthy without balancing during draw. I haven't pulled below 26.0V yet but have some cloudy weather coming and I am hopeful I can test the bottom side of the batteries as well.

Almost certain you'll see notable deviation in the lower knee. My gut says you're still much closer to a top balance than a bottom balance, but it's not a bad idea to actually KNOW what it looks like down there (that's what she said?).

As the batteries currently sit right now in absorption they are out of balance still but only with a delta of 48mv compared to the almost 200mv yesterday. I have not performed the top balance yet and I'm assuming since the voltage was only taken to 27.6 instead of my previous 28.4 it kept the over charged/balanced cells from hitting the max voltage level and or they are slowly leveling together with the other cells. What are your thoughts on this?

Exactly. As the battery is peaking at a lower voltage, it's taking longer to charge and feeding it progressively lower current as it fills. This gives more time for the cells to balance. For reference, a typical passive balancer will take a full 24 hours to burn off 1Ah of spare capacity

Also my max tail current allowed is 10a what do you recommend there?

10A then. My gut says it's going to transition to float due to the long absorption time, but it's hard to say with your relatively low initial charge current.

 

[brandonboosted]

After a couple weeks of intermittent trail and error, I am still having the same issue of one cell in each bank are hitting single cell overvoltage protection. With the absorption at 27.60v (CV) the cell slowly creeps up until it hits 3.65. I have tried hours of pulling down the single cell just for it to creep back up. Unsure if its relevant but the REPT bank is cell #5 and Eve bank cell #6. I understand energy flows to the path of least resistance so I am wondering if these means the cells are bad and what my options would be as far as getting the most out of them if its possible. I'm assuming id have to drop my absorption voltage to the point the cell doesn't hit protection but I may loose a lot of capacity. In absorption mode I have to pull around 3-4A out to prevent the cell from increasing by itself. I have quadruple checked connections and I have ensured that the bms is trying to balance the cell its just not enough current. I have also tried to charge the individual banks alone (only by turning charge/discharge off remotely). Really getting frustrated, especially with fact I need maximum. capacity in the winter months.

 

[sunshine_eggo]

After a couple weeks of intermittent trail and error, I am still having the same issue of one cell in each bank are hitting single cell overvoltage protection.

Have the lower cells caught up at all?

With the absorption at 27.60v (CV) the cell slowly creeps up until it hits 3.65. I have tried hours of pulling down the single cell just for it to creep back up.

Pulling down how?

When one is at 3.65V, what are the others at?

What current are you running at this time?

Unsure if its relevant but the REPT bank is cell #5 and Eve bank cell #6. I understand energy flows to the path of least resistance so I am wondering if these means the cells are bad and what my options would be as far as getting the most out of them if its possible.

With cells in series, there is only one path - through all cells equally.

Have you tried a deep discharge and observed the same high cells hitting the low knee and dropping out first?

I'm assuming id have to drop my absorption voltage to the point the cell doesn't hit protection but I may loose a lot of capacity.

3.4V/cell at ~3A is about 95% SoC.

In absorption mode I have to pull around 3-4A out to prevent the cell from increasing by itself.

Pulling out how?

I have quadruple checked connections and I have ensured that the bms is trying to balance the cell its just not enough current.

30mA isn't much.

I have also tried to charge the individual banks alone (only by turning charge/discharge off remotely).

What about individual cells? Have you attempted to charge the low cells to catch up with the high cells?

Really getting frustrated, especially with fact I need maximum. capacity in the winter months.

How much capacity are you actually using?

 

[brandonboosted]

Have the lower cells caught up at all?
As I discharged the one problem cell the bank averaged the 3.44 range opposed to the 3.37 average I had without discharging the problem cells.

Pulling down how?
I just meant discharging the single problem cells.

When one is at 3.65V, what are the others at?

3.44V

What current are you running at this time?
+2-4A


With cells in series, there is only one path - through all cells equally.

Ok thank you for the explanation.

Have you tried a deep discharge and observed the same high cells hitting the low knee and dropping out first?

I actually forgot I did this and I don't have the data because it happened at night while sleeping and my heater kicked off because both banks went into low cell protection but I do remember each bank had one cell only that was way off similar to the over voltage delta. If i remember correctly the other cells were around 2.90 and the single cell was 2.50. I can retest this to determine if its the same cell or not. If it is the same cell what would that mean for the cells?

3.4V/cell at ~3A is about 95% SoC.



Pulling out how?

I'm referring to myself pulling -4A drain on the single cell to keep the battery even with the rest of the cells.

30mA isn't much.



What about individual cells? Have you attempted to charge the low cells to catch up with the high cells?
I have not since that would be 7 cells. Do you think this is something I should try? all of the other cells are even with each other.


How much capacity are you actually using?

Roughly 350-400AH.

 

[brandonboosted]

Another thing worth noting is that as soon as discharge begins and single cell voltage is around 3.35ish (26.8V) the delta is .003 everything and is perfectly synced.

 

[brandonboosted]

 

[brandonboosted]

More pics

 

[sunshine_eggo]

Please try to hit enter and type below my quoted text, otherwise your reply gets cut out when I reply again.

3.44V @ 2-3A of charge is at an extremely high state of charge, likely 98% or higher. If that 2% is really important to you, then you need more battery/solar.

I see now that you're using a capacity tester.

Pull 5.6Ah (2%) out of both peaking cells now. See how that changes the dynamic. Keep doing that until those cells stop being the high cells.

On 4 separate 8S batteries, the strategy of using a power supply on individual cells rectified the problem and restored balance. I would seriously consider trying it. In one case, the 8 cells of the two separate batteries only took 5-10 minutes of charging @10A to 3.6V to completely restore the top balance.

In the second case, the battery was horrifyingly imbalanced, and it took 45minute to 5 hours of charging at 30A to restore balance. I doubt that you're to that level.

I actually forgot I did this and I don't have the data because it happened at night while sleeping and my heater kicked off because both banks went into low cell protection but I do remember each bank had one cell only that was way off similar to the over voltage delta. If i remember correctly the other cells were around 2.90 and the single cell was 2.50. I can retest this to determine if its the same cell or not. If it is the same cell what would that mean for the cells?

Yes, it's important to identify that the cells dropping out first are not the same as the peaking cells, UNLESS ALL cells are at a very low SoC. Drop out at 2.5 with all the others at 2.9 is not necessarily an indication of a major problem.

I'm referring to myself pulling -4A drain on the single cell to keep the battery even with the rest of the cells.

Rather than pulling to try and keep the cell below 3.65V, pull out fixed amounts at peak current as described above.

Roughly 350-400AH.

If you're only getting 350-400Ah out of these two batteries together, I would be concerned. Did you capacity test your cells?

Another thing worth noting is that as soon as discharge begins and single cell voltage is around 3.35ish (26.8V) the delta is .003 everything and is perfectly synced.

That's to be expected and not something to use to discount the top balance issues. The working voltage of LFP is extremely flat, so cells at different SoC under load, between the knees, will rarely be out by much unless there is horrific imbalance (near empty vs. near full).