LifePo4 cells getting unbalanced when drop below 3V

[MisterSandals]

I've observed that cells not getting charged above 3.32V and dropping to the 3.28V level very quickly now.

What is preventing charging above 3.32V?
At this voltage the cells will absorb a LOT of energy without the voltage rising (much). This is a good thing so keep charging!

 

[flashmark]

What is preventing charging above 3.32V?
At this voltage the cells will absorb a LOT of energy without the voltage rising (much). This is a good thing so keep charging!

It seems no capacity to hold the current. Even though charged at 30A, it discharges very quickly (I was charging through an inverter - Solar power). I saw that cells discharged quickly to 3.29V level even with a small load.
Now the usable capacity is in between 3.29 - 3.1V and That is very low.

 

[MisterSandals]

Even though charged at 30A, it discharges very quickly

If the batteries are accepting 30A of charge, you need to just keep charging it. Some of the discharge you are seeing is surface charge and some is being absorbed into the cells further.
If the charge current stops or drastically slows at 3.3V then something is wrong. Right now I think the problem is impatience.

 

[flashmark]

If the batteries are accepting 30A of charge, you need to just keep charging it. Some of the discharge you are seeing is surface charge and some is being absorbed into the cells further.
If the charge current stops or drastically slows at 3.3V then something is wrong. Right now I think the problem is impatience.

It was charged to 54.6V level in the first charge after the top balanced (at 3.65V and keep a day). The 2nd time, it started to kill the charge around 54V. The third time started to kill charge around 53.6V. That's why worried. Is this normal? On the first day, I was able to run through my battery pack the whole night. 2nd day little bit less and the third day was the worst so far. There is no huge difference in consumption.

 

[MisterSandals]

The 2nd time, it started to kill the charge around 54V. The third time started to kill charge around 53.6V.

What do you mean "it started to kill the charge"?
The charger slowed the charging amps?
The BMS cut out due to cell over volt? or ??

 

[MisterSandals]

Even though charged at 30A, it discharges very quickly (I was charging through an inverter - Solar power).

I reread from the top and cannot see any mention of what charger(s) (only solar charger?) you are using.
It would be helpful to know what charger you have and what the charge parameters are (a screenshot of charge setting would be mighty helpful).

I think most of your problems are from not having enough charge on your battery and you are experiencing very typical cell divergence at the low state of charge range (we top balance and expect balance cells there, not necessarily at low state of charge).

 

[Ampster]

How can I solve this problem?

There is no way to add capacity. You can balance at the top or balance at the bottom..There is an alternative to spend a lot more and purchase Automotive grade matched cells. Using 90 percent of cells that cost $150 per kWh is more cost effective than getting 100 percent out of $250 per kWh cells.
I actually only use 80 percent of my pack's capacity because I believe my pack will last longer that way.

 

[Ampster]

If you buy cells and they are from the same manufacturing batch and were tested cycled together ......

I don't think batches or testing is any guaranty that cells will be matched. The matching process is a result of binning cells after testing and matched cels do not necessarily come from the same batch.

Active balancers are useless if you have a bad cell or good cells that were not at the same state of charge when assembled.

It all depends. I top balanced 48 cells and when I assembled them into a 3P16S pack my 2Amp JK active balancer got them to .001 Volts of each other at the top within days in a few cycles. My $600 automotive grade BMS would have taken weeks and many cycles. @John Frum said essentially the same thing. My pack is 42 kWhs.

 

[hwse]

It really sounds like you are just not charging any where close to full. 3.45v/cell should be the lowest stopping point for your charge voltage. I use 3.5v/cell on mine. If what you mean by "kill the current" is that the current begins to drop off, that will be normal when you get to the max charge voltage set point. The way that the charger controls the voltage is by reducing the current to the level that will produce the desired voltage. Try setting your bulk/absorb to 55.2v. Given that the battery sounds like it is deeply depleted, it will take a long time to get to that voltage. Some screenshots from your BMS as you get near the upper voltage limits would greatly help us in understanding what you are seeing.

 

[flashmark]

What do you mean "it started to kill the charge"?
The charger slowed the charging amps?
The BMS cut out due to cell over volt? or ??

I reread from the top and cannot see any mention of what charger(s) (only solar charger?) you are using.
It would be helpful to know what charger you have and what the charge parameters are (a screenshot of charge setting would be mighty helpful).

I think most of your problems are from not having enough charge on your battery and you are experiencing very typical cell divergence at the low state of charge range (we top balance and expect balance cells there, not necessarily at low state of charge).

What I meant was after charging above 53.6V, battery cells discharge very quickly with a small load.
So charging beyond the 3.35V level is useless it seems.
The solar charger keeps charging there is no voltage drop.

 

[flashmark]

It really sounds like you are just not charging any where close to full. 3.45v/cell should be the lowest stopping point for your charge voltage. I use 3.5v/cell on mine. If what you mean by "kill the current" is that the current begins to drop off, that will be normal when you get to the max charge voltage set point. The way that the charger controls the voltage is by reducing the current to the level that will produce the desired voltage. Try setting your bulk/absorb to 55.2v. Given that the battery sounds like it is deeply depleted, it will take a long time to get to that voltage. Some screenshots from your BMS as you get near the upper voltage limits would greatly help us in understanding what you are seeing.

Is there any use in charging beyond 3.35V because cells can not hold the current (No capacity), right? It seems the difference between 3.35V and 3.45V is less than 1% in terms of capacity.

 

[robot_zombie]

it is normal for voltage to drop pretty quickly to ~3.25-3.3 once load is applied, but you need to keep charging until at least 3.45-3.5V, and ideally maintain that voltage until the current drops (~0.05C or so, <10A) for the pack to be fully charged. the cells will keep taking charge as mentioned without the voltage rising quickly then rise quite quickly after ~3.45 or so. The BMS SOC wont be very accurate until you get at least one complete charge cycle.

 

[hwse]

Is there any use in charging beyond 3.35V because cells can not hold the current (No capacity), right? It seems the difference between 3.35V and 3.45V is less than 1% in terms of capacity.

The cells will tell you when they are full. If they are still taking significant amps without the voltage spiking (no single cell over 3.65v) then the battery is not full.
At 3.35v/cell, my batteries are still taking my full charge current, so they are not full. I set the chargers for absorption at 14.0v (3.5v/cell) and it will take several hours at 100A of charging to go from 3.35v to 3.5v. Once the battery gets to 14.0v, the current rapidly drops down to around 20A and the charger continues to hold it at 14.0v for 30-minutes of absorption before dropping to 13.4v float.

 

[hwse]

What I meant was after charging above 53.6V, battery cells discharge very quickly with a small load.
So charging beyond the 3.35V level is useless it seems.
The solar charger keeps charging there is no voltage drop.

Can you post a screen shot of your settings page so that we can see how you have set up your bms? You will need to enter your password to get access to it.

 

[flashmark]

The cells will tell you when they are full. If they are still taking significant amps without the voltage spiking (no single cell over 3.65v) then the battery is not full.
At 3.35v/cell, my batteries are still taking my full charge current, so they are not full. I set the chargers for absorption at 14.0v (3.5v/cell) and it will take several hours at 100A of charging to go from 3.35v to 3.5v. Once the battery gets to 14.0v, the current rapidly drops down to around 20A and the charger continues to hold it at 14.0v for 30-minutes of absorption before dropping to 13.4v float.

Can you please explain why I need to charge to 3.65V since there is a 1% capacity after 3.4V? Because I have to use solar or grid current for many hours to get cells into that level (3.65V) but since batteries can not hold that, there is no use for it, right?

 

[John Frum]

Can you please explain why I need to charge to 3.65V since there is a 1% capacity after 3.4V? Because I have to use solar or grid current for many hours to get cells into that level (3.65V) but since batteries can not hold that, there is no use for it, right?

I know you didn't ask me but I will answer anyway.

You don't need to charge to 3.65 volts per cell but you do need to charge into the high knee on a fairly regular basis so that your bms can maintain the top balance of your cells.

If you have grade b cells its even more important to do this because they are less evenly matched and tend to drift more.

The high knee starts at about 3.425 volts per cell.
I typically advise to charge to 3.55 volts per cell because it is well into the high knee but still gives .1 volts headroom per cell.

 

[John Frum]

This pdf contains a plot of an LFP charge curve.

Dropbox

www.dropbox.com

Thanks to Andy at @Off-Grid-Garage

 

[sunshine_eggo]

This pdf contains a plot of a typical LFP charge curve.

Dropbox

www.dropbox.com

Thanks to Andy at @Off-Grid-Garage

Worth noting that those curves are for an extremely low charge rate. 5A vs. 110Ah capacity.

Higher charge currents will raise the curve overall, make the leg less pronounced, and increase the duration of the upper leg. This is a 0.75C charge:




EDIT: Please note that this particular setup is subject to a .05V drop, so the whole curve should be shifted down .05V

 

[hwse]

You don't need to charge to 3.65v/cell. I was suggesting that you charge to 3.5v/cell average. But you are likely to have some imbalance at the top end so the 3.65v for the highest cell is a protection cutoff. from 3.5v to 3.65v there is very little additional capacity, so I personally don't go there. But you are only charging to 3.35v and I have found that it takes a long time to charge from 3.35v to 3.5v.

I have one drop-in 460Ah pack that has a very large imbalance at the top and bottom so that I cannot charge it to any voltage above 14.05v without tripping the OVP at 3.65v on the lowest cell as seen in the first screen shot. In spite of this large 248mv inbalance, this battery gave me 445Ah out of the rated 460Ah (3.26% short). The second screen shot shows the same pack during charging after my full capacity testing. Notice that the average cell voltage is 3.35v but it is only at 41% of capacity and that was actually much higher than actual due to the FET locked error.

Voltage is not a good indicator of capacity. The only way that you know that the cells are full is when they quite taking in current at a set charging voltage which must be above the cell resting voltage. anything over 3.45v will do for most of the capacity but set charging voltages below that level are insufficient.

 

[hwse]

Have you been into the advanced setting on your JK? I have asked for screen shots of the advanced settings a couple of times, and you have not posted any which makes me wonder if you have been onto that page yet. On mine, the default settings were way out of whack starting with the fact that it was set for LI-ION and not LIFEPO4 as it needed to be.

Here is a link where Andy goes into the advanced setting page with his recommendations. I find them to be a very good starting point.

 

[flashmark]

Have you been into the advanced setting on your JK? I have asked for screen shots of the advanced settings a couple of times, and you have not posted any which makes me wonder if you have been onto that page yet. On mine, the default settings were way out of whack starting with the fact that it was set for LI-ION and not LIFEPO4 as it needed to be.

Here is a link where Andy goes into the advanced setting page with his recommendations. I find them to be a very good starting point.

Sorry. Here it is.

 

[John Frum]

Worth noting that those curves are for an extremely low charge rate. 5A vs. 110Ah capacity.

Higher charge currents will raise the curve overall, make the leg less pronounced, and increase the duration of the upper leg. This is a 0.75C charge:

Is that voltage measured at the battery terminals or at the charger terminals or something else?

 

[hwse]

Sorry. Here it is.

Those look like mostly default LFP settings, I would suggest changing the OCP Delay from 300-s (5-minutes) to something faster. I also changed the UTP from -20ºC to +3ºC and UTPR from -10ºC to +5ºC.

 

[flashmark]

According to this chart, charging above 3.4V is not good for cell health. Is this wrong?

LiFePO4 Voltage Charts (1 Cell, 12V, 24V, 48V)

Explore the LiFePO4 voltage chart to understand the state of charge for 1 cell, 12V, 24V, and 48V batteries, as well as 3.2V LiFePO4 cells.

cleversolarpower.com

 

[hwse]

According to this chart, charging above 3.4V is not good for cell health. Is this wrong?

LiFePO4 Voltage Charts (1 Cell, 12V, 24V, 48V)

Explore the LiFePO4 voltage chart to understand the state of charge for 1 cell, 12V, 24V, and 48V batteries, as well as 3.2V LiFePO4 cells.

cleversolarpower.com

For long term I would say yes. That is why I set float at 13.4v (3.35v/cell) or lower. You still need to raise it to a level above this to recharge the battery (bulk charging).
In the report you posted, they call for bulk charging at 3.65v/cell but that can only be done if you are perfectly top balanced. In real life, I think it is better to charge to a lower bulk voltage of 3.5v/cell with a short absorption time held at that voltage to get the cells saturated before going to a lower float of 3.35v/cell or lower after that