My 48V LiFePO4 Battery thinks 53.33V is 100% charged. Is that all right?

[Offgrid Aussi]

This might be totally normal, but the default parameters for the Over Voltage Protection and Recovery in the BMS suggest otherwise (57.60 and 54.00V).

The 48V LiFePO4 Battery is made of 16 EVE LF280K A-Grade Cells (3.2V 280Ah).
They are controlled by a Seplos BMS (3.0 200A with 2A Active Balancer), that indicates that the battery is 100% full, when charged to 53.33V (3.33V per Cell).

The Off-Grid Growatt Inverter (SPF 5000 ES) stops charging when the battery reaches the 53.33 V.
It is communicating with the BMS via Growatt CAN Protocol.

Default Params set in BMS:
Battery Over Voltage Recovery: 54.00 V
Battery Over Voltage Protection: 57.60 V
Cell Over Voltage Recovery: 3.400 V
Cell Over Voltage Protection: 3.650 V

Now I wonder, if these specific cells are fully charged at 3.33V and I need to change the parameters in the BMS accordingly or if I am not using the cells to their full potential and something is holding them back.
Is anyone using these cells as well and is having the same issue?

Your thoughts are greatly appreciated.

 

[BentleyJ]

LFP cells do not reach full charge until they get to at least 3.4V per cell. Refer to the attached charge curves. One is generic from the internet and the other is a screen shot of a top-up float charge from my system the other day.
You can see how the battery keeps absorbing current until it reaches 54.5V then the current drop off rather rapidly. Voltage will relax to 53.6V when the charger shuts off and there is only a small stand by draw from the battery.

Have you verified the voltages mentioned above with a meter?

 

[DIYrich]

Is that 53.33v just before charging (float) stops, or resting voltage after charging stops? 53.33v is close to my battery's resting voltage after a full charge.

 

[Offgrid Aussi]

Thanks for the charts BentleyJ.

Interesting fact, that the voltage is higher when charging, but it makes total sense of course.
And no, I haven't verified the voltages but I should have.
I also have to admit that the 53.33 V were right after the inverter stopped charging, so it might have been higher while it was at it, but I did not pay attention then.
I will monitor the charge process tomorrow, when the sun is up and the battery is charging again. Now ist dark down under.

Thanks again, I will post the results tomorrow.

 

[740GLE]

Be mindful, fully charged resting voltage is 3.33v a cell after a little bit

During charging does current tapper off? or is 50a going into the battery then stop immediately?

IMO if the current is tapering off going from constant current over to constant voltage, the you are only probably losing 5% and is conservative for the cell life. I’m in the mindset I won’t push cells everyday to 100% full but will make sure they get a proper good balance once in a while to keep them in check.

One thing that is beneficial is good balancing happens after 3.4v so if you wanted to bump up the Growwatt charging parameters to 3.45v (55.6v) you won’t harm anything

 

[Offgrid Aussi]

Hi all, Thanks for the advise.
I tried monitoring today, but there was not enough sun to get above 95%. I'll try again tomorrow.

I did find a log file from when I charged the battery via the Growatt Inverter from 240V grid power.
The Over Voltage Alarm was triggered at 57.62 V and subsequently charging stopped.
I have never seen anything above 54 since I charge from solar, but I can't say for sure, as I didn't log the voltage anymore. I do that again now, and hopefully get some serious sun tomorrow.

 

[AshleyL]

Hi all, Thanks for the advise.
I tried monitoring today, but there was not enough sun to get above 95%. I'll try again tomorrow.

I did find a log file from when I charged the battery via the Growatt Inverter from 240V grid power.
The Over Voltage Alarm was triggered at 57.62 V and subsequently charging stopped.
I have never seen anything above 54 since I charge from solar, but I can't say for sure, as I didn't log the voltage anymore. I do that again now, and hopefully get some serious sun tomorrow.

Ah......the great mistake of using CAN communication for Growatt inverter........Hopefully the Seplos BMS accounts for cell internal self discharge and BMS or any attached active balancer power consumption.
Remember, Growatt inverter relies on the reported SOC % to stop charging. If your BMS reports "inaccurate" SOC 100% to the inverter, example reported 100% even though none of the cell voltage reach 3.45V, Growatt inverter will stop charging immediately. Oh, you should know about the dreaded Growatt 95% - 100% algorithm too.

 

[Offgrid Aussi]

Uh.... I didn't know about that. But it sounds like I will have to read up on it. Thanks AshleyL.
Should anybody know how to teach the Seplos BMS what 100% charge are, I'd be interested to know.
Thanks again.

 

[AshleyL]

Uh.... I didn't know about that. But it sounds like I will have to read up on it. Thanks AshleyL.
Should anybody know how to teach the Seplos BMS what 100% charge are, I'd be interested to know.
Thanks again.

Seem like you are Australian.......perhaps you should view Andy of OffGridGarage Australia youtube channel

https://www.youtube.com/@OffGridGarageAustralia/videos

Scroll down to video from one month ago, there should be two videos on Seplos BMS......

 

[marionw]

You might have to micro-manage the battery charge by disconnecting the BMS communications cable. When the BMS tells the inverter it is at 100% SOC the inverter stops charging. Within the BMS this is generally not controlled by alarm or protection setpoints.

Typically when several cells in a battery reach the "Balance" voltage set in the BMS (usually around 3.45v) the BMS will start to balance cells and reset SOC to 100%. Depending on your charge-discharge cycles, over time you might have to once in awhile take control of the charge process to get the batteries to balance.

By disconnecting the BMS from the inverter you can set the charge voltages and current in the inverter (which get overridden with BMS communications) and thus control the charge. In my inverter I can view the various max bulk/absorption and float voltages as well as charge current on the inverter (or monitoring software) with the BMS communications connected and I can view them again with the BMS communications disconnected. With the BMS connected the values are set by the BMS and will revert to what I set in the inverter when the BMS is disconnected.

The Seplos BMS might have PC based software that you can use to monitor the charge as well as possibly change some of the BMS parameters.

 

[IGBT]

Aren't these batteries reaching around 99% charge and all this work is going into getting that last 1%?

 

[Offgrid Aussi]

Yes I've been wondering the same thing IGBT, but assume that between 53.3 and 56.7 there must be more than 1% there.
LFP Voltage Charts for 48V Batteries suggest that 53.3 is more like 85% than 99%.

I understand now, that the charge voltage is much higher than the float voltage, as the 'pressure' that pushes amps into the battery goes away and I can verify that with the log from today. The voltage does indeed climb to the over voltage protection limit set in the BMS (56.7V), and then the battery briefly goes into standby mode before discharging again. Here, the famous Growatt 95% - 100% algorithm kicks in, I think. The battery keeps discharging although there is plenty of sun to be harvested. This is with the CAN communication on Growatt Protocol. I thought I'd test that first. I'm going to cut the coms over the weekend and do another run. And watch the Videos on Andy's Channel, of course.

Thanks for all the comments and support, I'll keep you posted.

 

[gotbeans]

Imagine a battery like a cup of water, put the cup of water under a high flowing water and it'll never fill to the rim, but if you lower the water pressure you can get it right up to the top of the cup rim.

In order to "fill" those batteries at a lower 53.3 voltage it takes several extra hours, amps get lower and lower. The sun isn't out long enough unless they're already charged in the morning then it can sit at 53.3 and will probably fill to 99%.
To counter that we can up the voltage and they'll fill up faster with lower and lower amps.

 

[IGBT]

You could do a capacity test at the 53.3v level and see how many AH you get until the battery is down to near depleted. Maybe you have 85%, maybe you actually do have 99%.

 

[Lt.Dan]

Okay so I know a fair bit about the Seplos BMS's, and I might be able to help.

The Seplos BMS has a few different parameters for triggering 100% SOC. Andy at Off-Grid-Garage has documented what each version of the Seplos required for that 100% calibration, and put it all in a spreadsheet, ill try and find it and link to it.

How are the balance of the cells? The only reason I ask is because one of the triggers is a Cell OVP (Over Voltage Protection), where one cell hits 3.65v, the BMS assumes the pack must be full. So if you are hitting 3.65v on one cell and the pack is at 53.33v, then the BMS is going to tell the inverter to stop charging. This would be highly unlikely because the cell would have to be so crazy out of balance, but I gotta ask just to scratch it off the list.

If the SOC is inaccurately calculated, then when charging it will seem like the BMS hangs at 99% for what seems like forever, but its really waiting for one of the triggers to happen. It will not command the inverter to stop charging until the triggers are met and 100% is registered.

 

[Offgrid Aussi]

@740GLE and @gotbeans That makes sense and I can confirm that the amps taper off towards full charge.

@Lt.Dan Thanks for pointing that out. I have checked and the voltage diff between cells is 5mV at the moment. I have a 2A active balancer installed, that I run every few weeks when the battery is at least 95% full.

@marionw and AshleyL I will give that micromanaging a try now and cut comms to the BMS.
I have switched the Growatt program 05 from LI to USE2 and have set the charging limits (19 and 20) to 55V and the discharge limit (21) to 44V.

The Growatt manual suggests that if you use a lithium battery without BMS comms to set cutoff charging voltage (19) and float charging voltage (20) to the same value. Does that make sense?

Anyway, today I'm starting with 66.6% SOC at 52.52V and we will see how the charging process goes.

 

[Offgrid Aussi]

Ok, now the battery charges to 99.9% at up to 57.0V which then tapers off to around 53.4V at float state.
Under load the charging starts again at 99.7%. So now waiting for the charge to drop to 95%.
Only downside is that the Growatt doesn't log the SOC to the cloud anymore.

Thank you all for your support and insights.

 

[marionw]

The Growatt might only log SOC to the cloud when connected to the BMS.

The "Charge Limit" of 55 volts might be a bit low. Increase to 56 and monitor (16 cells x 3.5 volts per cell).

Another thing you can do (with BMS disconnected) just to get a better understanding of your system is to write down any of the settings (i.e. 19 and 20 and others) that affect how the inverter charges the batteries. Connect the BMS, verify there is BMS communications and then record the previous settings again to see if any where changed with BMS communications.

Don't know if you can change or at least verify any settings in the Seplos BMS such as the voltage cells much reach before the BMS starts cell balancing (normally 3.45 volts)

It appears you have at least made progress (and hopefully learned a bit more about your system and equipment).

 

[Offgrid Aussi]

Yes, I learned a lot. About time too. Thanks again. Running on 56V max charge voltage now.