Battery SOC Behaviour with Growatt SPF 5000ES

[Ricardo_]

Hello, I have a Growatt SPF 5000 ES with a 4,8kWh battery pack, they are connected and working with BMS, i have a strange battery SOC behaviour as you can see by the data below, these values are read from the inverter. As it reaches 85% the battery starts going down very fast without any change in the load, anyone had the same experience what could be wrong?

 

[AshleyL]

Can you post the battery voltage graph too?

 

[Ricardo_]

 

[AshleyL]

This is weird, your battery voltage seems to be fine. Assuming if your battery is using 16s configuration, 3.0V (considered as empty SOC) X 16 = 48V.
From your battery voltage graph, other than the one weird sudden dip at the last part, you still have a lot of power remaining in the battery.
Perhaps there is SOC calibration issue with your battery pack.
What is your battery pack model? DIY or ready-built pack?

 

[Zapper77]

4.8 kWh battery. Is that a 15S?

 

[AshleyL]

4.8 kWh battery. Is that a 15S?

Unlikely, based on your voltage graph, if it is 15s, 3.65v (The max safe limit for Lifepo4) x 15 = 54.75V.
Your voltage graph max out at 55.8V. Are you saying the inverter is overcharging each cell to 3.72V ( 55.8 / 15 )?
Are you using Growatt Hope 4.8L-C1 ? Just tell the model.

 

[Ricardo_]

My battery is a 16s Tensite 48V TS-L5000
The last dip in the end is when a 800 W load is switched on

 

[AshleyL]

My battery is a 16s Tensite 48V TS-L5000
The last dip in the end is when a 800 W load is switched on

Okay, 16s cell it is......any information of BMS protocol being used? Pylontech protocol?


Just check the DIP switch position at your battery.

 

[Ricardo_]

Okay, 16s cell it is......any information of BMS protocol being used? Pylontech protocol?

View attachment 187545
Just check the DIP switch position at your battery.

I use Can protocol as the battery manufacturer recommend for growatt at Chanel 1 in the inverter

 

[AshleyL]

I use Can protocol as the battery manufacturer recommend for growatt at Chanel 1 in the inverter

Hmmm.......Pylontech CAN protocol it is....

Ok, first, the 100% - 85% curve is normal because the battery is being used to power Growatt inverter. (60watt idle power).
Bad news, your battery BMS seems to track the SOC status via voltage level in a wrong way after comparing your battery power, battery state of charge (soc) and battery voltage graphs. A proper one would be in combination of voltage and coulometer tracking.
However, I can't be certain yet.


Question, how long have you been using the battery?
The battery BMS claims to have dynamic SOC calibration (assuming if you can trust the marketing claim).
Let check the individual cell voltage using Tensite app. I assume you have it installed?
Check the voltage of individual cell when the SOC starts to go downhill steeply.
I suspect there might be weak cell or unbalanced cell. You can have 15 "good" cells and 1 "troublemaker" cell to ruin the SOC state.
Example, 3.3v good cell x 15 = 49.5V plus one 2.6V weak cell = 52.1V

 

[Ricardo_]

I installed the battery in September 2023, i have the monitoring system with my raspberry for a week, so only now i started having data on this behaviour that i was already suspicious about .
yes i have the app i checked the SOC in the app is the same as in the inverter.
i will check the individual cell voltage.

 

[Ricardo_]

Hmmm.......Pylontech CAN protocol it is....

Ok, first, the 100% - 85% curve is normal because the battery is being used to power Growatt inverter. (60watt idle power).
Bad news, your battery BMS seems to track the SOC status via voltage level in a wrong way after comparing your battery power, battery state of charge (soc) and battery voltage graphs. A proper one would be in combination of voltage and coulometer tracking.
However, I can't be certain yet.


Question, how long have you been using the battery?
The battery BMS claims to have dynamic SOC calibration (assuming if you can trust the marketing claim).
Let check the individual cell voltage using Tensite app. I assume you have it installed?
Check the voltage of individual cell when the SOC starts to go downhill steeply.
I suspect there might be weak cell or unbalanced cell. You can have 15 "good" cells and 1 "troublemaker" cell to ruin the SOC state.
Example, 3.3v good cell x 15 = 49.5V plus one 2.6V weak cell = 52.1V

 

[AshleyL]

Are those cell voltages captured during the sudden dives after 85%?

Cell voltages are normal (seems very well balanced too) and overall voltage is normal 52.239V.
I suggest you contact your dealer and ask for firmware update for the battery BMS or configure it not to match the SOC state with voltage level.
Lifepo4 has a very flat curve between 10% - 90% of the charge level.

 

[Ricardo_]

Are those cell voltages captured during the sudden dives after 85%?

Cell voltages are normal (seems very well balanced too) and overall voltage is normal 52.239V.
I suggest you contact your dealer and ask for firmware update for the battery BMS or configure it not to match the SOC state with voltage level.
Lifepo4 has a very flat curve between 10% - 90% of the charge level.

I captured it in the morning at 7 am, today it started to dive at 80%, the time i captured the battery SOC was around 53%.

 

[AshleyL]

I captured it in the morning at 7 am, today it started to dive at 80%, the time i captured the battery SOC was around 53%.

After comparing your battery voltage graph and SOC graph, there is a correlation.
As if there is a straight line curve 55.8V = 100%, 51.9V = 50% and 48.0V = 0%
You mention around 53% at 52.239v, right?


55.8V - 48.0V = 7.8V

7.8V​	100% SOC​	55.80V​
7.41​	95​	55.41​
7.02​	90​	55.02​
6.63​	85​	54.63​
6.24​	80​	54.24​
5.85​	75​	53.85​
5.46​	70​	53.46​
5.07​	65​	53.07​
4.68​	60​	52.68​
4.29​	55​	52.29​
3.9​	50​	51.9​
3.51​	45​	51.51​
3.12​	40​	51.12​
2.73​	35​	50.73​
2.34​	30​	50.34​
1.95​	25​	49.95​
1.56​	20​	49.56​
1.17​	15​	49.17​
0.78​	10​	48.78​
0.39​	5​	48.39​
0​	0​	48​

Hereby lies the problem.
Lifepo4 battery will maintain flat voltage level of 3.2V or 3.3V between 10% - 90% = 3.2 x16 = 51.2V or 3.3 x 16 = 52.8V.
Only at the last 0%-10% where the voltage will fell from 3.0V to 2.5V steeply. Same for 90%-100% range where the voltage will suddenly rise steeply between 3.45V - 3.65V.

You really ought to request your dealer/manufacturer not to use voltage as SOC reference.......?

edit:
You sure the BMS is communicating with the Growatt inverter correctly? The more I take a look, the more I think there is no communication between the battery and growatt at all. Cause that pattern seems to be similar to Growatt "US2" mode where by a drop of 2volt resulted in steep SOC state.

Can you verify at Growatt display there is "LI" near the battery logo and percentage being shown there?

 

[Ricardo_]

After comparing your battery voltage graph and SOC graph, there is a correlation.
As if there is a straight line curve 55.8V = 100%, 51.9V = 50% and 48.0V = 0%
You mention around 53% at 52.239v, right?


55.8V - 48.0V = 7.8V

7.8V​	100% SOC​	55.80V​
7.41​	95​	55.41​
7.02​	90​	55.02​
6.63​	85​	54.63​
6.24​	80​	54.24​
5.85​	75​	53.85​
5.46​	70​	53.46​
5.07​	65​	53.07​
4.68​	60​	52.68​
4.29​	55​	52.29​
3.9​	50​	51.9​
3.51​	45​	51.51​
3.12​	40​	51.12​
2.73​	35​	50.73​
2.34​	30​	50.34​
1.95​	25​	49.95​
1.56​	20​	49.56​
1.17​	15​	49.17​
0.78​	10​	48.78​
0.39​	5​	48.39​
0​	0​	48​

Hereby lies the problem.
Lifepo4 battery will maintain flat voltage level of 3.2V or 3.3V between 10% - 90% = 3.2 x16 = 51.2V or 3.3 x 16 = 52.8V.
Only at the last 0%-10% where the voltage will fell from 3.0V to 2.5V steeply. Same for 90%-100% range where the voltage will suddenly rise steeply between 3.45V - 3.65V.

You really ought to request your dealer/manufacturer not to use voltage as SOC reference.......?

edit:
You sure the BMS is communicating with the Growatt inverter correctly? The more I take a look, the more I think there is no communication between the battery and growatt at all. Cause that pattern seems to be similar to Growatt "US2" mode where by a drop of 2volt resulted in steep SOC state.

Can you verify at Growatt display there is "LI" near the battery logo and percentage being shown there?

Yes it is being shown there, these values are being read from the inverter and if i connect with the app it is around the same, as we are talking i am thinking about more a of a BMS problem. I will contact the seller about the issue.

 

[AshleyL]

Yes it is being shown there, these values are being read from the inverter and if i connect with the app it is around the same, as we are talking i am thinking about more a of a BMS problem. I will contact the seller about the issue.

Around the same? Something isn't right here.....it should be exactly the same if BMS communication is working.

Can you check your Growatt Program 05? Refer to your manual book.
Cause if "US2" is selected, Growatt displays "LI" at the display too.

A proper BMS communication will use "LI" instead of "US2"

I use Can protocol as the battery manufacturer recommend for growatt at Chanel 1 in the inverter

Plus, in your picture attachment, your battery is configured to use Pylontech CAN protocol. For Growatt to communicate using Pylontech CAN, it should be using CAN L52 protocol. (What with Chanel 1?)

 

[Ricardo_]

Around the same? Something isn't right here.....it should be exactly the same if BMS communication is working.

Can you check your Growatt Program 05? Refer to your manual book.
Cause if "US2" is selected, Growatt displays "LI" at the display too.
View attachment 187771
A proper BMS communication will use "LI" instead of "US2"





Plus, in your picture attachment, your battery is configured to use Pylontech CAN protocol. For Growatt to communicate using Pylontech CAN, it should be using CAN L52 protocol. (What with Chanel 1?)

Ch1 awas the battery selection asain battery, yes it is using L52 protocol and inverter is on Li BMS mode

 

[AshleyL]

Ch1 awas the battery selection asain battery, yes it is using L52 protocol and inverter is on Li BMS mode

Uh oh ......then as I feared....the problem lies with your battery BMS.
If your BMS SOC percentage is following the voltage value in table I listed up there exactly, then you should really knock the designer/programmer head.....just kidding. No choice but to consult the dealer and ask for solution........

Alternatively.....you can use Growatt "USE" or "US2" mode and use overall pack voltage as reference, but this relies on a VERY BALANCED and MATCHED battery cells (or proper smart active balancer). Otherwise, the BMS might simply disconnect the battery from the inverter (Low voltage or high voltage disconnect protection) and leave you with blackout.

 

[Philectric246]

This sounds a lot like what is happening to me. I am using a Growatt SPF6000T with a Rosen 10kw 200AH 15S battery. I was using the Lithium setting and connecting the battery via CAN L52 but the inverter would switch back to the grid relatively quickly and when I checked the SOC on the battery it was usually very high, sometimes over 80%. Since then I've tried using US2 and USE with limited success. I have had a lot of code 03(low voltage) and code 04(over voltage) errors. When I get the over voltage errors the BMS(I assume) disconnects the panels from the battery, presumably to stop it from overcharging and this cycle just keeps repeating itself until I turn it off. I fixed the overcharging issue by reducing the voltage in setting 19 to 52.5v/3.5v per cell. I still don't understand why it switches back to grid when the battery is almost full.