Rosen battery SOC issue

[Francesco]

Hi,
my Growatt SPF5000ES shuts down (error 04, lo battery voltage) with 2 200Ah Rosen batteries showing positive SOC but very low voltage.
I tried to reset the batteries pushing the BMS reset switch but no luck..
The dischaging seems to proceed normally then, suddenly, the battery voltage drops around 42.9V and inverter shuts down.
(see pic's)
As far i undertand the Rosen battery is arranged in 15S4P cells configuration and equipped with a Seplos BMS

 

[sunshine_eggo]

BMS must be synchronized at 100% when fully charged to be reasonably accurate. The reset likely just forced the BMS to guesstimate SoC based on voltage, and that's rarely accurate.

15S @ 47.13 is very close to empty - about 10-15%, so a cut off 13% below when it read 47.13V sounds spot on.

Fully charge your battery to > 52.5-54.75V (will happen somewhere in that range), and the BMS should recalibrate to 100%.

 

[Francesco]

BMS must be synchronized at 100% when fully charged to be reasonably accurate. The reset likely just forced the BMS to guesstimate SoC based on voltage, and that's rarely accurate.

15S @ 47.13 is very close to empty - about 10-15%, so a cut off 13% below when it read 47.13V sounds spot on.

Fully charge your battery to > 52.5-54.75V (will happen somewhere in that range), and the BMS should recalibrate to 100%.

Thank you for your reply.
When (supposedly) fully charged the BMS never reches the 50V...
again: from fully charged to zero I am far from removing 200Ah of power

 

[rodrick]

Invest in a good shunt if you want accurate measurements of charge

 

[sunshine_eggo]

Thank you for your reply.
When (supposedly) fully charged the BMS never reches the 50V...

You need to ignore the % value.

again: from fully charged to zero I am far from removing 200Ah of power

I see no evidence this has ever been done.

Fully charge your battery to > 52.5-54.75V (will happen somewhere in that range), and the BMS should recalibrate to 100%.

When the battery no longer accepts current in that voltage range, you will know it is full. The current should taper to 0A. If it abruptly cuts to 0A, then the BMS has engaged charge protection, and your cells may be imbalanced. If that happens, lower charge voltage to about 0.2V below where it triggered protection.

 

[Francesco]

ok, but since the BMS SOC level seems to drive the Growatt charger, do you believe I should change the setting from LI to user 2 and then manually set the charging tension to 54.75V?

 

[sunshine_eggo]

ok, but since the BMS SOC level seems to drive the Growatt charger, do you believe I should change the setting from LI to user 2 and then manually set the charging tension to 54.75V?

10,000% except start at 51.75V (3.45V/cell) as I expect this battery will have imbalance issues.

But first, is it possible to see min and max cell voltages? If so, when 100% is indicated at about 50V, what are your min/max?

 

[Francesco]

10,000% except start at 51.75V (3.45V/cell) as I expect this battery will have imbalance issues.

But first, is it possible to see min and max cell voltages? If so, when 100% is indicated at about 50V, what are your min/max?

Here there are the single cells details; they refer to a single battery, the other one shows similar values

 

[sunshine_eggo]

Okay. Those look fine. I just wanted to confirm that there were no high outliers triggering the 100%. Those look fine, but unfortunately, until voltages are over 3.40V, voltage does not correlate to balance.

Recommend USE or US2 and set bulk to to 51.8V and float at 51.7. Do not increase it until you observe all 15 cells very close to each other at 3450mV.

 

[Francesco]

good, we have a beautiful sunny day here in Piedmont - Italy - today, so the battery are charging flawlessly
After reaching and checking the corret balance at 3450mV, I can set the values at 54.7/54.6 untill fully charged.
Then shall I reset the bms button? Is it a specific procedure/time for this operation?
Last: do you have (or where can I find) a reliable SOC/V table for such cells?
tks

 

[sunshine_eggo]

good, we have a beautiful sunny day here in Piedmont - Italy - today, so the battery are charging flawlessly
After reaching and checking the corret balance at 3450mV, I can set the values at 54.7/54.6 untill fully charged.

So all 15 cells showed really close to 3450 when current was VERY low - near 0A?

If so:

Bulk to 52.5
Float to 52.4

This should push cells to 3500mV. If they are all very nearly 3500 at 0A, final step:

Bulk to 53.3
Float to 53.2

Once you observe all cells at or near 3550, you can be confident the battery is good and truly fully charged.

You can then revert back to BMS communication or stay in use mode. If you choose to stay in USE mode:

Bulk to 52.5V
Float to 50.6V

Then shall I reset the bms button? Is it a specific procedure/time for this operation?

I am not aware of any button. The BMS should do it on its own.

Last: do you have (or where can I find) a reliable SOC/V table for such cells?
tks

I have corrected your request... "reliable" is not possible. There are certain rules that apply:

3.45V @ 0A input = fully charged.
3.65V @ 0.05C input (5% of rate capacity, i.e., 5A on a 100Ah battery) = fully charged
3.10V or below is at a very low state of charge.

Everything else is just an approximation and MUST be checked after the battery is at rest (no charge or discharge) for at least 5 minutes:

General LiFePO4 (LFP) Voltage to SOC charts/tables 12/24/48V

I've looked at several places for a simple, easy to read at a glance "chart document" for LFP. I could not find anything comprehensive. I discovered an XLS Worksheet in my stored file and took parts of that to create this chart, I did not...

diysolarforum.com

Please note that most 48V LFP batteries use 16S not 15S as 16S better approximates the operating range of 48V systems developed for lead-acid batteries. Do not use the 48V column. Use the Cell V column.

Again, these are approximate and for resting only.

 

[Francesco]

ok, I finally completed the procedure.
Cells V was around 3.550 at the end of the procedure (+/- 20mv)
when I reconfigured the inverter to BMS mode (100%), the battery pack showed a voltage of 50.9 that seems a little bit low.
Let's see in the next days how it will work.

Thank you so much for your precious advice!

 

[sunshine_eggo]

ok, I finally completed the procedure.
Cells V was around 3.550 at the end of the procedure (+/- 20mv)
when I reconfigured the inverter to BMS mode (100%), the battery pack showed a voltage of 50.9 that seems a little bit low.
Let's see in the next days how it will work.

Thank you so much for your precious advice!

50.9V/15 cells = 3.393V/cell - a very reasonable resting voltage AFTER the charge has been removed, and the cells have been allowed to settle.

LFP is like lead acid in that charging typically requires a HIGHER voltage than the operating voltage.

 

[Francesco]

I am closely looking to the behaviour of this BMS but not sure to fully understand.
Please look to the chart taken in the period 17.00 - 20.30 (no PV input at all)
I have the folloing values:
- 47% SOC 48.6V
- 35.1% SOC 48.5V

Isn't strange?

 

[sunshine_eggo]

LFP is notorious for having near constant voltage.



This is for illustration purposes/sanity check only, as it only applies to cell voltages at rest.

48.6/15 = 3.24
48.5/15 = 3.23

10% SoC drops vary from 0.01V to 0.03V per cell, so I don't find anything concerning in your results. Not strange.