valentin-silvera avatar image
valentin-silvera asked

Current fluctuation in the battery when SOC =100% off-grid system

I have a situation with an offgrid system that I would like to share with you to try to resolve.

This is the list of equipment used:

3x Quattro 48/15000/200-100/100 230V

1x Cerbo GX

1x GX Touch 70

1x Orion-Tr 48/24-5A (120W) Isolated DC-DC converter

6x SmartSolar MPPT RS 450/200-Tr

2x Inversor Fronius Symo 15.0.3

15x CEGASA eBick 280 PRO

The battery technology is LFP, with 15 280 Ah prismatic cells in series.

The layout is shown in figure 1. The Fronius is connected to the AC outlet. There is no dedicated energy meter for the load.

When the SOC reaches 100%, the voltage requested by the BMS (CVL) drops and enters a current fluctuation between charging and discharging. The battery charges and discharges at low current and quickly (figure 2). The attached video help understands the situation.

The system has been running for more than 3 months with the same behavior.

I have 2 hypotheses for the origin of this problem:

  1. The CVL drops from 51.9V to 50.9V. In this condition there is a voltage drop of ~0.2V due to loss in the wire. It may be that the voltage reaching the battery is very close to the battery's open circuit voltage. A possible solution would be to adjust the value requested by the BMS by increasing the voltage a little.
  2. At the same time that SOC = 100%, AC consumption also starts to fluctuate a lot, which to me doesn't make much sense. However, as the system has AC photovoltaic coupling and does not have an energy meter dedicated to the load, the AC Consumption value is an estimated value and not measured. The system then tries to reach the charge value and in this adjustment the battery compensates for the estimation error. There is also the fact that the CCL only goes to 41A and may be too low a value for the system to control. In this case, the solution would be to install a dedicated energy meter for the load.

Thank you very much to anyone who can help me!1.jpegLAYOUT.jpeg2.jpeg3.jpegvideo.mp4

battery chargingoffgrid
1.jpeg (140.0 KiB)
layout.jpeg (39.7 KiB)
2.jpeg (144.6 KiB)
3.jpeg (107.9 KiB)
video.mp4 (9.1 MiB)
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

2 Answers
snoobler avatar image
snoobler answered ·

Likely related to your AC coupled Fronius per #2. The Quattros will shift their frequency away from 50Hz to force the Fronius to disconnect when the battery reaches absorption/charged voltage. Basically the Fronius is connecting/disconnecting its output

Why would you have a GT inverter in an off-grid system? GT inverters do not regulate themselves. They expect to have an infinite grid in which they can dump their full array output at all times - not a great option when your grid is less than infinite. The Quattros must alter the grid to force the Fronius to disconnect.

As you've demonstrated, the BMS is sending different control parameters to the GX. This could also be due to internal cell imbalance within the batteries. Have you evaluated the min/max cell values during these periods?

Try limiting the charge voltage to 51V in DVCC. If that doesn't work, try 50.25V.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

valentin-silvera avatar image
valentin-silvera answered ·

Hi @snoobler , thanks for the answer and help
I used Fronius inverters because it is more efficient for powering the AC load and cheaper than mppt. Using the AC-coupled PV assistant, frequency control in offgrid mode works very well. The problem arises when SOC = 100% and the BMS demands very low current and voltage, so I believe that in these conditions the system is unable to fine-tune.

Unfortunately, the Carlos Gavazi energy meter that I need was not available in my country. We are purchased but it takes 4 months to arrive.

As another alternative, I am talking to the battery manufacturer about changing the CVL voltage values.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

snoobler avatar image snoobler commented ·
I think you've found the consequences of your choice to go with AC coupling over MPPT.

My suggestion with having DVCC limit charge voltage to 51V was intended to get away from the BMS communicating a reduced current. This effectively changes the CVL voltage values as you have requested from the manufacturer. This may worsen or improve the situation.

Did you try it?

0 Likes 0 ·
valentin-silvera avatar image valentin-silvera snoobler commented ·

Hi, just to update on what happened. I get the problem by updating the voltage values requested by the BMS to slightly higher values. To do this I had to ask the battery manufacturer to do a remote update.

0 Likes 0 ·