Pylontech SOC drift below SOC min and ESS in external control oscillates in pumping network

Hello,

I have a problem with ESS, using a quattro 48/10000 and 3 Pylontech US5000 batteries. The ESS minimum SOC is set to 20%. Rather than leaving the battery idle, when the SOC reaches 20%, the battery supplies the house, with the SOC dropping to 15–16%. Then the quattro draws power from the grid at full power to bring the battery back up to 20%. I observe that the quattro is in “external control” during this time (and not in inverter mode).

Single-phase system, venus 3.90 beta 9, quattro on v568, with batteryLife, no peak shaving.

This is quite energy-inefficient. How can I prevent this?

Thanks in advance for your help.

External control is the mode an ESS system will normally be in.
If SOC is drifting, this can happen at low states of charge, especially when there are cell issues, then ESS will behave like this as the SOC suddenly drops beneath the recharge threshold.
This is not inefficient, it is an indication of a battery with a problem, that is under load.
The only solution here, is to raise the minimum SOC to one where the battery SOC will not unexpectedly drop under the same load.

I would also look at cell imbalances and differences between min/max cell voltages at the top of charge and at deep levels of discharge.

ESS requires an accurate SOC to work correctly. Without one, this is normal behaviour.

Thanks for your input.

Batteries are new and there is no cell inbalance (within 10 mV). The drift below SOC min is not sudden, but slow and takes a couple of hours before ESS takes energy abruptly from the network to raise the SOC back to SOC min.

Do you have any DC loads?
What Battery monitor is configured in the GX system setup?
At min SOC, ESS should trigger state #1 and go into passthrough, at this point it should not be possible to discharge the battery unless there are other parasitic loads.

To add, new batteries tend to need time to balance properly. Pylontech historically have had SOC and cell warning issues in the initial period after installation, more so when adding new batteries to existing older batteries.

no DC load except cerbo GX. Battery monitor is pylontech. ESS in on #1 #2.

You will need to share screenshots of configuration of DVCC, ESS (what is the setpoint) etc.
You are also using a beta (and an outdated one), it would be worth trying a production version in case something in the beta is contributing.

If the BMS is the source of SOC in the GX and the battery is being discharged, which should not be possible in states 1 and 2 (batterylife) then it can only be an issue with the BMS itself.
It is not clear from the chart shown what the battery current measurement is once it hits 20%, this would show if the battery is being discharged and by how much.

Pylontech and quattro clearly don’t agree!

Pylontech and quattro clearly don’t agree!

I’ve just updated Venus beta 14. But that hasn’t changed anything.

The maximum charge voltage is set to 51.8V. This is insufficient.

As a result, the battery doesn’t charge properly to 100% and drifts.

You should set it to at least 52.4V. If these are updated US5000s (version 2.3), the BMS indicates 52.8V (CVL). You can trust this value and leave this setting unset in DVCC.

The principle of recharging once the SoC drops 5% below the limit is normal.

Why have you set a charge current limit? And why 140A?

I agree with Nick about the idea of starting to configure your system based on a stable version, unless you really need a new feature that hasn’t yet been incorporated into the stable version (which one?).

I lowered the maximum charging voltage to 51.8 V because the quattro has a VDC voltage measurement error and a DC current measurement error (as seen in a screenshot above). The result was that the DC voltage drifted up to 53 V and the battery went into an error state (I mentioned this in a post recently). Cerbo can’t manage the quattro properly because of this current measurement error. It’s a problem with regulating the DC current rather than the DC voltage. I thought 3.90 beta did a better job than 3.81. By lowering the VDC limit, the battery error occurs later or not at all. I hope Venus will soon be able to regulate things properly.

But the battery balances its cells. No problem on that front.

The charging current limit is that of the quattro (140 A).

In the ESS screenshot, you can clearly see that Pylontech says “discharge” whereas quattro says “charge”. Could that be where the problem lies? In any case, the battery continues to discharge below the minimum SOC.

I disabled “limit battery charge voltage”. Let’s see if it works better at 100% SOC.

Yes, the battery powers the house via the Quattro. No peak shaving.

ESS says “discharge disabled”, but it still discharges.

It’s surprising to have a voltage problem with brand-new Victron equipment.

My first thought would be a wiring fault, but looking at the graph below, I find that hard to believe.

Peak shaving could have been a reason. It allows the loads to continue to be powered when the grid supply is insufficient.

What is the battery configuration?

To avoid any confusion, I can see the BMS is chosen in DVCC, but it must also be configured here, sometimes this is left to be the inverter by mistake: