My final experience and some practical advice
Based on my experience, there are a couple of things that may help others facing similar behaviour.
First, I would strongly recommend adding additional battery interconnection cables, as recommended by the battery manufacturer. In my case, using 2×70 mm² cables significantly improved the predictability of inverter switching during grid loss.
Another thing that helped was adjusting the inverter protection thresholds. I lowered the shutdown voltage to 37.2 V and set the restart voltage to 42.1 V, which reduced the number of unexpected shutdown events.
However, based on everything I observed, my conclusion is the following.
A single battery simply cannot deliver the short-term peak current that the inverter demands during certain transitions. This peak demand seems to be either a hardware or firmware characteristic of the inverter, and apparently it is considered normal behaviour by design.
I discussed this with both technical support and official installer, and they all consistently told me the same thing:
for a 10 kW inverter you should ideally have at least a 20 kWh battery bank.
Personally, I’m not even sure that total capacity alone is the real solution. I suspect the number of batteries (parallel current paths) matters more. For example, having two batteries capable of delivering ~200 A peak each would likely solve the issue, because the inverter could draw the transient current from multiple sources simultaneously.
Right now, in many setups there is only a single battery with a BMS limited to ~200 A, which simply isn’t enough for the inverter’s short-term current demand.
Is this behaviour normal?
In my personal opinion - absolutely not. Considering how excellent the rest of the Victron ecosystem is, this particular behaviour was extremely disappointing.
Unfortunately, after a long period of troubleshooting I eventually made the difficult decision to sell my MultiPlus inverter and switch to a Deye inverter. The Deye system certainly has its own disadvantages, but it works with my battery without any issues and fulfills my primary requirement - acting as a large UPS for the flat.
I do not recommend repeating my path. I still believe Victron builds one of the best systems on the market. If you have the option to add a second battery, that is likely the simplest and most reliable solution.
At the same time, I would strongly encourage Victron engineers to look carefully at this behaviour. Situations where a system with only ~500 W load can trigger transient events equivalent to tens of kilowatts during grid transfer should probably be revisited.
For reference: all assumptions about possible grid instability were eliminated after replacing the inverter. The system now works perfectly under the same conditions. I even went further and installed a 12 kW inverter, and my 16 kWh battery works with it without any issues at all.
Apologies if this message sounds emotional. I genuinely wanted this system to work, but I simply did not have the physical space to install additional batteries, and using a smaller inverter was not an option for me.
I hope my experience helps others facing similar problems.
):
