Hi everyone,
I’m hoping one of the Victron developers can clarify whether the Dynamic PV Power Limiting implementation in Venus OS supports this particular configuration.
Existing system
I currently have a three-phase ESS system consisting of:
Cerbo GX running Venus OS 3.62
Three-phase Victron ESS (Quattro 15000/48)
3 × SMA single-phase PV inverters
One assigned to L1
One assigned to L2
One assigned to L3
All three SMA inverters are controlled by the Cerbo GX via SunSpec Modbus TCP using the new Dynamic PV Power Limiting feature.
The system works extremely well. I have 0 W export configured, and I can actually see the Cerbo sending different power limits to each SMA inverter depending on the load on each phase. During setup, each inverter is also assigned to its respective phase (L1/L2/L3), so it appears that the controller is performing per-phase calculations rather than simply applying one global power limit.
Planned expansion
I’d now like to add a Solis three-phase inverter, which also supports SunSpec Modbus TCP active power limiting.
The proposed system would therefore become:
L1 → SMA single-phase inverter
L2 → SMA single-phase inverter
L3 → SMA single-phase inverter
Solis three-phase inverter connected across all three phases
All four inverters would be controlled by the Cerbo GX using Dynamic PV Power Limiting over SunSpec.
My question
Can the current ESS Dynamic PV Power Limiting algorithm correctly manage this configuration while maintaining 0 W grid export?
More specifically:
Does the ESS controller support multiple controllable PV inverters contributing to the same phase?
Is a three-phase SunSpec inverter internally modelled as contributing to L1, L2 and L3 simultaneously, alongside the existing single-phase inverters?
Or is the control logic effectively designed around having only one controllable inverter per phase?
For example, if:
L1 is exporting,
L2 and L3 are importing,
the Solis is producing equally on all three phases,
will the controller intelligently reduce only the L1 SMA first, or will it also reduce the Solis (thereby unnecessarily reducing production on L2 and L3)?
I’m trying to understand whether the optimisation engine has a true per-phase model of all controllable PV sources, or whether this mixed single-phase/three-phase arrangement falls outside the intended design.
Has anyone implemented this exact configuration successfully, or can one of the Victron developers confirm whether this topology is fully supported?
Thanks!