ESS parallel to 3-phase delta grid (no neutral) with Dyness PowerBrick Plus — feasibility check

Hi everyone,

I’m planning an ESS installation and would love some feedback from people with similar setups before purchasing.

SITUATION

  • PV system: 10 kWp with SolarEdge SE6000H single-phase inverter + optimisers, on Net Metering (SSP) scheme
  • Grid: 3-phase 220V without neutral (delta configuration) — will be upgraded to 380V 3-phase with neutral in the future
  • Consumption: >12,000 kWh/year, all-electric house with on/off water-to-water heat pump ~7 kW, home during the day

PLANNED CONFIGURATION
AC-coupled ESS in parallel with the grid (no backup required, bill savings only):

  • Victron MultiPlus-II 48/5000 (non-GX)
  • Victron Cerbo GX MK2
  • Dyness PowerBrick Plus 16 kWh (1 module to start, expandable to 2)
  • Victron EM24 Ethernet near the utility meter with CT clamps on all three phases
  • Victron Lynx Distributor
  • Victron CAN cable Type B (crossed pinout PIN4/5 → PIN7/8)

QUESTIONS

  1. THREE-PHASE COMPENSATION WITHOUT NEUTRAL — With a 220V phase-to-phase grid and the MultiPlus connected across two phases, can the Cerbo GX correctly perform three-phase compensation reading the EM24? Has anyone dealt with Italian delta grid configurations?

  2. DYNESS POWERBRICK PLUS + VICTRON — The Plus model (16 kWh, IP65) is newer than the more common Dyness models. Has anyone already integrated it with a Cerbo GX? Is the CAN protocol the same as previous Dyness models?

  3. EM24 WITHOUT NEUTRAL — Can anyone confirm that the EM24 Ethernet works correctly in 3P configuration without neutral, measuring phase-to-phase?

Thanks in advance to anyone willing to share their experience.

Don’t waste your time on the solaredge. Get something that works and has real support.
Go look at the SE issues topic in the QA section.
As for the topology, suggest you read the docs.

A simple search will provide the answers:

Hi Nick,

thank you for the quick response and for pointing me to the docs.

I’ve read through the ESS documentation and I understand that a 3-phase Victron configuration (3x MultiPlus in parallel) is not supported on a delta grid — that’s clear.

However, my setup is different: I’m planning to use a single MultiPlus-II 48/5000 connected phase-to-phase on two of the three delta phases — not a 3-phase Victron configuration. The Cerbo GX would read all three phases via the EM24 Ethernet meter for three-phase compensation, but the MultiPlus itself would only inject/absorb on the single phase pair it’s connected to.

My specific questions are:

  1. Is a single MultiPlus-II used for ESS (parallel to grid, no backup) on a 3-phase delta grid (220V phase-to-phase, no neutral) a supported and stable configuration?

  2. Does the Cerbo GX + EM24 three-phase compensation work correctly in this topology, even if the MultiPlus only acts on one phase pair?

  3. With ground relay disabled in VEConfigure, is a single MultiPlus-II in ESS mode installed by wiring l1 and l2 on L and N of the inverter,on a 3-phase delta grid (no neutral, phase-to-phase connection) considered stable and functional?"

Regarding the SolarEdge — noted, I’ll have a look at the SE issues topic.

Thanks again for your time.

I did Further research on the forum and someone seemed to make it work but i have more questions i cannot just get good info about:

Grid meter: planning to use Carlo Gavazzi EM530DINAV53XS1X (confirmed compatible with Victron on delta grid via RS485).

— Ground relay: understood it must be disabled in VEConfigure. Question is whether ESS grid test still passes.

Has anyone successfully run Victron MultiPlus-II in ESS mode on a 3-phase delta grid (no neutral) using a single-phase isolation transformer (220V/220V, 5-7 kVA) to create an artificial neutral? Specifically: does the ESS grid test pass with the neutral created this way?

The issue is not the missing neutral, the issue is that the grid measurment will be wrong. Adding a 230-to-230V isolation transformer does not change that.

In order to correctly use a victron system, single phase or not, in a threephase delta grid, you would need to use a threephase delta-to-star transformer for the whole house, and then use its output as the grid measurment.

Theres been a few threads lately about this. In theory you can correct the wrong measurment through nodered, and use a single MP for example on L1 (connected to L on ACin) and L2 (connected to N on ACin), however the resulting load calculations would be wrong in VRM

Thanks for the clarification. Regarding the single MultiPlus on L1-L2 with Node-RED correction — has anyone actually implemented this in ESS mode with a Dyness battery and confirmed it works stably? What specifically is wrong in VRM — just the display or also the actual charge/discharge control?

The power control (charge/discharge) is wrong because the grid measurments are wrong. Victron assumes a star grid but is being fed delta values.

I have not tried it myself since i dont have a delta grid. But heres a thread that holds a lot of information about this, and maybe the person that asked about it has since tested this