[BUG] unnecessary energy shift via DC-bus

Hello,

I’m using three Multi RS in three-phase configuration (german grid code).

Multi RS FW: v1.28

Venus OS: v3.70

Unfortunately I’m still experiencing a misbehaviour when exporting to the grid with excess solar. This behaviour has been persistent since installation of the system one year ago.

As you can see from the screenshot, Phase 1 has a AC load of 4936W (internal load plus export), while it only produces 4297W of solar. The difference is coming from the other two inverters via the DC-bus.

I don’t see a reason why Phase 1 has to receive any power from the other Invertes via DC. This behaviour creates unnecessary losses and also leads to reduction of the solar output in extreme cases when the Inverter is already at its power limit but still receives extra power via DC.

DC bus power sharing has always been a feature of victron systems. Not just in the RS range.
So while it bugs you, it is not a bug but design. It has many advantages over other inverter brands that have individual battery banks on each phase who cannot share power.

If you wanted pure ac feedback, then the multi RS is perfectly capable of that. Just configure it as such and remove the DC side.

1 Like

Sorry, but I disagree. There are strong benefits of power sharing via DC, but these only apply for offgrid situations.

When connected to the grid, the system should have efficiency-optimized operation, which means to keep solar production and AC export within one inverter and phase.

The system could immediately switch back to DC power sharing as soon as the grid fails.

Sorry, but I have to disagree with this. If you have unbalanced loads as is virtually always the case, it is more efficient to be able to share power between phases at dc, particularly if a part of your system is dc coupled.

1 Like

I still didn’t see any technical explanation why sharing power between phases on DC is beneficial for efficiency.

When keeping the power within one inverter, the PV voltage only has to be converted to AC. When power sharing, the PV voltage has to be converted to the DC-Bus voltage first (50V), then moved to the other phase, then converted up to the DC-Link voltage and only then it’s converted to AC again.

Besides the inefficiencies, as I have mentioned before, the power sharing leads to additional de-rating at higher loads, I’ve already described this in the past in this thread:

Please define what “efficiency” means to you. For most users, it probably means “higher self-consumption.” For others, it might mean “higher utilization of the generated solar power, whether for grid feed-in or self-consumption.” These different definitions require different regulations.

2 Likes

Efficiency is output power divided by input power, so I mean higher utilization of the generated solar power.

The self consumption is always 100% as long as there’s excess solar and/or the battery has remaining charge. There’s no compromise in self-consumption when the power sharing is disabled (except when not having a grid meter that sums across all phases).

I wonder why my point is not understood well. Three MP2 combined with a AC coupled PV inverter behave more efficiently, that’s what I wish for three Multi RS Solar aswell.

The MP2 system doesn’t share power across DC, but loads the MP2s equally with the calculated excess across all three phases.

So additional requirements for the Multi RS system would be:

  • option to configure the grid meter as “sum of all phases” just like in a MP2 ESS system
  • don’t share power over DC when the meter is configured as sum of all phases and the grid connection is active

No, not necessarily. It always depends on what you want. And so for many people ist much more efficient to use 1 kWh solar energy by themself than to feed 3 kWh solar energy to the grid.

1 Like

Depends if your PV is AC or DC coupled.
DC coupled PV is converted from the PV voltage direct to the battery voltage, and is then available to provide power to any phase inverter.
If you are AC coupled, with individual systems, then it is impossible to move excess power from one phase to another, leading to export on one phase (if permitted) and import on one or two other phases. This may incur a metering loss as well if the utility has low rate for export and high rate for import.
If you have AC coupled PV with a common DC source, then it is possible to transfer power from one phase to another, as the battery inverter gore into the charging quadrant, this will transfer energy from Ac to DC, which is then available to another phase to convert back into AC. This is a little less efficient than DC charging, but a lot more efficient than not being able to do it at all. typical conversion loss for each step would be 95 - 98%

Not sure what you are referring to here, or if it’s relevant to Victron systems.

Looking at the description for the Multi-RS solar, it looks like the PV side of this is behaving as AC coupled Solar, but with a 3 phase system, they can get efficient power sharing by having a common battery.

The Multi RS Solar is a hybrid pv/battery charger/inverter. The PV is DC-coupled, but not in a classic way.

Look at the architecture:

The PV power from the MPPTs can be directly converted to AC. But when shifting the power to another phase, it has to go through the Battery DC/DC twice.

The voltage between the MPPTs, DC/DC and DC/AC converters is the DC-Link voltage.

To avoid this remove the battery bank and have ac pv mode only configured on the multi rs. It is a feature.
And if efficiency from that perspective is what you want that is the only way to achieve it with the RS.
What has been done is a system has been built a system that does not fit your desired operation.

Is there no desire at all to improve the system?? Sure seems like it.

All it needs is a software update to improve the behavior. The basic functionality is suitable for my desired operation, that’s why I bought and designed it that way.

All I’m getting is gaslighting and excuses from people who seemingly don’t even understand the problem. It’s very frustrating…

I think maybe too much is being read into the answers here.

What you are seeing misbehaving

Is part of the system design as the behaviour hasn’t changed through firmware updates.

Would you prefer uneven feedback due to the loads on the one phase using more power?

Do you have essential loads on the output of the inverter?

yes! there’s no benefit of even distribution of feedback across the phases.

No, the grid and loads are all connected on AC-in.

Edit: I also have another near identical setup where the loads are connected on AC-out (essential loads). The behavior is the same.

Hi @laserranger,

You are correct that this a known issue and a feature that could benefit from improvement.

The RS R&D team is aware, and have plans to address, but there is no ETA at this stage.

1 Like

Hi Guy,

Thank you for the feedback! I’ll wait for a future update then :slight_smile: