Under certain circumstances, it seems advantageous to me if the grid operator has no way of curtailing a PV system.
To achieve this, a PV system to be installed this year would need to meet the following conditions:
Module capacity of up to 25 kWp
Feed-in limit of 60% of the module capacity (i.e. a maximum of 15 kW)
An east–west system would probably meet the 60% feed-in limit on its own fairly well, meaning that hardly any kWh would need to be curtailed.
With a south-east–south-west system, only a few kWh would probably need to be stored so that nothing went to waste.
How could the peak output be capped simply with a MultiPlus II system?
To do this, I would simply set the feed-in value for the MultiPlus II system in the ESS settings to around 100 watts below the maximum permitted feed-in value, assuming that the maximum permitted feed-in value had been set on the standalone grid-tie inverter.
Could this work?
To save the stored electricity for a period when electricity prices are high, it would be even better if the MultiPlus II system only stored energy and did not discharge during the period of maximum solar generation.
Venus OS is the control centre for Victron.
It can be configured as an ESS, which is the standard for grid-connected systems.
In an ESS, surplus PV is charged into the battery at around midday, for example, and shortfalls (at night) are compensated for from the battery, so that the power at the grid connection is ideally always 0 watts.
The free “DESS” extension calculates the optimum strategy based on the PV forecast, consumption forecast, electricity prices and system efficiency. In practice, however, there are still some non-optimal scenarios. Search for DESS.
Using Node-RED or MQTT, you can also tell
Venus OS directly what it should do and when.
I hope this helps as an introduction to the complex subject of optimal energy management.
The question is whether the grid operator accepts a customer-adjustable configuration.
Both the grid setpoint and the feed-in limit are customer parameters.
If the batteries are full, the PV output would have to be curtailed. At present, this is only possible with AC-coupled PV and officially only with Fronius PV.
DC-coupled PV is only curtailed when no grid is connected.
I think I need to describe it in a little more detail.
With simple, grid-feed-only inverters (i.e. no hybrid inverters and no battery), regardless of the manufacturer, the feed-in power can simply be limited to 60%. This means that everything is permanently configured and safe from the grid operator’s perspective.
If I now also install a MultiPlus II system in parallel and set it, for example, to charge the battery when the feed-in power reaches 59%, then only 59% will be fed into the grid until the battery is fully charged. Once the battery is full, the 60% feed-in limitation of the simple grid-tie inverters takes effect.
Is what I’m planning clearer now?
Up to this point, everything should work without any programming, using the ESS settings (Set Gridpoint).
You’d need to tell the PV inverter when the battery is full and then throttle it back to 60%.
With current models, I believe this should be possible via the digital input in accordance with Section 9 of the EEG, “Curtailment”.
You could connect a digital output on the Cerbo GX to the inverter and use Node-RED to control the Cerbo’s digital-output relay when the SoC reaches, say, 95–98%.
The Cerbo would then essentially act as a home energy management system (HEMS).
I see the Grid Set Point as independent of this. It cannot regulate the inverters’ output either.
An alternative would be to integrate the inverters via SunSpec (Modbus/IP), disable DC surplus feed-in and set a feed-in limit of 60% of Wp. The inverters would then need to be dynamically throttled.
Unfortunately, I can’t test this because I haven’t modified my OpenDTU yet.
So you want to charge the battery exclusively—or preferably—with the peaks above 60%. It has to be said, though, that reaching over 60% hardly ever happens. Besides, by the end of 2028 (with the emphasis on “should”), every system is supposed to be fitted with a control box. That would do away with the 60% limit, and curtailment would then remain the absolute exception, depending on the grid area.
This isn’t easy to configure using the official ESS parameters, but it is absolutely feasible. It probably wouldn’t make much sense, though.
I control my installations using the grid feed-in limit set to a maximum of 60%, and it works. As you say, the maximum charging current is ignored under DVCC. However, the maximum feed-in limit is observed, as it only takes effect once the battery is fully charged.
You can test this very easily. Enable the feed-in limit and set the value. This also works with DC.
No, standalone PV inverters operate completely independently of the battery’s SOC. They shift the MPPT and reduce the feed-in power to 60%. This works entirely without an additional Multiplus II system with a buffer battery.
Exactly. That way, not a single kWh would be wasted.
For east-west systems, the AI forecasts a 1.1% annual yield loss, and for south-facing systems, up to 9%.
The yield benefit would increase further if additional kWh could be buffered using the additional Multiplus II system—kWh that would not be remunerated because of negative wholesale electricity prices.
From today’s perspective, it looks as though curtailment will remain the absolute exception. Ideally, there is no control box, or retrofitting one is not feasible because of grandfathering provisions.
Can even the simplest PV inverters communicate with the control boxes planned for 2028 for all systems?
Yes, that sounds about right. 9% seems a lot to me, but 5–8% is realistic.
Yes, they should be able to, actually. Either via four contacts or EEBUS. However, this only applies to systems installed after 01/01/2025 (date not guaranteed).
A battery storage system in itself will probably be worthwhile. In theory, your plan could also be implemented with custom controls, for example via Node-RED on the GX. However, you would need to be able to predict these peaks in order to reserve some capacity for this rare situation. Since these high peaks only occur in spring and autumn, i.e. when yields are generally still somewhat lower, I can imagine that even without additional control, a large proportion of the peaks will end up being stored in the battery.
If you don’t want to limit the PV inverters to 60% across the board, but only when the battery is full, because the MP2s can then draw more power, you need to tell them when to limit themselves.
With current PV inverters, this is done via a two-wire contact. So you need to check whether your PV inverters have a two-wire interface for limiting.
If they do, you can connect them to the Cerbo and control the relay based on the state of charge (SOC).
This means the Cerbo takes over the role
of the HEMS, and you need neither general limiting nor direct control.
My microinverters don’t have a two-wire interface, so I can’t verify this.
BTW: on the MP2, Aux1+2 can also be configured as inputs for Paragraphs 9 and 14. It’s the same principle.
Unfortunately not. The Aux contacts are no good. EEG 9 limits zero feed-in exclusively at ACout (i.e. via the MultiPlus sensor) and exclusively for the MPPT; in my view, it’s hardly usable. 14a switches off the chargers on the MP2 and doesn’t actually impose a limit.