I am trying to troubleshoot what appears to be a Dynamic ESS issue by negatives prices.
System configuration
3-phase grid connection
Victron MultiPlus 48/5000/70
Cerbo GX FW:3.75
AVM-3P75CT grid meter FW:1.11
ESS with Dynamic ESS enabled
Fronius 4.5 kW inverter (3-phase) connected on AC-in
SolarEdge 2.5 kW inverter (1-phase) connected on AC-in
Fronius is visible in VRM
SolarEdge production is not measured separately by Victron
Battery system operating normally Jk BMS
PV configuration
Fronius inverter is connected to East and West facing solar arrays.
SolarEdge inverter is connected to a West facing solar array.
Problem
When Dynamic ESS is disabled, the system behaves normally.
When Dynamic ESS is enabled:
Within a few minutes the Fronius output is reduced to approximately 29 W.
Battery charging almost stops.
In ESS settings, Feed-in limitation active changes to YES.
Battery SOC is around 44%.
Disabling Dynamic ESS immediately removes the limitation.
Fronius then returns to normal production levels.
Observations
With Dynamic ESS enabled:
Fronius production: approximately 29 W
Battery SOC: approximately 44%
Feed-in limitation active: YES
With Dynamic ESS disabled:
Feed-in limitation active: NO
Fronius production immediately returns to normal.
Additional observations
The issue appears only when both PV inverters are generating power at the same time.
In the early morning, when mainly the East-facing Fronius panels are producing, the problem is not obvious.
The problem typically starts later in the day when:
the West-facing strings on the Fronius begin producing significant power; and
the SolarEdge inverter is also producing
At that point Dynamic ESS limits the Fronius to approximately 29 W and battery charging almost stops. SolarEdge inverter is also producing and power is deliverd to the grid with negative prices so it cost money.
I also noticed that after resetting the solar forecast, the system appeared to behave normally again for a while. However, the issue immediately returned after re-enabling Dynamic ESS.
Questions
Could Dynamic ESS be making incorrect decisions because:
The SolarEdge inverter on AC-in is not visible as a separate PV source in VRM?
Dynamic ESS is only seeing the Fronius production and therefore calculating an incorrect PV forecast or energy balance?
There is a known issue with multiple AC-in PV inverters, where only one inverter is directly monitored?
Dynamic ESS is incorrectly activating feed-in limitation despite the battery still having capacity available?
Has anyone experienced a similar issue where Dynamic ESS actively curtails a Fronius inverter to around 29 W while the battery is only 75% charged?
Any suggestions regarding logs, diagnostics or settings to investigate would be greatly appreciated.
Dynamic ess is not actively controlling the acpv inverters. It just does instruct the system to “not feedin” during negative prices.
For general maths of dynamic ess to be correct, solar needs to be known to the system. This not only required for overall correct forecasting, but also for “per-second” setpoint calculations locally. (Solar Edge should be easy to add through ModBus TCP)
Are you running on individual phase mode? I think, then the issue is, that a one-phased Victron system could only absorb solar on phase 1. So, to eliminate Feedin on phase 2 and 3, throttling the 3 phased fronius is the only solution to obey the restriction (They cannot be throttled phase individual). With Multiphase balancing it should theoretically be able to keep the fronius producing upto it’s single inverter nominal power - but I don’t know, if the implementation would do that, and if that’s even possible based on the assymmetric load cobtraints of your grid code.
1)Our SolarEdge inverter is an old model that cannot be integrated into Victron (no Sunspec/Modbus TCP support). We also cannot replace it, because it still generates green‑energy certificates until 2031. So Dynamic ESS has no visibility of the SolarEdge PV production.
2)We are on a three‑phase grid, but only one MultiPlus is installed. The system measures the sum of all phases, but the MultiPlus can only absorb or inject power on phase L1.
3)Can you explain why Dynamic ESS is not charging the battery in this situation?
We are on a three‑phase grid, but only one MultiPlus is installed. Victron measures the sum of all three phases, and when I disable Dynamic ESS and use normal ESS, the system correctly brings the total power flow back to zero and the battery starts charging. So the system is clearly capable of correcting the total (sum‑based) power flow.
However, when Dynamic ESS is enabled, the battery does not charge, even though the total of the three phases shows export. In this case, Dynamic ESS does not try to bring the sum to zero, while ESS does.
Why does Dynamic ESS not perform the same sum‑correction that ESS does, even though the hardware is able to do it?
Then you could add the Solar Edge by adding a meter in front of it and connecting that to the gx with role “pvinverter”
Dynamic ess does not control anything directly. It just gives some advices to ESS basically.
I assume, you would see the same happening, if you just use ESS and disable feedin in the ESS Settings. I assume that’s an issue with having 3 phased ac pv, only 1 inverter and a feedin restriction.
When you set DESS to “My prices cannot be negative”, it will consider the lowest price 0 and not disable feedin iirc. (That at least would work around the symptoms)
Thanks for the explanation.
But this still doesn’t answer the core issue I’m seeing.
Why does Dynamic ESS not charge the battery when the total (sum of all three phases) shows export, while normal ESS does charge the battery in exactly the same situation?
Even with:
3‑phase AC‑PV
only 1 MultiPlus
feed‑in restriction enabled
ESS succeeds in bringing the total to zero and starts charging the battery.
So technically the system can correct the sum.
To rule out price logic, I also set the option “My prices cannot be negative” so that Dynamic ESS treats all negative prices as zero.
But even with the price fixed at 0, Dynamic ESS still does not attempt to charge the battery or correct the total power flow.
So the question is:
Why does Dynamic ESS not perform the same sum‑based correction that ESS does, even when the price is zero and the hardware is clearly capable of correcting the total?
As said, DESS is not calculating anything on that matter. Dynamics ESS just does “high level flow control”, it doesn’t do anything on basic control behaviour / balancing, it’s not even phase-aware on any value.
It may be possible, that the scheduler decides to idle the battery here and have it charged later. Then, ESS - along with a Feedin restriction - has no other choice than curtailing your acpv as there is no energy sink available (grid restricted, battery idle, consumption low)
What does the forecasted energy flows show for that hour(s)? Does it forecast 0 solar2battery?
Thanks for the explanation.
But I still see a clear inconsistency in how Dynamic ESS behaves compared to normal ESS.
Every day, the forecast shows that the battery should be charging during the hours with solar production. However, as soon as the price becomes negative, Dynamic ESS stops charging the battery — even though the forecast still shows solar‑to‑battery flow for those hours.
When I disable Dynamic ESS and use normal ESS, the system behaves correctly:
ESS charges the battery and brings the total (sum of all three phases) back to zero.
So the hardware clearly supports sum‑based correction.
To rule out price logic, I also enabled “My prices cannot be negative”, so Dynamic ESS treats negative prices as 0.
But even with the price fixed at zero, Dynamic ESS still stops charging and deviates from the forecast.
So my question is:
Why does Dynamic ESS stop charging the battery and ignore its own forecast whenever the price is negative (or forced to zero), while normal ESS continues to correct the total and charge the battery?
This behavior only happens with Dynamic ESS, not with ESS, and only during negative‑price periods.
If you could give me your system ID (numeric one in the vrm url), I can have a look, if DESS is applying a feedin restriction that may cause this. (Or if I can see something else)
I was under the impression it should not do that for exactly 0, just for negative prices - but if it works for non-0 prices, then that seems to be what is most likely causing this.
While DESS is active, you have ESS + feedin restriction active.
As long as the solar of the solaredge is “consumed away” and the ESS-control can see consumption, everything works as expected.
As soon as consumption drops bellow the solar edges production, ESS (0 consumption, no feedin allowed) decides to shutdown the fronius almost completly, as there is no need visible.
When you switch to disable DESS, you have ESS + NO feedin restriction - and it starts to feedin, and then increasing the battery chargerate.
So, how to proceed?:
I would disconnect the “invisible” solar edge inverter for a day or two and see if the issue then is resolved. If yes, get a compatible meter, put it in front of the Solar-Edge and add that with “Role pvinverter” to the gx.
That would btw. also fix your consumption values, as with the current (invisible) Solar, the system won’t be able to calculate a correct consumption for your home, as it is not aware of “how much solar” of the solaredge is directly consumed by loads.
I will have a chat with our ESS-Dev, if he has an Idea about this concrete situation. (And if it is something that could be resolved)
Another Test you could run would be to bring up consumption above the solar edges production, forcing ESS to unthrottle the fronius and then see if it starts to charge the battery again as well.
We performed the tests you suggested and observed the following:
Test 1: Increase load consumption
We first waited until the Fronius output had dropped back to approximately 29 W.
At that moment, we switched on additional loads in the house. As soon as the consumption increased, the Fronius inverter immediately started producing more power again.
When we switched those loads off again, causing the consumption to drop, the Fronius output gradually reduced and returned to approximately 29 W.
This behaviour was fully repeatable.
Test 2: Switch off the SolarEdge inverter
After the first test, while the Fronius was again limited to approximately 29 W, we switched off the SolarEdge inverter.
Immediately after disconnecting the SolarEdge inverter, the Fronius production increased again and started operating normally. Battery charging also resumed as expected.
We then switched the SolarEdge inverter back on and the issue returned. The Fronius output reduced again and eventually settled back at approximately 29 W.
Based on these tests, it appears that the presence of the SolarEdge inverter is directly related to the behaviour we are seeing from the Fronius inverter.
Please let us know if you would like us to perform any additional tests or provide further data.
