Unfortunately DESS has no control about that. If the scheduler sends a feedin restriction, DESS just tells the inverters / system, that it is not allowed to feedin.
And I think, that in this case the underlaying system prefers a “marginal grid pull”, as even a little feedin may be suspect to high fines in many countries, when a permanent feedin restriction is required.
I have a similar problem. Why limit PV output when the battery is nowhere near full?
I suspect the issue in my case is that the grid setpoint control is too slow. This causes power to be fed back into the grid briefly, triggering the Dynamic Power Limit. The Dynamic Power Limit and the grid setpoint control then end up fighting each other, resulting in the PV output being throttled further and further. After a while, the PV output recovers slightly (due to the Dynamic Power Limit). This problem only occurs during periods of negative prices. The issue is exacerbated when clouds drift over and PV output fluctuates sharply—likely due to the slow response of the grid setpoint control. Anyway, as far as I’m concerned, the Dynamic Power Limit shouldn’t need to operate at all while the battery isn’t full yet. (Perhaps for someone who has more PV power than the battery can charge…) Or the dynamic power limit / zero feed-in control needs to be set less strictly. For now, I’ve disabled the Dynamic Power Limit.
(v3.80-39 (stable version same issue), 3x mp8000, 80kWh, SolarEdge se17k)
Do you have a MultiRS or a VEBus System? (Multiplus / Quattro)?
Cause for VEBus, dynamic ESS is not using Grid-Setpoints for control. Also for the given situation (“negative feedin price”) the feedin Restriction is technically what is desired. A Grid-Setpoint is a “weak” constraint for the inverters that can be violated in favour for any other instruction present or just in the regular ESS-Case, when the battery can’t charge anymore.
I’ve send you a pn for your system ID, that is probably something to look into for our Inverter-Experts. Having a Feedin-Restriction (no matter if temporary or 24/7) should not lead to throttled solar, as long as the battery is able to absorb the available solar-power, i’d say.
The inverter is a solaredge connected via sunspec.
Imo feed in restriction is not desired when feedin is allowed but not profitable and a zero grid setpoint causes less grid usage and less costs.
When battery is full it’s still possible to regulate the pv inverter and maybe the scheduler should reserve 5% headroom for charging when the pv inverter suddenly delivers more power to keep the grid to 0 watt.
Anyway 300 watt grid draw is quite a lot and it hurts the profit.