I have now since 2 days ESS running on my Victron; goal is to keep my net grid power around 0 (or in the settings Grid-setpoint = 50W).
But I noticed a behaviour I was not expecting, so I’m curious if this normal.
My system is the following:
My Victron Multiplus-II-GX is connected via AC-in to Phase 1 on my house.
I have a battery which has its own BMS. This BMS is not connected to the Victron.
On phase 2 and Phase 3 in my house I have each a different PV system.
I have a grid-monitor system that measures every 30 seconds the total power (phase 1 + 2 + 3); which is send (and received) by the Multiplus.
As soon as the sun starts shining (like this morning), the energy measured is negative and the battery is set to charging by the Victron, and indeed the battery gets charged.
And when the sun stops shining the battery discharges and deliveres to AC_in.
So far the good part.
Now my observation:
The victron charges and discharges in “discrete steps” of maximal power (set to 2000W). So I see a pulse-train of +2000W at the output (3 minutes on=2000W, 3 minutes off=0W) when the sun is shining. And a -2000W pulse when the sun is not shining.
Is the ESS by default a PWM-like signal or should the amplitude also be changing?
(or did I mess up somewhere in the setttings)
I solved it, so for those with the same issues here a short explanation:
There were two issues with my system at the time of the above measurements: Issue 1:
In my case the BMS and the Victron are not coupled. According to the BMS the battery was 70% charged, so no issue.
But the Victron converted was still at its default setting, that is 50V as sustain voltage. In the case above it tripped each time just the 50V sustain voltage, making the Victron converter to stop charging and discharging; hence the pulse pattern. Changing the sustain-voltage to a slightly lower voltage solved this issue. And yes I’m working to get my BMS coupled to the Victron to solve it better.
Issue 2:
The above does explain the pulse pattern but it did not explain why it was the maximal current each time. That issue was caused by another phenomena: the update frequency of the energy measurement. The sensor that passes the energy value to the Victron converter was too slow, hence the Victron converter ramped-up to the maximal voltage as it did not see any change in the measured energy.
This was easily detected after issue 1 was solved: the battery now switched continue from charging to discharging in the frequency of the energy measurement.
A faster energy measurement will solve this.
Conclusion: I need to implemented both fixes on my system.
