I have seen some strange behaviour in a system I have.
Multiplus-II GX 5000VA, EM540 as grid meter, ET112 as PV meter.
I have a node-red algorithm that gives a full charge once a week using a 2500W gridsetpoint. The idea is to set the setpoint to 2500W and charge the battery with the excess power that the house is not consuming. Because this is overnight, the battery generally gets charged with 2200W.
Now, every 1 in 5 charges I get the following behaviour. Once the setpoint goes to 2500W, the loads increase to 1800W at exactly the same time. Meaning the battery only gets charged with the remaining 700W. I’m not physically at the site but to me it seems the meter values are wrong.
The AC-load graph goes up exactly at the same time as the gridsetpoint goes up, and goes down at the same moment.
It’s virtually impossible for the EM540 meter to have such a false reading. Meanwhile, it’s also virtually impossible for the house to have such a -square wave- sort of load during the time I want to charge the battery.
Does anybody have experience with a similar issue?
To me, it seems like the BMS is not accurately reporting the actual charge current.
Then, Power is pulled from the grid and fed into the battery - but since the bms reports a way lower value, the only conclusion the system can make is, that the difference of the power (grid in vs bms reported power) is consumption on ac l1.
So, as long as the battery is charging with 2000 watts, but reporting 800W only, you’ll see 1200 watt ac-l1 consumption, matching the square shape you are seeing.
Could you check, what the multiplus’ DC-Power showed at that time?
But also think carefully about if that could be a real consumption. My father was also hunting a “ghost consumer”, and after hours of research it turned out, that his air-ventilation system has a electric pre-heater that kicks in, when outside air temperature drops bellow 0° and then is running for 60 Minutes due to hysterisis settings of that system.
Good question about checking the reported DC power of the multiplus.
The voltage and current reported by the BMS and the Multiplus DC system are (nearly) exactly the same.
I use the Seplos V3 BMS.
The scale is not the same but this is the comparison. The issue starts at around 15:00.
