I have attempted to explain my predicament below. Apologies for the long story, but struggled to abbreviate and still ensure I cover everything I have seen thus far.
I have a 3ph system with 3 Multiplus II 48/5000, a Cerbo GX, three MPPT 250/100 and Pylontech US3000C battery system. I have also installed an EM24 grid meter that is connected via USB to the Cerbo GX.
I have installed approximately 20kW battery power and 13kWp between the three MPPT's.
The system seems to record incorrect data and this causes an erratic response. The EM24 seems accurate. I have a 6kW 3 phase element in the one geyser and can see when this comes on that the EM24 shows an accurate reflection of this power as can be seen in the picture below.
During the past hour the element was on while the MPPT's was charging the batteries. Midway though the time where the element was on, the system decided the batteries is sufficiently charged it can assist. As expected the grid will now reduce while the inverter is pushing back the power to the non-essential side on the AC Input of the inverters.
The concern is that AC Consumption (Total load calculated by the system) now spikes to almost 40 kW on L1 as well as on the other phases although little less. The grid meter is actually seeing that we feeding back into the grid on L1 and L2 although L3 remain stable (L3 now correctly replaced with the power from the batteries). Also the Multiplus AC Output power becomes extremely erratic, which I would believe is only the supply to the essential side and should not reflect the power drawn by the element.
Then as soon as the batteries go back to charging the system become stable again with the Grid meter showing the drawn power.
I have disabled the Grid feed-in (Both AC and DC) settings last night which seemed to have improved the situation, but it was still not stable this morning when the geyser switched on. Below is a graph from 3am to 6am
At 4am the geyser switched on and the noise started as the batteries still had capacity. Just before 5am the batteries got to minimum SOC (35%) and it started pulling from the grid. After noticing this I decided at 5am to change the Multiphase regulation back to all phases. This improved these peaks and dips seen on the both sets of graphs, but the issue still remain when the batteries are being utilized.
If anyone can give me some advice as to what to look at to improve stability during these discharging cycles of the batteries. Once I have this improved I would go back to Individualphase regulation to ensure I remain net positive on each phase as I have three single phase meters for my 3ph supply that I need to manage.