I suspect what’s happening is the energy integration is being performed on different input power sensors.
One is using the combined DC output of the PV arrays while the other the AC output from the inverter after conversion.
The DC values will naturally be higher than after conversion to AC. It’s only a hypothesis, I can’t say for certain this is what’s going on.
However, some information to explain why I think this is the case…
When I look through the sensors the Sigenergy2MQTT add-on provides there are two different outputs for Daily PV Energy:
sensor.sigen_0_daily_pv_energy
which yesterday reported 22.27 kWh:
and
sensor.sigen_0_inverter_1_daily_pv_energy
which reported 24.84 kWh
Meanwhile the mySigen app reports daily production as 22.27 kWh, which matches the sensor.sigen_0_daily_pv_energy sensor, while the other sensor is 11.5% higher.
That’s not just some sampling rate error.
Then when I look at the sensors for the individual PV arrays, the outputs yesterday were:
PV1: 10.10 kWh
PV2: 6.03 kWh
PV3: 8.71 kWh
which totals 24.84 kWh, the same as the sensor.sigen_0_inverter_1_daily_pv_energy sensor above.
So I think perhaps another look at the inputs being used to calculate daily and other energy accumulation values is worthwhile.
.
We consume and are billed on the AC energy output of the inverter, not the DC side, so it’s the AC energy data that matters most, to me at least.
The DC data is of course interesting and helpful for diagnostics, comparing strings etc but as inputs into Home Assistant Energy Dashboard and other tracking, then it’s the AC data I want to use.
How do I know this data is the right one to use?
Well I compared what the outputs for energy we export to the grid with that reported by my Utility’s billing meter (a Class 1 smart meter sampling the data in the multiples of milliseconds). IOW any error in PV output will propagate to export energy values. Overstate PV output and export will also be overstated.
The mySigen app data and the Sigenergy2MQTT outputs (and hence the HA Energy Dashboard using these outputs) matches my Utility’s billing meter inside of 0.5% which is not the case with the other integration.
.
I had “got used to” MQTT from previous integrations for my off-grid solar/battery system at a previous home and also my car integration uses it as well, so using it for this was not a stretch.