question

@mvader (Victron Energy Staff) could you kindly pass this question on to the dev team as well? I suspect they’re the only ones who can answer exactly what the yield and consumption metrics actually count/don’t count. Thanks :)

Firstly both values are very rough estimates. They are concluded by looking at the 2 amp meters and the two voltmeters (which only sense under load furthering the error because the cant account for the resistance of the cables) which can all be out by a quite a bit and cant be zeroed either. It will than integrate over time and the error is carried and potentially exponentiated.Mathematically drastically changing values are a problem for the accuracy of integration. Since Solar irridiation is ever changing there is another error source

Secondly the beginning and end of "a day" is for lack of an internal clock estimated by pv voltage rise and drop typical of sunset and sunrise. So a day might not be exactly 24h long if you go from a foggy day to a perfect blue sky day but skewed towards one day or another. Not a problem for the long term but since you want precision it might just.

Yield is whatever amps are coming in at the Pv port times the PV volts and that over time. So everything coming in, going to the battery AND the loads (AND self consumption, ie heating up the device under load)

Consumption is amps at the load port times battery voltage over time. So basically whatever goes out that door is counted. It does not differentiate where it came from be it the PV directly or the battery.

Battery voltage is a rough indicator of SOC and requires to settle for some time of being unladen of loads or charges. You are likely getting wrong and low readings because the loads are pulling the voltage down if you dont disconnect them.

If you are interested in the SOC, get a BMV. It measures the amps more precise and has a second line to the battery to have the exact voltage. It can however only look at the energy going in and out of the battery, it wont give you any indication of the yield that is consumed directly. There is no Victron device to measure yield more precisely. You could put a second BMV or similar there but the correct voltage on the panels is a challenge. The integration of the time you would have to do have to do yourself then externally. A lot of trouble for not much gain I should think.

Personally I go with the kWh overall on a monthly basis thats good enough for me.