On the whole, the EMS has been fairly stable with inputs from IoTaWatt, Kauf plugs, and some powercalc entries. I’ve been monitoring 4 circuits with CTs (that I want to use elsewhere) until I complete my 5kW solar inverter/BMS HA integration, the latter of which I have been testing.
However, today, yesterday and the day before, as I’ve been testing by sending dummy value messages through MQTT from an RPi monitoring the inverter vendor’s website, huge kWh spikes in consumption and in grid export occurred in the hour I was testing each day respectively.
None of my input devices (IoTaWatt, Kauf plugs, Athom plugs, etc) show anything remotely like this, though the numbers could match up with dummy data I’ve been using as input.
Oddly, none of the entities I’m testing have any connection whatsoever to the EMS input entities, aside from being in the units of kWh.
Does the EMS pick up any entity change related to kWh? Is there some special considerations I should be undertaking to avoid corrupting my ‘real’ energy data?
And I should note that I have not experienced any of these spikes when I was not testing the inverter integration (just to entities with no bearing yet on any of the entities used to drive EMS).
For some reason, the energy entities are resetting at 2pm. Odd this hasn’t shown up as an issue before, as I’ve been running almost 3 weeks in EMS without this surfacing before. But sure enough, as I look through other days that appeared normal, the same reset was taking place.
Do I need to set a time value somewhere for it to reset at midnight in my timezone?
Or do I need to override or ignore these, and manually create a Riemann integration for each power entity explicitly?
Update: It turns out I already did a Riemann integration for the SolarOut. Looking at the Riemann integration entity for SolarOut, it is a accumulating entity without reset. So should I conclude that I should use a non-resetting energy entity for all EMS inputs?
