Huge discrepancy between Solis data and meter reading

10-10-20

If you're seeing variation between ESB reported and inverter reported energy sensor while using Home Assistant, have a look at the integration method used by the HA sensor.

https://www.home-assistant.io/integrations/integration/#integration-method

In my view a generic CT clamp isn't a precision instrument and only has a loose physical-link to the current conductor, so there's that element of loss of precision. Secondly there's the update interval too, having that too far apart is going to reduce the precision of the calculation and that will lead to the integration calculation being significantly off.

Just as a test, I'll configure a set of different integrations and see what I get at the end of the day. I suspect that it won't be off by much as the sensor reports back very frequently, but a less chatty sensor would lead to a bigger difference.

# Integration calculation testing:

  - platform: integration
    source: sensor.shelly_em_esb_meter_channel_1_power
    name: test_shelly_em_LEFT_power_daily_energy
    method: left
    unit_time: d
    unit_prefix: k
    round: 3

  - platform: integration
    source: sensor.shelly_em_esb_meter_channel_1_power
    name: test_shelly_em_RIGHT_power_daily_energy
    method: right
    unit_time: d
    unit_prefix: k
    round: 3

  - platform: integration
    source: sensor.shelly_em_esb_meter_channel_1_power
    name: test_shelly_em_trapezoidal_power_daily_energy
    method: trapezoidal
    unit_time: d
    unit_prefix: k
    round: 3

# END Integration calculation testing