I have a Shelly energy meter connected to Venus OS on my Cerbo GX, configured to measure my EV charger consumption. Yesterday around 10-11 AM, the Shelly reported a massive power spike – roughly 219 kWh in a single hour – which is obviously incorrect.
The spike is visible in the VRM dashboard graph (see screenshot), distorting the y-axis and making normal data unreadable. However, the total Consumption figure at the bottom of the dashboard appears correct, as that is based on a separate grid meter, not the Shelly.
My setup:
Cerbo GX with Venus OS
Victron ESS with Multiplus
Shelly meter measuring EV charger (configured as AC load)
Separate grid meter for total consumption
My questions:
Is there a way to delete or correct the erroneous data point in VRM?
Has anyone experienced similar false readings from a Shelly meter integrated via dbus-shelly?
Any suggestions on how to prevent this from happening again?
There is no way to have data in VRM corrected unfortunately. It would require manual alignment on a lot of datapoints and therefore is not done “just because it looks ugly”.
That error happens most likely - you should be able to validate through vrm advanced - if the EnergyForward Counter of the shelly reports “0” for a split second, before it returns to it’s regular 248kWh value - then, the system will have detected that 0 as a reset of the counter and anticipate further values to be real deltas again.
I will forward this to the dev of dbus-shelly as well, maybe we can “ignore” 0 Counters, if they ever appear, as that usually always is “wrong”.
In Case of an intedet reset, it wouldn’t hurt as well, if the 0 is ignored and the reset then happens on 0.001 or sth.
You can’t compare the two sites as one is testing functionality that has not, or may not be pushed to production. That is what happens with betas. If you want to report issues for betas there are dedicated sections for this, we discourage the general sections being used for beta topics.
