Power returned to the grid does not correspond to the actual returned power

Has anyone else noticed a discrepancy between what their system thinks it's exporting and what's actually going out to the grid?

Background: I've got a Victron setup on my garden office — MultiPlus-II, Cerbo GX, and I recently added a Carlo Gavazzi ET112 on the AC output to get better visibility of my solar yield. Since fitting it, the figures on VRM have gone a bit wonky.

Specifically, the "Grid feed-in" reading on the dashboard seems to be understating the actual returned power. My DNO smart meter is showing noticeably higher export figures than VRM is reporting. The gap isn't massive but it's consistent — roughly 8-12% out depending on time of day.

Before I fitted the ET112, VRM and the smart meter were broadly in agreement. So I'm fairly confident the meter is either:

• Positioned incorrectly (clamp orientation?)
• Configured on the wrong phase setting (even though this is a single-phase install)
• Creating some kind of double-counting issue with the existing grid meter input

I've checked the CT clamp direction — arrow is pointing the right way toward the grid — and I've set it up as Grid Meter rather than PV Meter in the Cerbo settings, which might actually be wrong now that I think about it.

Is there a definitive guide for where the ET112 should sit in a single-phase off-grid/hybrid setup when you also have SEG export metering? The Victron docs aren't totally clear on whether it should be classed as a grid meter or a PV meter in my config.

@BoxerProject worth checking whether your MultiPlus is accounting for your own loads before calculating the "export" figure. If you've got anything drawing power in the garden office — even standby devices — the inverter sees that consumption first and only the remainder goes to the grid.

The reading your DNO meter shows is what physically crosses the boundary, so that's the ground truth. Victron's portal figures can drift if the CT clamp on your grid meter isn't positioned correctly — mine was slightly off on the van setup until I repositioned it past the main fuse.

Also, are you on a smart meter? SMETS2 meters can show a slight lag/rounding on export readings compared to real-time Victron data. Probably not a huge discrepancy but enough to cause confusion when you're comparing figures side by side.

Good point from @CoveMick there. Another thing worth investigating is your grid meter versus what Victron is reporting — the MultiPlus measures AC output, but if you've got any loads between the inverter and the grid connection point, those won't be visible to the system and it'll overstate the export figure.

Also worth checking whether you have a Grid Meter (e.g. the Carlo Gavazzi EM24) configured in your setup. Without one, the MultiPlus is essentially estimating based on its own measurements rather than reading actual grid flow. Adding a proper grid meter made a noticeable difference to the accuracy of my export readings. What does your system topology look like — are your office loads on the AC output or AC input side?

Classic Victron moment — the system is technically correct, it's just using its own definition of "correct" 😄

Had something similar on my static caravan setup. Turned out the MultiPlus was merrily reporting export figures that bore little resemblance to reality because the CT clamp was positioned on the wrong side of a junction. Moved it about

@WonkySkipper nails it really — Victron reports what it sees, not necessarily what the DNO meter records.

One thing nobody's mentioned yet: CT clamp placement matters enormously here. If your clamp is on the wrong side of your consumer unit, or oriented incorrectly, you'll get skewed export figures consistently. Had exactly this on my boat setup when I first installed — the numbers were baffling until I realised the clamp was reading net household consumption rather than the grid connection point.

Also worth checking if you have a Grid Meter configured in VRM versus relying on the MultiPlus's internal calculations alone — they can diverge quite a bit without one. A Victron Energy Meter (the EM24 or similar) on the grid connection gives you a proper reference point to compare against.