AC Loads overview: separate Shelly Pro EM50 measurements and unexplained 25 W difference

DIY
, ,
1

Hello,

I have a Victron ESS system consisting of a MultiPlus-II 48/3000/35-32, Cerbo GX and SmartShunt.

The Cerbo GX is currently running Venus OS 3.80.

I am using a Shelly Pro EM50 to measure my AC loads. The two EM channels are used to measure two different load groups:

  • Essentials
  • Terrace

The measurements are available in the Cerbo GX and are shown under Loads → AC Loads.

For example, at the moment the display shows:

  • AC Loads total: 315 W
  • Essentials: 288 W
  • Terrace: 2 W

My questions are about how these values are presented and calculated.

1. Displaying the individual AC loads in the main overview

Would it be possible for the main AC Loads box in the Overview screen to display the individual Shelly measurements, in a similar way to how the Battery box displays voltage, current and power?

For example:

AC Loads
Essentials: 288 W
Terrace: 2 W
Total: 290 W

At present I have to enter Loads → AC Loads to see the individual measurements.

Since the Cerbo GX already receives the two measurements from the Shelly Pro EM50, would it be technically possible for the main AC Loads box to display these individual loads?

I understand that the MultiPlus-II itself does not independently measure AC-Out 1 and AC-Out 2. However, in my case these loads are being measured externally by the Shelly Pro EM50, so the question is whether the GX UI could use these measurements for the overview.

2. Difference between AC Loads total and the individual loads

There is also an apparent discrepancy which I would like to understand.

In the example shown in the attached screenshot:

  • AC Loads: 315 W
  • Essentials: 288 W
  • Terrace: 2 W

Therefore:

315 − (288 + 2) = 25 W

What does this additional 25 W represent?

Is the 315 W value calculated from the MultiPlus-II measurements, while the 288 W + 2 W values come from the Shelly measurements?

If so, what exactly is included in the 315 W value that is not included in the Shelly measurements?

3. Difference between Grid power and AC Loads

The same screenshot shows:

  • Grid: 339 W
  • AC Loads: 315 W

Therefore:

339 − 315 = 24 W

This is very close to the 25 W difference above.

Could someone explain what this approximately 24–25 W difference represents?

Could it be related to the self-consumption of the MultiPlus-II, or is it another calculated value in the GX energy-flow model?

4. Battery power

The battery is shown as:

  • 53.22 V
  • −0.3 A
  • −15 W

Could the approximately 15 W shown here be the DC power being consumed by the MultiPlus-II itself?

Or does this value include other system consumption/losses?

In other words, is there a specific GX/VE.Bus value that represents the actual self-consumption of the MultiPlus-II, separately from the AC loads?

I would mainly like to understand how the GX calculates these values and why the individual Shelly loads do not add up to the AC Loads total.

Thank you.

imagen
imagen972×881 60.4 KB

2

Are you aware about the Ammeters accuracy

image
image837×503 24.4 KB

Mutiplus 2% (5%)

I have a VM-3P75CT it change every 0.1 second

image
image252×236 8.81 KB

3

Thanks Fred.

I understand the Multiplus itself has a measurement tolerance of around 2%, but the Shelly Pro EM50 is specified by Shelly at ±1% accuracy, so I don’t think meter accuracy alone explains a 25 W gap on a ~315 W load.

This makes me think the 315 W “AC Loads” figure might actually be coming from the Multiplus’s own internal AC-out current measurement (used for Power Assist/transfer), rather than from a sum of the two Shelly CT channels.

Could someone confirm: is the AC Loads total in the Overview/VRM calculated from the Multiplus’s internal reading, and does it NOT take into account the Shelly’s CT measurements at all? If so, is there a way to make the GX use the Shelly channels (Essentials + Terrace) as the source for the AC Loads total instead of the Multiplus’s internal estimate?

Thanks again.

4

Just disconnect your load meters, what remains is the multiplus metering.

5

Thanks Nick. Without the two Shelly channels, the readings at a given moment are: grid consumption 409 W, battery charging with a consumption of -16 W (i.e. drawing 16 W from the grid), and the AC Loads indication showing 329 W.

As I understand it, total consumption should be: 16 W (battery) + 329 W (AC loads) + 64 W (Multiplus II’s own consumption) to add up to the 409 W of total grid consumption measured by the ET112. That’s my conclusion, since 16 + 329 = 345 W, which leaves 64 W unaccounted for out of the 409 W.

I don’t know if it’s possible to see that ~64 W figure anywhere, or whether the Multiplus reads it from an internal CT or simply calculates it. Thanks again.

6

IMO, you can ignore differences of up to 100 W in the IN/OUT calculations. Since most of the values are calculated, there’s also a considerable margin of measurement error.

My RS MPPT 450/200 often shows output power that’s 100 W higher than the sum of the MPPT channels.

I’ve observed similar discrepancies with my MultiPlus units in 3P-ESS. You can’t infer self-consumption from this any more accurately than by estimating it.

7

If the estimation or calculation seems to lack accuracy, I think it would be a good idea to use the actual grid and output readings when they are available, instead of relying on an estimate. Since we know the consumption of the loads (essential and non-essential) and the grid power, the difference must be the consumption of the Victron system. In that case, the error would simply be that of the measuring devices; however, knowing that it is a Victron estimation, we will assume that the indicated value is approximate (for example, with an error margin of up to 100W). Thanks.