Why am I importing power when running from batteries?

[dangermouse]

Got my smart meter fitted yesterday so I've been testing the inverter and batteries I installed recently. Note I have no PV connected yet, so I'm just testing running stuff from battery power. But I've seen some odd discrepancies in the figures that confuse me.

The set up is a Sofar HYD3600ES with 2 Pylontech US2000C batteries.

I set up a reasonably constant house load of 220W (with freezer etc turned off so nothing would kick in half way through the test).

With the inverter in "passive" mode (ie doing nothing), the smart meter shows 220W import, and the inverter shows the same. As expected.

I put the inverter in "self use" mode (ie run the house from batteries). The inverter now shows 0.00kW import, 0.16kW from batteries and 0.16kW to the load. The smart meter shows 35W import. These figures don't add up to 220W.

I then put the inverter into force discharge mode, and set it to export 500W.

Now the inverter shows 0.16kW going to the load (as before), 0.28kW export and 0.43kW from the batteries. Smart meter shows 240W export.

Main question is - why does the smart meter show an import when the inverter should be supplying the entire load, in self use mode?

Also, why does the inverter measure the load correctly when running from the grid (0.22kW), but show a much lower figure (0.16kw) when running from batteries?

[Countrypaul]

What was the 200W load? If it was not pure resistance then cold power factor be affecting things? Do both meters measure loads with a power factor less than one accurately?

Can you try putting a 500W or 1KW resistive load on and see whather the figures are closer to what you expect?

[nowty]

So passive mode is actually a special mode in order to control via a third party device.

Most inverters cannot self calibrate very accurately with their CT clamps so quite often there is a constant small import. Your solar should have an offset capability, I know the Solar ME3000SP does using the later firmware versions.

[dangermouse]

What was the 200W load? If it was not pure resistance then cold power factor be affecting things? Do both meters measure loads with a power factor less than one accurately?

Can you try putting a 500W or 1KW resistive load on and see whather the figures are closer to what you expect?

I thought of this just after I posted. The original test load was mostly inductive (TV, xbox etc) and the inverter showed a power factor of 70%.

Turned all that off and set up a mostly resistive load of 260W (old fashioned light bulbs) power factor 95%. The results were pretty much the same, so I don't think that's it.

[dangermouse]

So passive mode is actually a special mode in order to control via a third party device.

Yes, but it's also a quick way of forcing the inverter to do nothing, ie the batteries go into standby mode and neither charge nor discharge.

Most inverters cannot self calibrate very accurately with their CT clamps so quite often there is a constant small import. Your solar should have an offset capability, I know the Solar ME3000SP does using the later firmware versions.

This would be the "CT calibration" menu? I played with that but whatever figure I entered didn't seem to affect the readouts on the inverter. But maybe it would affect the power the inverter pushes out when exporting ... I never tried altering the calibration with the inverter actually operating.

Thanks both for your suggestions, I'll have another play and report back.

[dangermouse]

I think Nowty has cracked it!

So re-running the first experiment with a load of 200W. Inverter in self use mode, smart meter shows 33W import.

I set the "CT calibration" to -30W.

Now the smart meter shows 3W import!

However this doesn't affect the figures shows on the inverter LCD display, they are still "wrong". But what was mainly bothering me was importing when I shouldn't be. I realise that a few watts discrepancy is unavoidable, but when it was enough for the smart meter to say "this is costing you 1p per hour" that was too much!

So, the conclusion seems to be, the value set in "CT calibration" is subtracted from the inverters power output. So a negative value increases the export power, and vice versa.

Thanks chaps.

[dangermouse]

Just to follow up on this ...

After a few weeks of observation, with and without PV connected to the Sofar, I think I can conclude that the inverter (HYD 3600-ES) is not very good at balancing import/export power when supplying very small loads.

If you're pulling around a kW, supplied from batteries, the smart meter pretty consistently shows a big fat zero.

But if the load is only say 100W, also from the batteries, the smart meter shows the grid power fluctuates anywhere up to 30W in or out.

When I connected my PV, I observed that the inverter was even less good at balancing small loads, especially when the panels are not generating anything.

However, on a day when I can run totally from a combination of PV and battery power, I seem to only be importing in the order of 200Wh per day, which I can live with.

[Stinsy]

Just to follow up on this ...

After a few weeks of observation, with and without PV connected to the Sofar, I think I can conclude that the inverter (HYD 3600-ES) is not very good at balancing import/export power when supplying very small loads.

If you're pulling around a kW, supplied from batteries, the smart meter pretty consistently shows a big fat zero.

But if the load is only say 100W, also from the batteries, the smart meter shows the grid power fluctuates anywhere up to 30W in or out.

When I connected my PV, I observed that the inverter was even less good at balancing small loads, especially when the panels are not generating anything.

However, on a day when I can run totally from a combination of PV and battery power, I seem to only be importing in the order of 200Wh per day, which I can live with.

Have you looked at the 30-min readings on your bill?

You'll probably find that all these fluctuations add up to a few pence a day.

Don't worry about it.

[dangermouse]

Have you looked at the 30-min readings on your bill?

You'll probably find that all these fluctuations add up to a few pence a day.

Don't worry about it.

Indeed, 0.2kWh at 40p a unit is 8p - as you say I am not worrying.

It's more just a desire to understand how the system operates.