Accuracy of clamp-on metering when PV inverter present

Submitted by Guest on Fri, 02/03/2012 - 14:00

I was intrigued to find this site and to see that you're using clamp-on meters to measure consumption in PV installations.

I have a couple of these which I have used for some time to measure my household consumption, but in the presence of my newly installed Fronius inverter they seem to be horribly inaccurate.

As soon as the inverter is connected to the domestic supply, before it starts to generate, the reading on the clamp on meter (monitoring flow through my supplier's meter) jumps up by 180-200W. My first though was that the Fronius was actually consuming power. However careful monitoring of the electricity supplier's meter established that it wasn't.

Once the PV array started generating I monitored the readings. At the point where the sun just managed to balance the household need (i.e. the electricity supplier's meter just stopped rotating its disc), the clamp-on meter said there was still 300W flowing through that meter.

Tests with the inverter isolated indicate that the clamp on meters are usually pretty accurate, but as soon as the inverter is connected on the AC side the readings go to pot.

Has anyone else observed this? Is it an issue with the shape of the wave form generated by the inverter?

TIA,

John

Re: Accuracy of clamp-on metering when PV inverter present

Submitted by TrystanLea on Fri, 02/03/2012 - 16:50.

Hello John, It may well be a waveform issue, can you measure voltage as well as current to get a real power, apparent power and power factor reading?

We found that the power factor can get quite low when the power output is below around 15% of array peak capacity. The real power reading may be 133 Watts and apparent 220W for example (on a 2.9kW array). However above 15% the real and apparent power readings are usually the same.

We've actually just had a pv system installed here last couple of days and have the monitoring on it, take a look: http://vis.openenergymonitor.org/pages/solar.php see how the apparent power diverges from the real power reading at low power.

Trystan

Re: Accuracy of clamp-on metering when PV inverter present

Submitted by S Bateman on Sat, 03/03/2012 - 07:57.

Hi John,

I've got a similar set-up with a SMA inverter and have been amazed at how accurate the readings are. I know that clamp meters are not terribly accurate as we use them at work to measure laser input power (150A) and it often can vary depending on where you put the clamp. The LEM type seem to be a lot better but you have the break the cable to fit them, anyway I digress.

Compared to the internal logging on the SMA inverter the solar output (actual) as measured by the EMONTX is within 2% of that recorded by the inverter which was pretty amazing really.

I calibrated my system using a Fluke DVM in series with a 4 way extension lead (and a RCD) and adjusted the calibration factors to get the current seen by the CT to the same as that measured by the Fluke. For load I used a couple of hairdryers as I could vary the power demand by changing the power setting on the hair drier.

Re: Accuracy of clamp-on metering when PV inverter present

Submitted by smp2012 (not verified) on Sun, 04/03/2012 - 11:37.

This is a general query, arising from SBateman's reference to LEM clamps (thanks very much, by the way, for the charts).

Npower insisted on installing a digital 'like-for-like' meter beside my generation meter after I went solar PV (3.9 kWp) in the summer using an Mastervolt Sunmaster Inverter. The changed meter wiring looks as though I may not be able to use a clutch-clamp for monitoring, leaving me wondering whether, if I have to use LEMs on either meter, breaking into cable(s) might invalidate guarantee(s).

Advice really welcome but no rush! I am definitely at base camp. Cheers smp2012

Archived Forum