Harmonic distortion causing high loss??

[Scat97]

I am busy investigating why my household has an unusually high electricity consumption. Over the past few weeks I have been manually measuring the consumption of many electrical devices in my house (such as the TV, computers, washing machines etc) in order to determine what could be using high amounts of energy. After accounting for all the big energy-consuming devices (and some rough estimates of lights etc), I still have an unaccounted 45kWh of electricity usage.

As I am not sure which devices could possibly still account for this usage, I am starting to think some electricity is leaking somewhere. However before I want to go over all the wiring in my house I wanted to check if perhaps the power quality in my house is bad perhaps accounting for a higher power draw than I am assuming.

In order to determine this I measured the AC voltage from the plug using a resistive voltage divider (and floating the oscilloscope) which showed some distortion (I think). I was wondering whether someone here might know if this distortion is significant and if it could be worth renting a proper harmonic power analyser to further analyse this problem...
Furthermore while doing this experiment, I found that occasionally there was a large voltage spike, but was not sure whether it was a quirk from my oscilloscope.

I look forward to getting some thoughts on this. And if someone has tips on further investigating unexplained electricity usages I am all ears!

 

[Avo Mk8]

Intriguing!
Is your unaccounted usage figure for a day, a week, a billing period!?

Are you using a scope probe on that resistive divider?
I imagine the artifacts you observe may be due to the set-up - the waveform doesn't look that horrendous to me! I have no knowledge of how meters deal with harmonic distortion, but I naively assume they would make a good fist of measuring the power you actually consume, within reason.
Others here will surely know.

If you have a clamp ammeter I would be inclined to monitor the feed and see if the (obviously varying) consumption tallies with your assumptions. Could there be a load you've not accounted for?

I think you would know if something was 'leaking' that much!

 

[Avo Mk8]

Below an academic paper looking at the accuracy of semiconductor/IC electricity meters, under 'distorted waveform' conditions. The ones examined seem to have reasonable performance to my mind.

The report does make passing reference to real world random performance of the occasional meter!

https://res.mdpi.com/d_attachment/energies/energies-14-00390/article_deploy/energies-14-00390.pdf

 

[Lucien Nunes]

What are you using to measure the energy consumption? Beware that many domestic energy monitors do not correctly measure non-unity power-factor loads, whereas suppliers' energy meters do. Therefore with a mixed load a discrepancy between the two is common. With certain low power-factor items such as electronic modules using capacitive droppers in white goods on standby, some energy monitors have been found to be over-read by a factor of 5x.

Leakage is often proposed as a reason for high bills but is rarely significant. Energy dissipated by leakage is manifest as heat and anything over a few watts usually makes itself known. 50W is the heat output of a large soldering iron which would rapidly burn out the insulation of a faulty junction box or cable, but would only add €0.25 per day to the bill if somehow it were able to persist.

Please do not float the scope ground, it's always bad practice. Not only due to the shock risk if the casing becomes live, but leakage from the scope AC power innards to the measurement ground can cause significant errors. I suspect this is part of the cause of the waveform you are seeing. There's no excuse - you have a differential mode available. Connect the two probes to two taps in the divider and set the inputs to invert and add.

Don't forget that impulse voltages can blow your divider to smithereens unless it's rated for Cat III or more. If you just put some resistors together, you did allow a margin on their limiting element voltage, I hope?

If that is in fact the waveform of your mains, yes it is a bit distorted but probably within the capability of the suppliers' meter to measure to rated accuracy. The impact on most domestic load efficiencies would be trivial, so I do not think that is a major player in your presumed excessive bills. At least, it is probably hidden inside the error brackets from your estimating process.

 

[pc1966]

I have seen cases where a non-RCD circuit has a fault to earth that is able to drain several amps but fails to trip the over-current protection (MCB/fuse) due to the modest earth resistance.

This sort of genuine "power missing" case is only likely to be to earth as if it were a fault L-N in a cable, etc, the power density involved would start a fire or cause the initial fault to escalate to a short.

You say "unaccounted 45kWh of electricity usage" but you do not indicate the billing period for this: per day, per month, per quarter, per annum?

45 kWh is roughly £9 and per year is an average of 5.13W which is easily the stand-by power of a satellite TV box, etc. Per month it is 61.6W which is harder to account for, and per day a more worrying 1.8kW and that will make your bill eye-watering expensive!

 

[Scat97]

The unaccounted 45kWh is per week. So that's a bit more than 6kWh per day, so I guess that's around 300W that I am losing which is quite significant.

For the measurement with the oscilloscope I was using probes on a voltage divider with a 10x reduction to stay in a safe region. But yeah I agree that it is a bit dodgy to measure the mains using a scope like this.

The reason why I was a bit sceptical about faults in the metering is that I can (vaguely) remember that there was some scandal a couple of years ago in the Netherlands where electricity companies were deliberately installing meters that measure apparent power. But I can't remember what ended up happening (or whether it was just a big conspiracy theory). Furthermore I can't imagine our power factor is too bad as we do not have a lot of especially inductive/capacitive appliences as far as I am aware...

For now I think I will look into measuring the power consumed for the different groups in the house (perhaps I'll invest in a clamp meter), perhaps I can narrow down a problematic group and focus my investigations there.
All the wiring done in the house is very dodgy (by the previous owners). I've found semi-live neutral wires in the past. Furthermore the neutral wire of the 3-phase wire going into a second distribution box melted at some point. And when it rains hard some circuit breakers randomly trip, perhaps also indicating a leakage somewhere...
I am afraid it might be time to stop ignoring the red flags and start rewiring certain sections of the house...

The weekly usage I have accounted for so-far is shown below, with the total chart amounting to approx 105kWh (per week). If someone has any tips on other big consumers I might have missed let me know . I measured the power consumption using one of those standard plugin meters.

 

[pc1966]

The unaccounted 45kWh is per week. So that's a bit more than 6kWh per day, so I guess that's around 300W that I am losing which is quite significant.

Yes, that would be observable on any moderate meter.

For the measurement with the oscilloscope I was using probes on a voltage divider with a 10x reduction to stay in a safe region. But yeah I agree that it is a bit dodgy to measure the mains using a scope like this.

You can get CAT-III probes but they are not so cheap:

Staubli SET Isoprobe Series 68.9559-23 Oscilloscope Probe, Passive Type, 500MHz, 1:100, BNC Connector | RS

uk.rs-online.com

All the wiring done in the house is very dodgy (by the previous owners). I've found semi-live neutral wires in the past. Furthermore the neutral wire of the 3-phase wire going into a second distribution box melted at some point. And when it rains hard some circuit breakers randomly trip, perhaps also indicating a leakage somewhere...
I am afraid it might be time to stop ignoring the red flags and start rewiring certain sections of the house...

First of all you really, REALLY should be getting that checked out professionally!

At the very least an global IR test (L+N to E at 500V) to see if there are any faults to earth, and more generally all of the circuits tested/inspected for problems. There may be some folk on this forum who are near to you can could give you a sound EICR (electrical installation condition report) as a starting point for any work that is needed. Quite likely they would also identify the problem of any "lost power" in the process. If you can post your town or post code (no further details) then someone might get in contact.

Background to EICR coding can be found in the Best Practice Guide #4 downloadable here:

Best Practice Guides | Electrical Safety First

Electrical Safety First produces, in association with other industry bodies, a range of Best Practice Guides which provide definitive information and guidance on a range of technical subjects.

www.electricalsafetyfirst.org.uk

 

[pc1966]

OK, just seen you are in the Netherlands so probably few folk here are local!

But the same basics apply, get some professional testing done.

 

[Avo Mk8]

If you procured a clip-on current transformer power monitor, you could monitor fixed wiring rather than individual appliances, eg

PZEM-022 Open and Close CT 100A AC Digital Display Power Monitor Meter Voltmeter Ammeter Frequency

Only US$15.13, buy best PZEM-022 Open and Close CT 100A AC Digital Display Power Monitor Meter Voltmeter Ammeter Frequency Current Voltage Factor Meter with Split CT sale online store at wholesale price.

uk.banggood.com

 

[Mike Johnson]

This thread came up in the side bar, and then I noticed the date, it was interesting so I read it all and was getting excited by what the OP had found, I suppose now we will never know.