Avoiding being overcharged by your smart meter by using AC line filters.

[uer166]

The current transformers should be resonable tolerant to a DC part. With some loads a DC component can happen. A relatively common example are hair dryers on the lower power setting that use a diode to reduce the power to half. AFAIK for hand held use this is allowed in many regions.

I would not expect the CT to saturate under some 100mA, as few materials get saturation under 1 A/m and 10 cm magentic path is still relative small. So it would be hard to make the CT saturate so early. Chances are they would prepare for some DC and intentionally use a material with current needed for saturation / leaving the good range.

None/few of the US meters are designed to tolerate DC, and in my experience you need very little current to saturate, single amps range. For MID metering in europe, they require to be tolerant to a half-wave rectified load of full current rating, and therefore you need to design CTs in weird ways. E.g. you usually have a dual (or even sometimes triple) core CT, where one material is high-permeability for high accuracy for normal loads, and a lower permeability to keep it working in case of 30A DC or whatever, but at slightly degraded accuracy.


Re: even harmonics: for a DC-tolerant CT, I think you're right, they can't measure DC but may be able to measure even harmonics. In general all these edge cases can corrupt, and break:

• RMS measurements of V and I
• Power factor measurements
• Reactive power calculation
• Frequency measurement (e.g. when it's done based on zero-crosses

Does that imply that the active power becomes inaccurate? No, it does not!

OP, I think you're really confused about how conducted EMI works, and how high-frequency stuff can affect metering. All that 10kHz or 100kHz or whatever, gets attenuated to nothing in the input filters, it simply doesn't meter it.

[coppice]

From the website you linked:

Here, the electricity being consumed no longer has a perfect waveform, instead it acquires an erratic pattern. The designers of modern energy meters have not made sufficient allowance for switching devices of this kind.

When they dismantled the energy meters tested, the researchers found that the ones associated with excessively high readings contained a ‘Rogowski Coil’ while those associated with excessively low readings contained a ‘Hall Sensor’. Frank Leferink (Professor of Electromagnetic Compatibility at the UT) points out that

As someone who designs and certifies MID meters, both 3-phase and 1-phase, the above smells of BS, the designers certainly have allowed sufficient allowance for high harmonic content. There are a few cases where a lot of the meters in the field read extremely low when presented with a half-wave rectified load, namely the current-transformer based ones. I don't see how anything can read 582% high unless you're trying to defeat the device in some non-real-world edge case. Also, hall and rogowski? Seriously? The author couldn't have chosen anything more exotic than that, household meters are CT or shunt-based.

I have found the actual paper scihub and going through it now, I suggest you do too to understand how the authors may have been mistaken (or not).

There are some major manufacturers of electrical energy meters who use hall sensors and rogowski coils, especially in Europe. CTs and resistive shunts are more common globally, but there is nothing strange or rare about a domestic tariff meter using either a hall sensor or a rogowski coil.

[uer166]

In that case I stand corrected. Do you know of specific manufacturers/part numbers? I remember doing the math concluding that a class B meter wasn't really possible to do with Hall.

[coppice]

In that case I stand corrected. Do you know of specific manufacturers/part numbers? I remember doing the math concluding that a class B meter wasn't really possible to do with Hall.

Its several years since I worked in metering. I would probably mix up who does what now. A small British company did a lot of work trying to get accurate results from Rogowski coils constructed as PCBs, while limiting their influence from external fields. Their approach has gone into some high volume meters, although the results we found when experimenting with their approach was quirky. One of the European makers used to use Hitachi hall sensors for three phase designs, which I think were only 2% accurate. Very cool designs, though. 100% of the electronics was on a single PCB, that clicked into place, with the hall sensors on it. When clicked in, the sensors were in just the right place above the high current terminals moulded into the case. That's the cleanest meter design I've seen. At least 2 people have their own ASICs, which have the hall probe and all the conditioning electronics within the ASIC. They claim extremely high accuracy. I think L&G are one of the companies that do this.

[Conrad Hoffman]

I tend to be skeptical of any non-rigorous measurement or discussion of power. There are a lot of pitfalls one can stumble into measuring power. Errors in measurement account for most claims of over-unity in free energy scams. I've been measuring low PF gadgets on the job for many years. We're well equipped and some high frequency measurements are still hard and prone to errors. Sometimes we admit that getting a super accurate result isn't worth the time and effort it would take and we just move on. I have a fair amount of faith in the people that design commercial power meters, as the technology is well established and the stakes are high.

[coppice]

I tend to be skeptical of any non-rigorous measurement or discussion of power. There are a lot of pitfalls one can stumble into measuring power. Errors in measurement account for most claims of over-unity in free energy scams. I've been measuring low PF gadgets on the job for many years. We're well equipped and some high frequency measurements are still hard and prone to errors. Sometimes we admit that getting a super accurate result isn't worth the time and effort it would take and we just move on. I have a fair amount of faith in the people that design commercial power meters, as the technology is well established and the stakes are high.

The standards for domestic energy meters focus on their accuracy from power factors of unity, where maximum accuracy is expected, to 0.5, where a relaxed accuracy is expected. Generally the approvals and calibration processes for meters do not check them below +-0.5PF, which I think is weird. Having a very relaxed spec seems appropriate, but not checking if the accuracy goes completely crazy doesn't. Its the same with small currents. Between the creep current (the minimum the meter is supposed to even register as power, so its readings don't accumulate small amounts of noise) and 5% of full scale, they don't generally even do a sanity check of the accuracy.

[Fflint]

Op here, I haven't gotten around to measuring my power use using CTs, because my use as charged did come down significantly after my meter was replaced when I installed my PV system. Previously I fairly routinely got bills between 600 in summer to over 1000  kWh in winter per 2 month billing period. After I installed my PV system (2.3kW only, projected annual production 2500kWh) my bills in the warm period (May-September) I got bills of few bucks per 2 months. So I basically used almost exactly the same as was generated. Now during winter my last 2 bills were on the level of summer use before. This is just anecdotal evidence, but it does seem significantly lower (by a third) than before.

Furthermore, the New meter gives me an hour by hour breakdown on use/generation and as far as I can tell it is accurate. I might actually make a refund claim based on that so the energy company sends the old meter for testing. (I hope they still have it).

This is just anecdotal "evidence". Anyone can say "maybe you're using less now that you have direct visibility of use", but I sure haven't made any extra effort to do that.

[Circlotron]

There are a few cases where a lot of the meters in the field read extremely low when presented with a half-wave rectified load, namely the current-transformer based ones.

Often wondered how accurate are old timey rotating disc watt-hour meters with a half wave rectified load.

[Zero999]

Dave did a video on the power wasted by smoke detectors, the apparent power of a nations worth of smoke detectors was a significant number in the MW.

That's not a problem, because smoke detectors are capacitve loads, which will be more than compensated for by the huge number of inductive loads, especially in industry, where motors are the main load.

[coppice]

There are a few cases where a lot of the meters in the field read extremely low when presented with a half-wave rectified load, namely the current-transformer based ones.

Often wondered how accurate are old timey rotating disc watt-hour meters with a half wave rectified load.

Why would half wave loads give trouble with CT based meters? They are all specified to use CTs with a pretty high DC tolerance. That spec usually comes from the meter's customer (the utility) rather than a standards body. Its there, none the less.

Most of the old disc (Ferraris wheel) meters work well with highly asymmetric load waveforms, unless the load is extremely high. Like the CT based meters, the core in most Ferraris wheel meters is specified to have considerable DC tolerance. I've seen some very lightweight meters, which are a bit suspect in this area, but most are OK.

[uer166]

They are all specified to use CTs with a pretty high DC tolerance.

That is not true of any US meters I've tested, and multiple MID EU certified meters that should have DC tolerance, but don't.

[AheadOfTheTimes]

I know there was some question in this thread as to whether the overbilling reported was real. Now a systematic study has been done by Euramet, which is the umbrella group for the national metrology institutes in Europe; more than half the models tested showed the problem, and the overbilling was as much as 20-fold on a water pump that was tested. It does seem to only happen with Rogowski coil sensors, that may be sensitive to the very high slew rates of pulse type loads. It looks like quite a mess for consumers who have these meters. The the full report, and references to supporting reports, here: https://www.euramet.org/research-innovation/search-research-projects/details/?tx_eurametctcp_project[project]=1543

[thm_w]

I know there was some question in this thread as to whether the overbilling reported was real. Now a systematic study has been done by Euramet, which is the umbrella group for the national metrology institutes in Europe; more than half the models tested showed the problem, and the overbilling was as much as 20-fold on a water pump that was tested. It does seem to only happen with Rogowski coil sensors, that may be sensitive to the very high slew rates of pulse type loads. It looks like quite a mess for consumers who have these meters. The the full report, and references to supporting reports, here: https://tinyurl.com/y2t63cvv

Here is the water pump link you are talking about, its from 2019: https://www.icrepq.com/icrepq19/205-19-have.pdf
4 of the 10 meters tested showed significantly incorrect readings.

"The power consumption related to these levels is between 23 and 130 Watt, respectively, specified by the manufacturer."
Worst case deviation was 2500% at 23W level. So would read 575W instead of 23W.

[AheadOfTheTimes]

PS: in terms of the original question, the old Isobar filters pretty good, with toroidal chokes, multiple stages. I don't think modern ones are like that. A brute force approach would be to plug in large capacitor filters such as sold by Greenwave or Stetzer (or have an electrician wire in a filter). The capacitor-only types are of course bad in terms of PF, and they draw reactive current all the time. You'd need to have two, one on each phase (talking about US power here).

[Siwastaja]

A pretty interesting case because this is exactly why we have EMI/EMC requirements and qualifications. Either the meter or the pump is clearly non-compliant and put illegally on the market, would be interesting to know which (or possibly both).

[tszaboo]

I know there was some question in this thread as to whether the overbilling reported was real. Now a systematic study has been done by Euramet, which is the umbrella group for the national metrology institutes in Europe; more than half the models tested showed the problem, and the overbilling was as much as 20-fold on a water pump that was tested. It does seem to only happen with Rogowski coil sensors, that may be sensitive to the very high slew rates of pulse type loads. It looks like quite a mess for consumers who have these meters. The the full report, and references to supporting reports, here: https://www.euramet.org/research-innovation/search-research-projects/details/?tx_eurametctcp_project[project]=1543

Well, I have a suggestion then.
At home batteries that you install, with custom inverter/charger.
Charge up the batteries with nice since wave, low harmonic content.
Discharge the battery and feed back to the grid with large harmonic content.
Free money.
Hopefully it forces them to change the electricity meters that are not guess-o-meters.

[Kleinstein]

A pretty interesting case because this is exactly why we have EMI/EMC requirements and qualifications. Either the meter or the pump is clearly non-compliant and put illegally on the market, would be interesting to know which (or possibly both).

The article shows some waveforms for the pump. These look really bad with high current pulses of one sign and what looks like some DC component.
The requirements for the power factor and harmonics may differ by region, but I would not consider this an acceptable current waveform for a consumer that is powered on for an extended time.
The setup shown has the meters in series and they thus measure not just the pump, but also the consumption of the other meters down the chain. With the odd current waveform with rel. high RMS current, but low power the losses in the meter can be higher than normal.

[coppice]

A pretty interesting case because this is exactly why we have EMI/EMC requirements and qualifications. Either the meter or the pump is clearly non-compliant and put illegally on the market, would be interesting to know which (or possibly both).

The article shows some waveforms for the pump. These look really bad with high current pulses of one sign and what looks like some DC component.
The requirements for the power factor and harmonics may differ by region, but I would not consider this an acceptable current waveform for a consumer that is powered on for an extended time.
The setup shown has the meters in series and they thus measure not just the pump, but also the consumption of the other meters down the chain. With the odd current waveform with rel. high RMS current, but low power the losses in the meter can be higher than normal.

If those current waveforms are accurate that is a very peculiar pump. I wouldn't expect even a reciprocating pump to give results like that. I wonder what the model is? The range seems fairly limited. There are 10 rate steps, but they only cover a range of 8000 l/h to 15000 l/h - less than a 2:1 ratio. That's quite a high flow rate, though. Several litres per second sounds like it must be for a large commercial aquarium.

[Siwastaja]

The pump waveforms are indeed absurd, looks like it's half-wave rectified, then pulsed. Even without knowing exact legal requirements, I'm quite confident it would be illegal almost anywhere. You can get by with strange loads, but only with very modest power levels. Tiny unidirectional (resulting in DC bias) short pulses approaching typical fuse ratings (thousands of VA); like those old "coffee maker half power switch using a diode" but in steroids; at least the half-wave rectification follows the sinusoidal shape.

[thm_w]

The setup shown has the meters in series and they thus measure not just the pump, but also the consumption of the other meters down the chain. With the odd current waveform with rel. high RMS current, but low power the losses in the meter can be higher than normal.

From the paper: "the energy meter reading of the static meters is corrected with the consumed energy of the following static meters"

If those current waveforms are accurate that is a very peculiar pump. I wouldn't expect even a reciprocating pump to give results like that. I wonder what the model is? The range seems fairly limited. There are 10 rate steps, but they only cover a range of 8000 l/h to 15000 l/h - less than a 2:1 ratio. That's quite a high flow rate, though. Several litres per second sounds like it must be for a large commercial aquarium.

Its some kind of fish pond pump, but I can't find the exact model. Similar to this style: https://www.alibaba.com/product-detail/BOYU-SPM-3500D-5500D-8000D-Aquarium_1600878847903.html?spm=a2700.7724857.0.0.127e1fddENkcCD
It seems more efficient than these ones, if it meets the rating.

[Kleinstein]

The setup shown has the meters in series and they thus measure not just the pump, but also the consumption of the other meters down the chain. With the odd current waveform with rel. high RMS current, but low power the losses in the meter can be higher than normal.

From the paper: "the energy meter reading of the static meters is corrected with the consumed energy of the following static meters"

There are 2 parts to the consumption, one is the supply current that is normally relatively constant and easy to correct. The other is the voltage drop at the shunt / current transformer and wiring. This part depends on the load current and with a current transformer can get even higher with the odd waveform. The distorted mains waveform from the nasty load can also effect the consumpton, up to the point that a MOV triggers from voltage spikes created by the current spikes. This current dependent loss is likely not corrected as it would be very difficult to do so.