How To Trick Your Electrical Meter By Saving Power

A group of Dutch scientists have been testing out some of today’s “smart” electrical meters to check their accuracy, among other things. Not ones to disappoint, the scientists have found consistently false readings that in some cases are 582% higher than actual energy consumption.

With experiments lasting for six months, the researchers tried to focus on meters representative of those commonly used in the Netherlands and manufactured between 2004 and 2014. Moreover, the researchers tried to reproduce standard household energy consumption patters rather than focusing on stress tests.

Their results? Well, “results varied wildly, with some meters reporting errors way above their disclosed range, going from -32% to +582%. Tests with uncommon results were repeated several times and the results were within a few percents of the original.” Moreover, “The greatest inaccuracies were seen when researchers combined dimmers with energy saving light bulbs and LED bulbs.” Not constrained to energy saving light bulbs, the inaccuracies are, ironically, tied to devices with integrated energy saving features. (Certainly makes us want to keep a close eye on our electric meters.)

“The reason for faulty readings appears to be the current sensor, and the associated circuitry,” said researchers. “The experimental results […] show that static energy meters can be pushed into faulty reading (positive and negative) if sufficiently fast pulsed currents are drawn by the consumer”

It is worth noting that there is contradictory research published by “the European voice of the providers of smart energy solutions” that maintains that “there is no reason to question smart metering technology”. Still, we wouldn’t blame you if you wanted a second opinion.

Thanks [acs] for sending this in!

95 thoughts on “How To Trick Your Electrical Meter By Saving Power

      1. I did too, and have already started gathering information on the types of meters, their measurement methodology, and the testing done prior to selecting these new meters that was done by our public utility.

        I tell you, I have yet to see an organization start STALLING that quickly in my life. They know, and they just realized we know… and they’re scared.

        1. Current sensors can go really high now (100A for an $8 sensor). It should be nothing to mirror the meter measurements using an microcontroller as a data-logger and determine at least if the same scenario is occurring here in the USA.

          1. And if you don’t want to put DIY equipment near your power lines (even if using relatively safe CT sensors rather than inline readings), or merely have more money than time, there’s several options you can buy online that use CT sensors and provide various web or otherwise interfaces for the data, most even support measuring individual circuits in addition to the mains coming in.

          2. That’s exactly what the meters are doing, and the problems result because the system isn’t fast/sophisticated enough to handle the spiky loads.

            What makes you think an $10 homebrew meter will be more accurate than your house meter?

          3. One of the problems with current sensors used in meters is that they are cheap and doesn’t have the frequency response to handle some high-frequency loads. The frequency response on CT sensors are usually less than the 8th harmonic (< ~400-450 Hz) which means anything above that are lost.

            I'm not that surprised that new digital meters reports the wrong energy consumption, just imagine all those switched PSU's introducing harmonics and uneven power draw which can't really be measured correctly unless you sample both the voltage and the amps accurately up to a couple of kilohertz

            I remember one case where the meters didn't take into account out of phase current due to the loads having high impedance/capacitance (ie. electric motors/florescent lighting) since they only measured the voltage and the current draw.

            So, you can quite easily build your own powermeter but you need to take into consideration how you measure both the voltage and especially the amperage – any off the shelf CT coil doesn't cut it.

          4. >”What makes you think an $10 homebrew meter will be more accurate than your house meter?”

            Because you can attach it to a $1000 computer to analyze the signals it gives.

          5. The micro is the problem – it’s ADC is not fast enough to notice the short pulses, which according to the research is the reason the meters are so inaccurate. Fast ADCs are expensive and power hungry, plus you need more crunching power to process the data.

          6. So if a current transformer is no good, would the old fashioned method be better? IE an inline resistor? When I say resistor I suppose I mean a couple of feet of thick copper cable, something in milli-ohms or the like. Then measure the voltage across, using a sensitive amplifier.

            While there are lots of very fast switch-mode PSUs around now, they’re usually pretty low power. Phone chargers and the like. I suppose consoles and computers use them, TVs too. Actually they’re probably more common than I thought. Not sure anything uses old-fashioned heavy iron transformers any more. Still, surely most power consumption is from things using raw mains, big motors and heaters? Little DC electronic doodads must only be sucking a few dozen watts in an average house.

            As far as spikes too fast for the meter to notice, would putting a capacitor across the mains help? To effectively short or buffer any quick spikes. Small enough that it doesn’t carry significant current at 50 / 60Hz.

          7. If you ran two small wires out to the pole, one for each leg, and tap two more into the breaker box at the input– then you already have that section of cable, it only needs you to calibrate the micro (mV/A) in some kind of permanent way.

          8. nowait, facepalm. Trying to produce a differential value using a differential signal to “provide” one half of the differential signal is just about the dumbest thing I thought of all day. In fact I suppose each leg in the mains cable itself would be the other interference-cancelling wire in the pair running the same length and it would be a non-issue, so just 2 small wires… I shall wait here for the police and receive my reprimand

  1. >>> It is worth noting that there is contradictory research published by “the European voice of the providers of smart energy solutions” that maintains that “there is no reason to question smart metering technology”
    >>> by “the European voice of the providers of smart energy solutions”
    >>> contradictory research


    1. To be fafir, their tests of their systems are going to have the same baked-in assumptions that went into designing those systems in the first place. If there’s something they aren’t accounting for, it’s going to cause problems at both ends.

      Not saying they *aren’t* trying to screw consumers, just that there are potential explanations that don’t involve dishonesty. They’d stonewall investigation and resist independent verification regardless just from risk aversion and beancounter asshattery.

    1. Maybe I know of a friend’s, cousin’s, boyfriend’s, aunt, who observed a greater than expected drop in consumed power when 2 Walmart Great Value brand A19 60W replacement LED bulbs were put in a fitting dimmed with a standard old rotary dimmer…. or maybe I don’t, or maybe that’s not the reason it happened, or maybe I dreamed it. After all, the relevant “authorities” claim it’s impossible right?

      Also if you don’t want to be accused of any nefarious activity, make sure you don’t read up on Power Factor Correction as a possible mechanism as to why this happens, and the best way to take advantage of it, even through the free proxy on in case your searches are logged.

      I would also not recommend that you build a table lamp with such a dimmer in, and try it around the house, in case the distance to the meter makes any difference, or it’s most effective on the socket with the shortest path to it.

      I would also not recommend reading this post, or having done so inadvertently, recommend that eye-bleach or the nearest flashy thing be used ASAP.

      This concludes our comedy science fiction interlude for this evening and we return you to your regularly scheduled commenting. ;-)

      1. This is like a casino getting angry at a customer who does something like literally counts cards in their head and then wins more than “expected” and then is accused of cheating and is asked to leave.

        Literally entire books have been written about this, and probably several movies as well.

    2. HAHAHAHAH!!! the funny thing is if you get that -32, they are going to have a tech over there ASAP. and they are going to be VERY demanding about it, almost threatening if you will.
      at least that’s how it went when our smart meter was put in and supposedly reading low.

      1. I don’t doubt your story, but how did they know that you’re meter is “wrong” and that you behavior (energy footprint) hasn’t changed in the same period?
        Big brother is watching you consume…

        1. Yep. Get a solicitor and sue them, claiming that all your readings in the past were too high! Work out the difference in how much you’re charged, and claim back til the point you moved into the house.

          1. I had to do this at one point. They talked so much bs and diverted so much not even the legal help knew what to do to stop the debt collectors banging my door and posting me a letter that cost $300 every day.
            My readings were below average for a single person, they refused to accept anything less than what a 5person family would use and demanded full access to my bank account as the only means of payment. Dirty bustards the lot of em. They had orders to get everyone on the card system as that cannot be undone once installed and costs twice as much.
            I could have been paying for me and nine imaginary people.

        1. Where my parents live at least they do “make up” (or estimate, you can choose how to word it) numbers and only actually read the meter once a quarter or so to correct the bill. Granted, this meter still needs to be physically read by someone driving several miles out of town.

          1. Heh. My dads water company did this. The bill would be some random number centered around a reasonable amount. Even during the time he had a massive leak. There was no way they were actually reading the meter as the box had been filled with debris for years (and it didn’t have the remote reader thingy).

  2. These measurements are deviations from electromechanical meters so I’m curious as to how accurate those are a measuring non-linear HF switching loads found in ‘efficient’ household electronics?

    1. I had the same thought. In the full paper (it’s on IEEE and probably paywalled, but I have access) they found that meters that measure current with a shunt resistor are in very close agreement to the mechanical meter.

      So it’s fair to say that mechanical method is about right.

      Hall effect sensors seem to be too slow to accurately measure the spikes, and Rogowski sensors vastly over-react to spikes.

      1. Get an old meter, paint black stripes on the old spinning disc, stick in a reflective opto sensor, Bob’s yer uncle. Connect that to as much computerised Inter-gubbins as you like. That’s assuming they haven’t sold all the old meters for scrap by now.

        If the modern meters use Hall-effect sensors, does that mean the old trick of sticking a magnet on the side will still work?

      2. Open access link to paper:

        Interesting read. The study uses various loads, and notes that using dimmers (which introduces nonlinear and/or non-sinusoidal current draw) seriously wacks up some meters. The ESMIG response is “nuh uh, that’s not fair and nobody would use their power like that!”

        Because humans _never_ do strange/stupid/weird things with their wiring

  3. it sounds to me that the researchers are talking about average power and the meter reads peak per cycle. If you load their lines with a poor power factor, that should cost you money.

    1. Unless your service agreement says you are responsible for power factor, you should not be charged for that. It is not fair to charge for service in a way that was not agreed to in your service contract.

      1. Then again, manufacturers should be made responsible for the poor power factor of the devices they sell, because the parasitic power turns up as extra I^2R losses in the transmission grid and transformer hysteresis, and the “energy saving” bulbs etc. don’t actually save as much energy as they claim.

        It’s especially bad for LED products which may literally blink at 100-120 Hz to claim that some 6-7 Watt LED (400-500lm) is equivalent to a 60 Watt (850lm) indandecent bulb – yeah, it may seem so because of the stroboscope effect.

        A good way to do that is to make consumers pay for the power factor – include it in the meter and suddenly all manufacturers have to add proper PFC instead of just claiming they do PFC and then not.

          1. Yah, don’t forget we’re still working on explaining why it’s a dumb idea to yank the ground pin to plug protected equipment into 2 pin extension cords.

            It’s only about 10 years since WiFi became ubiquitous, but the modern consumer has already forgotten that internet happens mostly by wires coming into their house, CATV or POTS lines, and you’ll hear them saying their wifi is broke or they need new wifi and realise they mean internet. Then trying to get the concept over that there’s something that converts their internet to wifi, is getting like kicking dead whales down the beach. It’s getting to the point where average people too dumb about it to attempt to help them. Get some thread “my wifi sucks, what’s the best wifi.” and when you’ve finally got it sorted out that they’re actually really having ISP problems, some Charlie Know-Nothing pipes in with “I’ve got a DLINK it’s an amazing wifi and they’re on sale in walmart for only $60” and bugger knows whether he’s talking about his VoIP box or his router, and whether it’s even the same unit on sale in Walmart, but thread derail and victim goes off to buy one, after cancelling current ISP, back 2 days later on a friends internet connection, all stress and panic because yeah, it did nothing. So you haul the thread around again to ISP and suspect it was physical infrastructure probs with either the POTS or CATV in their area, so nudging them to try the other alternative, and Charlene Know-Nothing turns up and is all “Maxleech ISP is the greatest, OMG OMG” and before you know it they’re signing up for a re-seller of their previous problem… aaargh…

            These are the people you’ve got to get to understand PFC, good luck with that.

          2. @RW ver 0.0.2
            In Germany there is this great abbreviation “DAU”==duemmster anzunehmender User==most stupid user you can expect/imagine :-)
            (It’s deviated(?) from “GAU”==groesster anzunehmender Unfall==biggest accident you can expect/imagine on a nuklear power plant.)

        1. The irony is our local power company had a program to trade in you old incandescent bulbs for LEDs. Got a bag of 15 of them. They also gave me $50 for my 30 year-old freezer (it had to work, or so they claimed, as the guy who picked it up was running behind and just took it).

  4. Timely. A lot of people in our electric companies service area are pissed about their bills doubling, or spiking when there’s nothing to account for it, not to mention WiFi interference. I’ve wondered what the readings would be if another independent meter was placed in series? Might help answer some of the questions. Sort of a SamKnows for electricity.

        1. In the US it would likely be a lot more work than that, as most of us have our power meters mounted outdoors. Installation would require having a licensed electrician plumb in a new meter base, calling the power company out to temporarily shut down your power (good luck with that once they ask you why), wire in said base, have it inspected by local authorities, call the power company out to turn your power back on (if they still agree to sell you power), and then pay most of the people involved – Or some scenario similar to this.

  5. My smart meter is consistent with the old analog meter. I was tracking my power usage for a couple of years before the smart meter was installed, and kept watching for another year. Now I use the power company’s data feed and it’s still consistent.

  6. Further point for serious discussion about the accuracy and the “no reason to question”:
    -32% — +582% is not a reasonable tolerance
    That feels an awful lot like +275 (in intriguingly round number) ±307 % : in this usage scenario… Hate to pull out the tinfoil hats, but…
    Also, if anyone ever tells you there’s ‘no reason to question’…
    I guess I need to find a way to put up an classic meter back up between my meter and my panel. I certainly had a jump in my bill when they switched me over…

    1. It’s definitely not +275% +/- 300%

      One particular design of meter (Rogowski) was consistently 500% over
      Another different design (Hall) was consistently 30% under
      One type (shunt) was within 5%

      1. Oh, whew (sorta),
        It was late for me when I read the post, and hadn’t tried to read the study yet. Thanks for the quick clarification!
        Gotta look up what mine is now, I think I hoping for shunt… With my solar panels, hall might under read my input too…
        Now to actually rtfa ,)

        1. Oh “crud”.
          Cursory research shows that my meter is a Rogowski…
          Double coiled to account for alien interference, but still.
          May be a sourced, cited open letter to the power co. and Public Utilities Commission will be in order.
          Too busy a day today full research will be forthcoming!
          Might also look into adding mechanical meters!

  7. The UK has had several thousand cases of stupid smart meters with one man racking up a £5K bill in one day when he wasn’t even home!
    News reporters came around and recorded the meter reading false right in front of their eyes!
    “Oh we are aware of issues with them and were working on it” seams to be the normal response.

  8. Cheap meters measure instantaneous power and average the data samples. AC loads are more complex than that.

    The Current phase can be behind or ahead of the Voltage phase from inductive loads like motors / compressors or in the other direction from CF lighting, Florescent lighting or devices that only use part of the phase like ceiling fans and light dimmers.

    This is called power factor and you need to assess the phase of both voltage and current to calculate true power consumption and cheap meters don’t do that.

    If you run enough power (load) via a LC circuit, you can actually tune the LC to drive the meter backwards on simple meters or if the software permits on complex meters.

    1. The actual power consumption is correct if you measure voltage and current at very high frequency (> 600Hz for 60Hz AC for example), multiply them, and sum them. That’s the discrete integration of power (the mathematical formula is using a integral sum see the section active power from here: ) and it does take into account the power factor obviously. So the any “cheap” meter doing so (using shunt resistor for current in order not to be affected by induction or capacitance issues) will give much better result than the fancy “non-invasive” meters.

      However, because it’s very trouble some to ask consumer to adapt their power factor, most utility simply charge active power and take the reactive power induced loss on their own loss. So either they increase their kWh cost for the average power factor loss, either they cheat the measurement in their meters. The former seems more honest to me, IMHO.

      1. They have already calculated the kWh price for private usage for a non ideal power factor. The electromechanical meter always measured only “real” power since decades, since it’s invention. They never “took the losses” they always calculated it into the price. Therefore there is no justification for increasing the price or cheat in the measurement.

    2. Even the most expensive digital laboratory power meters work this way… the key is to sample U and I simultaneously, that way the phase offset is taken care of. And obviously your ADCs must have the bandwidth to capture all the harmonics.

  9. I’ve worked in the R&D of 1 – 3 phase smart electricity meters.

    A lot of effort goes into detecting and correctly measuring different complex loads, this results in a very complex measuring logic with a very high execution frequency. The idea is really for the meters to measure as exactly as is possible even in very noisy environments.

    In our case, the calibration of each meter was performed in our factory which had been certified by a third party certification body. The equipment for calibration was purchased from yet another third party company which specializes in that kind of stuff.

    The power companies have no control over any of these steps. From an electricity metering production company point of view, they are just the end customers with no insight into the production of the electricity meter.

      1. Deliberately selecting the +582% is a liability nightmare. But it would make them more money. Whether that’s the rational choice even unconstrained by ethics would depend on the payout of the lawsuit vs how much money it would make you. Which value is greater?

        I suspect you don’t know and are mistaking cynicism for realism and insight.

      2. To sell electricity meters, you have to have them calibrated by a certified body (in this case by our own production line).

        The certification to do calibration is legally binding, and the meters are randomly checked now and then by the third party certification body. If the calibrated meters are not within specifications, the calibration certification is revoked and the site can no longer provide calibration. This is more or less a death sentence to electricity meter production companies, or at the very minimum a large economical sink hole because they have to hire a certified company to perform the calibration for them.

        In short, if a certified company messes up the calibration, they are going to regret it. There is likely no economical incentive which will make up for taking such a risk,

        1. I should add that this is how it is done in Europe. Smart metering is the norm here, and the processes to safeguard the calibration process has been around for many years and is (hopefully) mature enough to be relied on.

  10. A few years ago my gas company replaced the meter because they said I was using an unreasonably low amount of gas. After a huge spike in gas prices the year before, causing my bill to go from $50/month to $450/month, I set my thermostat to 50 degrees F and wore sweaters in the house. They never even asked if my usage really dropped.

    1. 50F is only 10°C. OK, this temperature is to be considered to warm for the fridge. But for anything I want to sty and live in it is way too cold. I would accept such a temperature only if I would be really poor and the decision would be if I should buy gas or food. Of course 450/month is incredible.

        1. Interesting, sounds like the house needs some better insulation. I have 320 m2/3400 sqft house that has a total energy bill of ~2000 US$/year – that’s both heating and electricity for a family of 7. And that’s on a latitude further north than eastern Canada..

  11. really reassuring article for somebody that will have his electrical meter replaced by a smart one this or next week… Probably a good thing i have written down all my readings of the last few years (i look at the meter every two months and transmit the reading to the company (btw something that really is no problem for me, i don’t need that automatic stuff) so i get a REAL bill, not a bill based on estimations with a “official” reading + bill correction only once a year), will see if something changes. I don’t remember how the stupid uh i mean smart meter that will be installed here works, seems like the electrical company don’t really wants to be open about his new meter – trade secret and so you know…

      1. Of course, but the argument they use is “trade secret”. The meter that will be installed here is super secured with encryption of the communication and tamper detection that will send a message to the electric company and alos erase the stored cryptographic key(s) – at least thats what they say, but nobody is allowed to verify… I tried to search on the web if somebody has done some analysis of the communication (just because of curiosity, here on HaD people should understand :-) ) but there is so much stupid crap (“used for brain control by the gouvernment”, “will cook the brains of our childs”, …) returned by the search engine i quickly gave up… It’s sad because all this crap hides the REAL problems with these meters: privacy and hackability (by black hats) – and maybe accuray problems too.

  12. I’m a bit of a beginner in this but if i recall knowledge from audio speaker crossover and buck converters: Is there a way to add coils between the power meter and the rest of the house to eliminate high frequency and spikes?

    1. My guess is that that would amount to a very bulky and/or costly solution. It’s probably easier to try and replace all electronics that generate the harmonics in the first place. Best solution would be to tell the electrical company to sod off with their new fancy electrical meter :P

  13. Gone are the days of sticking a neodymium magnet to the side of the meter.
    Interesting that you can add credit with the right frequency. Although I expect they’d cotton on if you were selling it back to the grid.

    1. sure but it might offset any large power consumption one might have and they might blame it one general inefficiency or loss, dunno if they have a way to detect it it but it wouldn’t surprise me.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.