We Have A Problem: Household Electrical

Hackaday, we have a problem. The electricity in your house is on. It’s always on. How fast are those kilowatt-hours ticking by and what is causing it? For most people the only measurement they have of this is the meter itself (which nobody looks at), and the electric bill (which few people actually analyze). Is it silly that people pay far more attention to the battery usage on their phone than the electricity consumption in their abode? I think it is, and so appears another great seed idea for Hackaday Prize entries.

A Better Way to Measure

breaker-panelThe tough part of the problem here is getting at reliable data. Just yesterday we saw an incredible resource monitoring project that uses an optical sensor to measure the turning or the wheel in an electric meter. We’ve seen similar projects for meters that have a blinking LED, and a few other methods. But in many cases the electrical meter is outdoors which makes cheap, easily installed sensors a difficult goal to achieve. Even if we did, this still provides just one stream of data, the entire house.

Alternatively you could tap into the breaker box. We’ve seen [Bill Porter] do just that and there are some commercially available kits that include an octopus of clamp-style current sensors. This is a bit of an improvement, but still requires the user to open the electrical panel (don’t scoff at that statement, you know most people shouldn’t be doing that) to install them. I’m sure there are other methods that I’m missing and would love to hear about them in the comments below.

The Point

To sum up what I’m getting at here, think about the Kill-A-Watt which proved to be a very interesting hack. People liked not just seeing how much power something uses but extending where that data can be accessed. We don’t remember seeing any successful efforts to move the concept ahead a few generations. But if someone can crack that nut it could yield a wave of energy savings as people are able to be better connected with what is using a lot of electricity in their homes.

Your Turn (and Lessons from Last Week)

As with last week, now it’s your turn to come up with some ideas… wild, fantastic, good, bad, outlandish, let’s hear them. Better yet, document your idea on Hackaday.io and tag it with “2015HackadayPrize“. You can win prizes just for a well presented idea!

Speaking of last week, I shared the idea of adding some feedback to how long you’ve been in the shower. There were many opinions about the value and worthiness of that idea so I thought I’d close by covering some of them. Yes, there are much bigger wastes of water (and electricity in this case) in the world but why limit our solutions to only the largest offenders? The low-hanging fruit tends to be stuff a lot of people can understand and relate to. If we only talked about large-scale fixes (I dunno; reducing mercury emissions from power plants?) there is little momentum to crank-start a movement. If you found yourself thinking the ideas from this week and last are far too simple to win The Hackaday Prize that means you better get your project going. The world is hacked together by those who show up.

I’d love to hear suggestions for future installments of We have a problem. Leave those ideas in the comments and we’ll see you here next week!


The 2015 Hackaday Prize is sponsored by:

106 thoughts on “We Have A Problem: Household Electrical

  1. Rainforest Automation makes a very cool energy monitoring unit.

    http://goo.gl/70oR16

    It connects over Zibgee to your meter and provides live readings of your power consumption and your current bill rate. They even have an API you can interface to grab the data from a USB port on the back. It works with any meter with a “OpenWay” logo on the front.

  2. Just a thought….

    Would some local building codes even allow installing “extra” doodads in the breaker box?

    My dad used to do all sorts of crazy electrical updates to his house as I was growing up but the cardinal rule was it had to pass muster with code. Kind of strange really since he didn’t feel that way about the plumping…….

    1. The breaker box is nothing more than a box with protection in it, nothing more. There’s nothing that says it needs to even be a box.

      Distribution boards like this typically can have all sorts of things in them.

  3. Uhh? Wireless CT coils are already widely commercially available for wireless real time energy use monitoring. It’s basically a wireless clamp multimeter that you install over your mains wire or branch circuitry or both that feeds usage data out.

  4. http://www.neur.io/spec/
    I just supported this kickstarter, and received my unit in the mail. It’s AMAZING.

    I’d like to send my photos, but I’m not sure how to put them in the comments so I will summarize.

    It gives you live updates of your Watts around twice a second.
    When you turn on the stove I notice a 2kW increase. Then you turn it off and see it drop back down.
    I noticed something last week was drawing my house up to 8kW, so I began investigating.
    Turns out that my drier uses 5kW continuous while drying clothes.
    As a reference my Nissan Leaf uses about 1.3kW since I only have a 120V trickle charger.
    The electric kettle I use every morning uses about the same 1.3kW but obviously only for a moment.
    The microwave uses around 2-3kW, the oven uses around 3-4kW etc….
    The house uses around 200W at night with nearly everything off.

    I know all of this now, but I knew none of this a month ago.
    It’s on my phone and they are working to build out their app to support automatic appliance detection for the big ones.
    They are also working toward IFTTT and Wink integration. Thus… Send me a notification if the usage is over 500W at 11pm. Etc.

    They could use support, and if you buy now you can use a promo code: MYFRIEND to receive 20% off.
    Happy kW hunting.

    1. Looking at their website, it doesn’t seem to require any more input information than any other commodity energy monitoring unit, ie a single current clamp (or in the case of some other units, reading the flashing led)

      So, nothing new here.

      1. It’s all in the software and accessibility.
        I agree with you. Someone commented on my FB post about using a multimeter… :)
        I think the difference is, when you finish vacuuming… then think… I wonder… pull out your phone then see exactly how much energy your vacuum uses from your smart phone, it becomes a powerful experience.

        1. Hey Sean

          Thanks all the great comments. You are absolutely right about Neurio being a powerful energy investigative tool. You can instantly see what each appliance is using and how much it costs you each time you use it.

          We also go beyond energy though, and truly make your whole home smarter. We will let you know if something is left on by accident, or if your fridge is about to breakdown. We will even be able to tell when your kids are home and what they are doing, ie. playing Xbox.

          We also work great with other home automation products to make them smarter. We can feed our data into smart thermostats like Nest and truly smarten up there home presence and energy features. We also are open platform and love developers using our APIs to create other amazing things. The possibilities are endless.

          You can check out our community page at community.neur.io for lots of great user stories and feedback.

          1. I’d buy one of theseYESTERDAY, if there is a way to use it WITHOUT beaming all our private household data outside our house.

            The reliance on the inherently insecure cloud is a non-starter for me.

            Your security, like most better funded banks, insurance companies and utilities, can and will be broken. This point is not up for debate.

            I want a completely closed system where you (or anyone else for that matter) do not see or store our data, period.

            Let me know when you have something that runs privately over my own network, without outside connections, around or below your current cost, and I, along with other security conscious homeowners, will be much more inclined to buy.

            Thank you.

          2. you will not get what you want, but your sense of entitlement is amusing. Your minority concerns do not a market make… but its very cute you think your electrical data is special or interesting. Go away snowflake.

        2. But uploading all the power usage to a remote Internet site, just makes my skin crawl. Someone bypasses their security and profiles all the data and works out when would be the best times to rob houses. geoiplookup’s for course grain profiling to narrow down the data to saleable regions. Every time see cloud, I think cheapest provider, and the collected data can be sold on to third party advertisers. Look customer X from their electricity usage is expecting a new baby, time to spam them with nappy ads.

          1. Hey Truth

            Privacy is definitely something we don’t take lightly. We go to great length to ensure everything is secure. We store user data, like name and email, in one encrypted server and sensor data in another encrypted server and they can never be crossed referenced. We have even held tons of hack sessions to see if outsiders can penetrate anything and point out potential flaws. So far nothing.

            Neurio

          2. Would you rather have a stranger enter your property to read your meter each month? Also not to hard to watch the speed of the disk spinning on the old analog meters to see when someone is home. You don’t even need to be a super-hacker to do that…
            With all of the effort to secure the data that these SmartMeter projects put in, I see them as a step-up for privacy concerns.

          3. Unfounded concerns. Rep is too polite to explainthis to you because he would have to admit you are not special. Of course dreaming up overly complrx hypothetical fears is some kind of past time thes days. beware super hacker burglers lol not shady meth heafs with a brick and a couple hours free time to see when you leave lol

  5. Many utilities are using “smart meters” that automatically send usage back to the utility for billing purposes. No more meter-reader. There oughta be a way to tap into that info.

    1. Not only are the utilities using SMART Meters but many are forcing customers to accept a SMART meter whether they want one or not. Some like PG&EIN California even go so far as disconnecting a customers power if that customer has had the smart meter replaced with a traditional analog meter.

      If the need for a SMART Meter is as the utilities claim, to get usage data, which is clearly a violation of ones privacy and home (because the SMART meter allows the utility to directly control the system and any smart meter capable device) then why is it not at least OK for the customer to get that same data?

      I believe the reason the utilities want this data and are not wanting customers to get access is so that they can manipulate the metrics and over charge customers. The SMART Meter and Grid were marketed as being a way to save customers money on their power bills yet there are numerous examples on the web of people being over charged and or not seeing any improvements to their power bill.

      1. Hi, I design “smart meters” for a living.

        Smart meters are more accurate than electromechanical meters. I know; I saw the data, and I was there when my former employer dismantled their electromechanical electric meter production line.
        It is not a violation of privacy. They use time of use information for one purpose: peak shaving, which means knowning when to turn on the expensive natural-gas generation capacity because they are reaching peak capacity on their base load generation.

        This Luddite approach to smart meters does not behoove the hacking community. Get off it.

        1. This is true, I work for an energy company, time of use discount plans are very beneficial to the consumers that choose those plans. We call it interval data (hour by hour usage), and it also helps with forecasting usage and planning product offerings in unregulated markets. And it’s not new. Large customers (factories, mines, etc.) Have been using interval data billing since the late 70s (as far a I know). The meters had tapes that they would collect and manually load the data into the mainframe.

        2. Er, can’t you tell how much power people are using, at the plant itself? Surely they only need to know how much power everybody, as a whole, is using. Why would it matter to them how that’s split up between individuals?

          And don’t usage patterns tend to vary predictably, with weather, time of year, things like that? As well as major TV events. What’s any of that got to do with knowing who switches on each appliance and when? Surely the best predictor of tonight’s usage is last night’s and the night before’s, not what I’ve been doing this morning.

          It’s not Luddite at all. Being concerned with information about yourself, and who has it, is very hackeresque. As is not trusting gigantic companies when they tell us “Oh this? It’s just a thing, go away, nothing for you to see”.

          1. no, you determine grid load by monitoring the frequency.

            you can try this yourself, get a motor and connect nothing to it, see it’s easy to spin (no load)
            short the terminals (huge load) it gets harder to spin.

            under low load situations grid frequency will increase, and under peak demand the frequency reduce.
            monitoring this allows you to take two courses of action.

            either ask some people to turn stuff off. (and in the UK at least if you can turn off a large supply with a sub 30 second notice then you can get paid for that – E.G a data-center with large backup battery, and rooftop generators can isolate from the grid and stay alive.

            Or you can turn on more fast generation. (e.g. hydro) where a stock lae can drain and provide huge amounts of power instantly.

            you take these instant response measures whilst you light your gas burner in your fast response station.

            For all intent and purpose you only want to use something like coal to generate the power used to run freezers since they are constantly on. they (coal) stations are very slow response. wind and tidal power and solar are also very slow response, and in any case you can’t just put a bit more sun out because a lot of people want a cup of tea in the same way you can light a furnace ready for the end of “insert popular drama here”

            these methods combined can stop grid blackouts. since each connection point in the grid will isolate a grid section which is believed to be dead. there is a case (I can’t remember exactly buk I think in the UK) but of a single power station failing and causing a huge black out, because the effect of it failing meant that section of grid was isolated. and then adjoining grid sections were also isolated. (grid sections are isolated because mostly a lack of power at an interconnect is caused by a downed cable, and at that point you want to ensure that the cable that’s on the ground is not live.

            in practical terms what this means is…
            No. smart meters are not used for determining instant grid usage, because there is a far better way.

            smart meters are very good for metering.
            traditionally, people read meters, then those people fill in their paper work and take it back to the company.
            these people take five – ten minutes on your house, and can therefore read about 6 – 10 meters per hour.

            companies then asked people to read their own meters. but consumers can’t be bothered, and unreliable and lie.

            so we have smart meters, which either are read by a near field device (i.e a street of 50 houses can be read simply by driving down it. which takes about 1 minutes) or can phone home themselves.

            This saves the energy company money, since fewer people can great greater amounts of meters.

            This does of course save the company money. however, the savings aren’t great.
            that guy that read meters is probably getting minimum wage. whilst he once took 1 day to read 50 meters he now reads them in a single minute, (great news you may think…)
            but think about the math.
            1 guy minimum wage (~£6 GBP) for 7.5 hours is £45 or ninety pence per house hold. (50 houses).
            meters are read four times a year for quarterly billing. so each person paid £3.60 on their bill to have their meters read.

            Now even if that guys job is now so easy that he does it for free. the saving is still only ever going to be in the region of a 0.5% or less saving on the bill.

            knowing how people have used power is a reasonable indicator or what sort of generation demand you will face. but historic grid load is a better indicator than looking at when you turn your lights on…
            and in fact the company don’t need a meter to say when lights will go on. given that there are perfectly good resources to tell you when sunrise/sunset is each day.

          2. Power isn’t that simple, at least in the unregulated markets we work in. The basic rundown is this: the local utility (LDC) distributes to the end customers and meters usages. In unregulated markets you can choose your provider, and that’s what our company does. We collect usage data for all of our customers from the various LDCs in the markets we serve, and then we bill the customer based on their selected rate plan. On the other face of the business, we use all that historical usage data, account add and drop rates, plus weather forecast data, add a little statistics, and use the resulting data to schedule capacity with distribution and generation companies, on behalf of our customers. The better the forecast the more accurate the scheduling, the better rates we can get and thus better rates we can offer our customers.

          3. can’t reply to Dan, thread limit reached.

            Dan, it might cost £3.60 to read the meter and these smart meters might save the utility that.
            However a couple of points.
            1) in the UK it’s a fact that we are having cost added to our bills to pay for smart meters.
            2) we’ll never see the £3.60 returned, it’ll go straight to shareholders.
            3) smart meters offer us zero savings as consumers, it’s all bluster and marketing BS.

            I personally see no benefit to me what so ever in having a smart meter and I’ll be refusing it’s installation as is my right to do so (read the HMG website about it).
            It offers me nothing of practical value and I’d suggest you go read the T&C from the various providers on who owns the meter, who is responsible for the meter and who is actually going to foot the bill if it or the remote reader device gets damaged.
            Hint: It aint the utility.

            I’m happy to read my own meter and thanks to that I’ve made savings in the past for example when Scottish Power charged more in winter than in summer (circa 2009) for gas and electric usage (higher verses lower unit charges).
            So funnily enough I used less in the winter than the summer and ended up with a cheaper bill due to my declared readings.
            If they to try to screw me I’m happy to payback the favour.

            Smart meters just put more power in the hands of the utilities.

      2. Respectfully you can’t have it both ways. Questioning what the utility says to make assumptions and not questioning what you read on the web; from whom I assume largely remain anonymous I’m curious how those persons went about figuring out if they where over charged, any links to how they measure their consumption? Measuring it in same exact way how the utility measures to determining billing.

      3. “I believe the reason the utilities want this data and are not wanting customers to get access is so that they can manipulate the metrics and over charge customers. The SMART Meter and Grid were marketed as being a way to save customers money on their power bills yet there are _numerous examples_ on the web of people being over charged and or not seeing any improvements to their power bill.”

        Way to submit to the FUD machine … got any REAL examples??

    2. Hi, I design smart meters for a living; presently, it’s water meters, but have designed electric and gas before.
      Note to the hacking community: the metering companies, and their utility customers, do NOT want you to have this data. Most everyone is going to AES encryption of the packets sent by the meter.

      1. I wouldn’t want my smart meter to broadcast unencrypted usage data. At least when it’s encrypted, I know only one consumer gets the data – the power company. If it’s not encrypted, anyone could grab the data and use it. For example, suppose a smart thief drives around scanning power meters for low usage indicating unoccupied houses.

        If I want my usage data, I can get it with any of the monitoring products listed in the post and thread. I don’t want it to be broadcast unencrypted. That’d be a much bigger privacy violation, wouldn’t it?

      2. The way your post is worded makes it sounds as if metering companies want to actively prohibit the end user from measuring their own consumption. Do you believe that to be the case, or is it as Nate says that it’s merely moving to properly encrypting the data.

        I don’t know if it stands to reason that utilities would unite in a cabal against the consumer in the way I feel your message intones.

        1. It’s not too technically difficult or expensive to install your own metering stuff on your side of the meter. Your house’s wiring, after a certain point (usually where the meter feeds into the fuse box), is your own. Either a clamp meter, or if you’re brave, wire a current meter in. They can’t keep the information from you when it’s you who’s creating it.

      3. I worked on a SmartMeter project recently. I can tell you that security was a very very high concern. Everything is encrypted and signed typically at several layers and sent over private networks. The meters had a slew of keys each for a specific purpose. Many keys are device specific so that if a hacker somehow pulled keys off of one meter, it was only good for that one meter. Even with keys to the kingdom, there was no way to say shut off the power to multiple endpoints. Additionally, monitoring the security of meter data traffic was not taken lightly. There were regular tests of key roll-overs and many other regular security mitigation procedure drills.

        Customer access to data was also a huge part of the project. Any customer of this particular utility can go download multiple years worth of data quite easily. Additionally, those customers could purchase devices that would show near instant power use through a zigbee connection (encrypted with keys specific to that device of course) like that mentioned in the first post.

        As a side note, this utility (and all in the state as I understand it) allowed customers to go back to the old analog meters for a fee. They had to pay someone to go out to your house each month to read it of course. But some people had weird ideas about “electromagnetic sensitivity” even though the meter put out less EM energy per day than your cell phone does in a year or more.

      4. My electricity supplier lets me have the data … but only on their website several hours late in hourly intervals.

        You are however free to install secondary meters after the main one, and use S0 interface to do some logging, if wireless CT is not good enough.

        1. That’s how they get it. If you want real-time high resolution readings, call them up and ask what kind of in-home displays they support. If they are using one of the major SmartMeters, they probably support all zigbee devices that implement the latest “smart energy profile”. Your meter will send this data at sub-minute speeds (resolution depends on how the utility configured the meter). But your utility never gets this data. You’ll have to get them to provision the device with them which pairs the meter with it.

    1. would be so much simpler to use an IoT enabled breaker wouldn’t it?
      Replace your standard breakers/fuses with bluetooth or wifi enabled versions.
      Then your smartphone app could tell you not only WHEN a breaker tripped, but using a relay, RESET the breaker remotely! No more flashlights, and poorly labeled panels.
      Down sides would be larger breakers, and therefore larger panels(probably double sized, mostly for the reset relay…. the IoT part could be pretty darn small). More expensive, and hackable(so ONLY allow reset if tripped, so wardrivers cannot turn your security alarm breaker off)

      1. If by reset you mean reconnecting the load, i doubt that is a good idea. In an ideal situation the breaker will not trip unless there is a problem, so maybe person should be there to investigate the problem and react to whatever happens when reconnecting.

        I like the idea o IOT in each thing, but people would most of the times choose the lowest cost that does the job. Not so many people out there would like to spend extra money to have this bonus, even if it is dirt cheap. If a breaker costs $5-10, would you pay $5 more to add Bluetooth to it? But what about if they could add bluetooth for $1? A device that has 20 inputs sampling all the outputs of the breakers could tell you which is ok and which is not for a lower cost, but it would be more difficult to add.

  6. Few have that analog meter anymore pictured above. Smart Meters have been rolled out nationally…with no public input or debate, and they are dangerous on a number of levels.

    1. Hi, I design “smart meters” for a living.
      No, they are not dangerous. At all.
      They do not cause cancer. They don’t spy on you, or email your porn movie rentals to the authorities.
      They barely have enough flash memory for their one and only mission: send time-of-use and consumption data to the utility company. You know what a struggle it was to get the engineering management to let us use C instead of assembly, because of fears of “memory bloat” and that it might cost them another 20 cents US for a bigger microcontroller?

        1. Most use 900Mhz and a mesh network (encrypted of course). The power transmission is very low (I think the ones I worked on were 250mw) and so they tend to work best in high to mid density areas. Additionally, they only transmit for a few fractions of a second a few times a day.

      1. > No, they are not dangerous. At all.
        Except if you lick the wires ;-)
        > They do not cause cancer
        I’ll agree with this
        > They don’t spy on you
        The meter doesn’t, those who get the data … I’m betting they don’t call it spying but rather marketing or input to other customer services. There’s someone at the utility trying to figure out a way to sell that data to someone.

      2. Well, now I am convinced they are bad. If the utility company sends out the shills with copy/paste forum responses to defend their technology, then they must be hiding something.

    2. Except for cooperatives or in inatance the utility was publically owned, I doubt there was any public input or debate when the electric grid was first rolled out. In most if not all States the public can provide input when it come to rate increases. I never have read of any customer input periods when it come to infrastructure changes, unless it involved right of way on private and public property. Expecting public debate and input on the roll out of smart meters that could effect the roll out is like expecting similar whenever Walmart changes it use totally up customer purchases.

  7. Look up TED5000… in fact…

    http://www.theenergydetective.com/

    I have one of these and they are awesome. works great, trivial to install if you can avoid licking the wires in the electrical panel, and it has a Full XML page of all data for your own processing if you want it.

    dont bother getting the kit with the remote display, just get the basic kit.

  8. Here’s what you do, take a watt meter, test the appliances, remember the (real) values, or jot them down on a label .
    At that point you know, the devices don’t change much. The only real variables are the heating/cooling, but if it’s a fridge you have no choice and if it’s a heater or AC unit you bloody well know that at it uses a lot of power and you don’t need some dial to tell you.

    1. Except when they DO change. Energy usage is a good indication of an issue. One December morning I got emails from my (self-created) energy monitor app and my thermostat about a heating issue. My heat pump compressor died and my heater was running on 15KW worth of heater strips instead of the 4KW heat pump. Without these devices I might not have found the issue until a very large electric bill. .

  9. If I was going to do this commercially then I’d either:

    1) Design a fuse board that had a CT per fuse/circuit.
    2) Design a plug in adapter that goes into an existing fuse board and the fuse then goes into it.

    One size does not fit all. I’m in the UK, so I know about UK electrics not French, or US which I understand are radial/ UK is ring wired for most circuits.
    One thing I’ve always done in a new house is to map out the sockets and light fittings and make a note of which fuse they connect to. In larger houses this can should up oddities in the ring wiring. In smaller houses it’s often shown up potential overloads.

    But what exactly are we trying to monitor here ?
    Usage? Why ? What’s the purpose ?
    The comment above about “usage over 500W at 11pm” I just don’t understand.
    Why do you not know what appliances you’ve left turned on ?
    You’re the guy that goes out and leaves the oven on, right?

    I’ve got two fridges.
    They are my base load along with a couple of mains powered devices that are always on, like timers, the router, couple of switches and a PC.
    Everything else I do this magical thing with::

    I turn it on when I want to use it, then turn it off when I’m finished with it.
    Wow. Imagine that.

    Thus no surprises.
    I don’t understand how people don’t know what their usage is (or what is using it), thus want to spent considerable sums of money on monitoring their usage (the irony being it uses more energy) when it seems like they are just really being lazy or stupid.
    Convenience seems to be the buzz word for lazy.
    Monitoring for the sake of monitoring I get. I do that too. It serves no actual purpose. It just makes pretty graphs.

    Want to know what’s using your power? Turn everything off (at the wall) that you can and then start tripping fuses/breakers till you find no draw. Map your circuits, see what’s plugged in.
    Bingo. You’ve found that plugged in set of curling tongs you dropped behind the sofa 3 years ago..
    Sheesh.

    1. Thank you, a shred of common sense goes a long way! A friend of mine once told me his grandmother used to have something plugged into every outlet so that the electricity wouldn’t leak out of them.

  10. how about a new SWITCH BOARD – for wall outlets.

    from the outside looks like a normal switch-board, 8 switches lets say
    YES, an electrician will be needed to put it in – from the back it should have CLEAR screw connectors for lights etc.

    from the inside, it has BLE or Wifi or RF (2.4GHz connecting to one central switch board with Wifi/BLE/network connection) – Has a current meter to measure the consumption of the room, can check consumption of individual devices by some clever programming and overall house stats can be checked/visualized/tested.

    * I am willing to make a PROJECT around this, have got some work already done. so any ideas/suggestions…???

    1. Can’t do that anymore… at least with Itron meters. The data is encrypted. This was possible in the very early rollouts in the proof-of-concept markets. These have all since been upgraded with much more advanced communications.

  11. While I applaud the intention , this has been tried to be solved by several domotica systems already and there is simply no way around clamp-style current meters or a dedicate current/voltage monitoring modules .
    Stuff that needs at least some basic level of electrical knowledge (and i agree – you better keep as many people as possible out of the electrical panel ) to install.
    You have also local regulations involving stuff like this and plenty of different meters (if you want to to leverage those) around in different country’s / regions.

  12. Don’t fit a Kill-a-Watt to your house!
    You need several kV-rated voltage sampling circuitry, for the resistor divider and PCB clearance. Or else…
    Province of Saskatchewan has several house fires from commercial Smart Meters arcing.

  13. The problem is that once you go through the hassle of installing current transformers (clamp or just running a wire through a non-clamp one) you quickly figure out what outlet or light circuit draws what current and that it doesn’t change enough to be important – or – more importantly YOU will not change your habits to make any sort of difference in the end.

    The kill-a-watt plugs in, tells you what you need to know and life moves on. The same thing with a simple clamp style mulimeter or multimeter attachment. Measure your lighting circuits or dryer or whatever, and do a little math based on how long you are using them and you have your data. It’s not really going to change enough to be meaningful to have a 24-7 log running.

    You know your DVR or washer and dryer uses x amount of energy. SO WHAT? You going to stop recording your shows or washing your clothes? You only need to record data once, and then it’s up to you to unplug the thing or go stinky!

    “Measurebaters” as my wife likes to call them, are silly people who think logging every detail or fuel consumption or energy usage is going to somehow magically make the world a better place by having detailed data that is really kind of meaningless.

    And here’s the best part. If you design an electronic device to measure how much energy you are wasting, you are only wasting more energy!!! Even energy “harvesting” devices are a load that must be fed. Physics is a cold hard bitch, I know.
    No free lunch. Even the water one, which you would think restricts flow and therefore uses less water, but it must be made in a factory(A LOT OF ENERGY), and it probably puts a higher load on a pump somewhere upstream.

    The answer to all of this lies in you! – Turn it off, unplug it, go outside and breathe the fresh air and stop over thinking things.

    1. If you are like me at my house, there are always various things plugged in like chargers and what not that are using a little energy. also there are much less than efficient light bulbs that get turned on and off, fridges and other appliances which are a bit older and maybe not the most efficient. who knows what else. I haven’t done a check, probably because of lazyness, but the electric bill isn’t all that bad and I don’t worry about it. However the goal of reducing global energy consumption is a good one, really I should re-evaluate what I do and make some changes, I’m not all that incentivized to do so though. In steps some product in this category which applies a social and gamification structure to your energy usage, now maybe I have a bit more incentive to reach a “low score”, do better than my neighbors, etc…

      1. I got one of those 4 gang extension cords with a switch on each outlet.
        it wasn’t much more than the 4 gang that i would have needed to plug in myphone + work phone + tablet charger etc anyway.

        flicking a switch is easier than unplugging a cord

    2. far be it from me to break you clouded view of the world but:

      yes, I measured energy in my house and made changes.
      I swapped down light bulb (50w halogen) bulbs in the kitchen/dinning room for 3w LED ones. (with ten bulbs (500W) and electric at 15p/Kw hour and around 4 hours use each day) I once had a 30p/day bill (£109 per year) that was cut to 30W 1.8pence per day or £6.57 over the year a £20 cost from ebay for the bulbs saved me £100 in the first year.

      each singular halogen bulb that you change to LED saves you between the cost of the bulb and thirty times the cost of the bulb depending on the amount of time you use it and the bulb cost (the ones I got were £2 each.)
      £65 per year constant use for a £50W bulb vs £3.90 for a 3w bulb.

      I love driving in a “spirited” fashion. however, the difference between driving like a knob and driving carefully is getting just under 500 miles from a tank and just over 600 miles from the same tank. – so whilst I love driving as though I desperately needed to be there yesterday, most of the time I don’t.

      Most people, at least most sensible people) do modify their behaviour to save costs where they can, and where it is convenient. (so no I won’t stop recording my favourite shows because I’m happy paying the £65 a year (or 17p per day) that a 50W PVR costs (that assume pretty constant disk use also!

      at the same time I swapped the mechanical disk in my PVR for a solid state one that uses less power (and makes less noise), I swapped my file server from the old pentium 3 desktop rig that I’d kept going for years at 100W (£120 per year) for a raspberry pi with USB disk attached (5W + 10W) £18 a year to run.

      1. None of this actually needed you to monitor the usage.
        It’s obvious that a 50W bulb is going to use more than a 3W LED bulb.
        You can do the maths on paper (hint 47W) against your KWh rate.
        And not to include the standing charge now (nice how our energy companies are now screwing the people that use the least electric and gas in the UK)

        When swapping your hardware I hope you factored in the cost of purchasing the new devices and the payback of that cost over x years before you again replace them.
        Fag packet maths suggests a good payback.
        That said, my P4 PVR goes to sleep when it’s not recording. So it’s not running 24/7 and thus is costing me around the same as the raspi + changing out all the hardware (PCI DVB cards at present).

        On the car thing, I get 42mpg and I drive with a binary throttle. I can get 50mpg if I don’t and if I baby it I can get 70mpg indicated – I did several tests to see how much my ave mpg could get down to over 90mins of urban driving.
        The savings in going from 42 to 50mpg do not replace the value in hoofing it everywhere, thus I don’t bother. 70mpg driving style is impractical.
        Now if it were 15mpg to 20mpg, different story.

        In my fun weekend car, I don’t monitor the mpg. It’s a leisure activity.
        The money on parts far out weighs the mpg so it’s irrelevant :)
        But then I do all my own servicing and engineering on the cars (as a hobby) so that saves a fortune on what most people do.

  14. IMHO, every wall socket and every lamp socket should have its own current and voltage meter and report/be acessible to one or more master nodes for reading and controlling. Extending the idea a little, every such wall or lamp socket could be controllable (ON/OFF, for instance) through the same protocol.

    The comunication will probably be mixed between wireless and wired.

    I think the challenge will be to produce a simple and cheap device that, at the same time, consumes very little energy so the losts as minimized. The communication poses another problem but I think it is manageable with current tecnology.

    Looking a the bright side, if you have a way to meassure and turn on or off the loads, it would be possible to schedule every shutdown so you can minimize energy use when you are not at home, or every night and so on…

    1. If Leviton or whoever thought they could sell smart outlets and wall switches or breakers or whatever, they’d already be doing it. It’s not like this stuff isn’t kind of already around. the technology certainly is.

      I’d say they have looked hard at the market and also considered the things in my post above about the data being really kind of pointless and decided to stay out of it because the profit wasn’t there.

      $0.99 for a dumb low end light switch or $40 for a glorified IoT switch and a hub to gather all the data? If I was building a house, I’d probably opt for the cheap switches even though I have a couple of Z-wave wall switches and a smartthings hub already in my current home. Things like outlets and wall switches add up quickly and no body has really nailed down a firm communications protocol. If I build my home around z-wave, who’s to say hubs that talk it will still be around a year from now? And I find automation to be wayyyy more interesting and useful than recording energy usage, and I love stuff like this. Joe Hombuilder isn’t going to lay out any more $$ than he has to for sure.

      Even wifi radios change, and who wants to upgrade a whole house worth of outlets and switches every few years?

      And again, if you are monitoring energy usage to conserve energy, you better be considering how much energy you are using to log the data and if it’s even worth it!

    2. A standard wall outlet cost about $5, one with USB ports is about $20. I imagine that a “smart” one with wifi is about $50. My house is about 2,400 sq ft and has ~80 outlets.

      Do the math and it would cost $400 for normal outlets and $4,000 for “smart” outlets. How long would it take to save that $3,600 by saving a few pennies here and there? This isn’t even counting the cost of the base station and time it would take to get everything running.

      I’ll take my “dumb” outlets any day of the week.

      1. A “smart” outlet would be the USB one (for the power supply) powering some litle Atmel or PIC. And I bet the USB one includes plenty of profit and tax on it, manufacturing price will be nothing like that high. It’s always easy to find things to add on to the bill if you want a high number (and things to ignore if you want a low one).

        RS-485 would be fine for data, just another couple of wires to run to each socket when you’re building the house. If the house is already built, then use one of the little radios that are all over the place, few dollars each, for an ad-hoc mesh network. The nodes are going to stay in the same place, you can always add a wire antenna if one of the nodes is in a dead-spot, and a ceiling node for your lighting would give good radio coverage. The few bytes needed for “I am socket 5 and I am supplying 7 amps” every minute or so is almost nothing.

        This isn’t really a complete design I’ve thought of. Just making the point the technology is old, cheap, and easy. You can buy a set of 3 radio-controlled mains sockets for 15 British pounds in the supermarket. Adding consumption-reporting and a 2-way radio would cost very little extra.

  15. I made this : http://i.imgur.com/icqgxjT.png (the design is work in progress, but the data is good)

    It works with the regular french electricity meters that everyone has here (I mean if it has been installed over the last 15 years or so) providing that the “TIC” output is enabled (they can enable it for ~30€). The data on the TIC output is ASK modulated and once you convert that to actual bits, it reads like a normal serial stream of data (1200 bps, 7bits/char, 1 stop bit, etc.) I read the “indexes” (not sure if you call it that in english, the number of Wh since the meter has been installed) every 20 seconds and store them in a mysql db.

  16. Has anyone made some simple (wireless?) sensors that can be attached to devices to measure their usage? Then you also collect the aggregated power usage of the whole house and overlay the sensor data.

    You could record when the door of your refrigerator is actually open, and then correlate it with spikes in power usage. But you never need to directly measure how much power the fridge uses.

  17. You know that picture of the breakers in the article made me think. Breakers can be replaced, why not make a breaker with a wireless chip inside that sends usage data to your network?

      1. In what universe? Here on earth 120V @ 20A is available in many wall outlets, certainly 15A, and people plug wi-fi routers into them all the time without starting any fires. Now the battery in your cell phone, on the other hand, those occasionally catch fire.

  18. I’m currently using 2 Brultech ECM-1240s to monitor both phases in my home (USA, 120vx2). I dump the data to my db and then analysis the data. It helped me figure out that I had a bad drop last summer.

    And as far as smart meters not spying on you, they don’t but don’t be so naive to think the utilities won’t turn around and harvest the data for resale. That’s a different type of spying.

  19. I think a really good home automation system with addressable outlets and switches would do far more. I know this all already mostly exists, but it is a mix&match of several systems and never seems possible implement exactly the way you want.

    For example, I would love to be able to push one button in my bedroom as I lay down to sleep that just shuts off every unnecessary outlet and light in the house, make sure doors are locked, windows closed, etc.. I know that is kind of simple, but that’s just an example. My biggest waste of energy is the fact that I forgot I left a light on somewhere or remembered as I was dozing off and said ‘screw it, I’m not traipsing all the way across the house flipping on every light in my path to go shut it off.’ Second would be things you normally leave plugged in like chargers, cable box, etc… None of that stuff needs to be running while you are asleep or gone, but who goes around unplugging everything every night?

    I have standard fluorescent lights in my basement work shop which should be pretty energy efficient. But I was really surprised at how much it reduced my bill just making sure they were off. Knowing how much energy they draw isn’t really going to prevent me from dismissing it when I realize after I am comfortably in bed or gone.

    1. Better yet, switch to battery power at night (after ensuring every uneccessary outlet is shut off) and just shut the whole house down.

      We do wasteful things out of convenience. We know it is wasteful, but just consider it the cost of convenience.

    2. Before considering a switch in every outlet you should consider how much power you save by forgetting something is on vs how much power that electronic switch consumes.
      I think in terms of lights, the biggest and simplest automation is to install motion sensors and dark/light sensors.

  20. More hackaday fail, The powerplant itself is the biggest user of electricity. It’s silly to even bother with low hanging fruit. This is akin to fixing a dent on a car that won’t run, when in actuality the distributor coil is broken.

  21. I’ve thought for a long while now that houses should have a DC power grid in addition to the AC system. Nearly every electrical device in my house save for the big appliances and desk lamps use either a wall wart or an internal DC producing power supply.

    It seems to me like one large AC to DC conversion system at the breaker and a DC power grid coming from it would result in less energy waste than a thousand and one small converters with lower efficiency in the conversion. I’ve seen some people on youtube show off “DIY solutions” using old PoE routers to convert their homes to DC power and claim energy savings from it.

    1. The problem is that the DC is mostly low voltage, usually 15V or less. Furthermore, it’s often slightly different voltages for different devices, so you still would need converters. If you try to distribute that low voltage, unless your current draw is REALLY low, any efficiency gains you get from a centralized converter are eaten up in the I2R resistive losses in the transmission lines. The general rule to maximize efficiency is to keep the voltage high for as much of the distance as practical. 12V is OK for distributing power over something about the size of a car. 120V is OK for distributing power for something about the size of a house. A few kV is OK for distributing to a small subdivision. Higher voltages are required for longer distances.

    2. I’m personally very surprised the use of POE had not caught up more.
      I’m moving to new home and I have in every room at least 2 RJ45 Ethernet slots who I’m connecting to a nice switch providing POE to all sockets.

      1. I looked at using PoE to replace all the wall warts on the various devices I have around the home.
        Conclusion, its’ not cost effective to do so, retrofitting DC-DC injectors at the device end to replace efficient modern wall warts.
        For a device which is already PoE powered yes. But the switches in larger sizes can be expensive. It’s much cheaper to get a 48V-56V PSU, an old patch panel and build your own passive injection setup for home distribution.
        There are some consumer grade routers and access points that are capable of being PoE powered. I use DLink stuff because of that to save on having to power the end device via a mains socket (only reason) and thus can easily fit in voids and difficult places that are easy to get a CAT5 to, but not a mains socket.

        When you’re buying devices (CCTV cams are a good example) the cost difference between the PoE and non-PoE version can be considerable.
        I’m talking the BOM/manufacture cost, not the end user mark up.
        It would absorb the cost of a wall wart often enough. Convenience costs once again :)

        Curious what you plan to plug into all of those sockets that is PoE powered and standalone in such a manner that it wouldn’t be easier/cost effective to use a wall wart.
        A crap load of IP phones, IP cams, and or….. ?

    3. There would be little or no savings. The wire gauge required for the low voltage DC would be huge, and to have one centralized power supply to cover all the DC voltages required would be wasteful. If you are only using the 5V supply do you turn off the 12 or 15 or whatever other voltage you don’t need?

    1. This thing has exactly the same issues as the one I posted(see upper): the sensor is sensitive to magnetic field which comes from all the circuits around it, so you will need an algorithm to isolate just one circuit. It is extremely complicated to calibrate it properly, and once you have done that, it is very difficult to keep accurate.

      1. interesting, if this is true they certainly did a great job brushing these issues under the carpet in their paper and their talk, including the discussion uncertainty. I believe they have also started a company to distribute them with a cloud monitoring solution.

        1. It’s brushed under the carpet a bit indeed. First, they calibrate by putting known loads in the outlets: this is a good initial calibration, but very difficult to do. What about unknown loads connected to a dedicated breaker? Still, unless you can hold those sensors in place very well I don’t know how stable the measurement will be over time. Of course, it depends on what accuracy you really want. If you can put a current transformer on the main input and then use a system like this to just separate per circuit and can do with some significant % in accuracy…yes.
          I really don’t see anyone going through all this trouble to install a system like this, feels simpler with current transformer. But sure, if you have no access.

  22. There are a bunch of “commercial” electricity meters on the market. They all have benefits and pitfalls. Kill-a-watt is a great example of a simple approach and you are not going to get anything better than that without going into the panel.

    Panels are dangerous, so you do not want anyone just going in there because they can. The nice thing is that in the USA as long as the work is done to code, it does not matter who did it. So all the DIYers could probably handle it but it is definitely not a good idea to encourage it. People bring up safety and it is very difficult to design a good product that can handle this environment. Potential lightning strikes and the pure current going through these panels a ver inhospitable place for general electronics. The open source meter project is a great starting place but even they do not have the proper safety certifications and could easily be not allowed by an inspector.

    I am surprised that accuracy has not been heavily discussed yet. Most the cheap meters or monitoring solutions do not measure reactive power or power factor. If you really get into AC monitoring you will notice that you need to understand power factor and make sure you have a meter that can record volt-amps (VA) in addition to power (W). Power factor is basically how well the voltage and current waveform line up and that describes the types of loads you have on your system. Most things line up really well but the inductive loads like motors do not and that makes them need more volt-amps than watts.

    Accuracy is especially important because a 5% error in the power reading over the course of a month could add up to significant differences in your meters kWH reading and the utilities. Especially if you are in a business where they charge you for your peaks as well as your usage. Essentially making the number you are seeing and the one you are billed for much different. Even among the commercial meters that let you access this data easily, very few have any sort of accuracy claim. The good one to look for is a statement on conformity to ANSI C12.1 or C12.20. These are what the utility meters are certified to, which makes the utility legally allowed to be billed off of its reading.

    The clamp on sensors, called CTs, are the best way to get the current signal for measurement. They magnetically couple to the AC wire, much like a regular transformer. CTs can be surprisingly accurate if they are characterized well. The wireless ones are cool but you still have to install them inside the breaker panel, which is a metal box, so the signal is not going to get far.

    I have had to look into all the available options for work and honestly the one that seems to provide the best value is eGauge (http://egauge.net/). It is definitely not the most polished solution but take a look and see what you think.

  23. These sound like some awesome life hacks to save money. I’ve never really considered how much I’m spending on electricity each month that I don’t need to. My only question though is, how safe are some of these methods? The last thing I need is to have an electrical surge do damage to my home and be a safety hazard.

  24. The fact is a fair few appliances use can be detected from a simple feed of current usage ala. Currentcost meter clamp and the display unit attached via 3v uart. Aka deaggregation. Appliances such as lights/room lighting circuit without dimmer that use <5w have individual profiles that can be detected by switching them on and off a few times and looking at the change in current usage. I wrote a python script that does just this. It can detect Tv's, bathroom fans, fridge compressor etc. With a bit more work you could build a more intelligent profile per appliance I'm sure.

  25. We need a cheap, simple, modular, open platform…

    Smart-Meters actually are nothing new. Stuff like X10 was never commonly used. I think among the main reasons for this are:
    – not always retrofittable (extra wires mean lots of work)
    – uneasy setup (people can replace a fuse or maybe a switch or outlet but installing a more complex system might need a professional, even for the hobbyist tinkerer)
    – price: stuff is too expensive compared to “dumb” meters/switches/…
    – no real standard/vendor interoperability

    Also I’m surprised, that lots of modern products rely on wifi and even take an accesspoint for granted. After all security and failsafeness is some kind of issue and home routers do die from time to time ;)

    What I think is needed is an open platform:
    – It must provide powerline communication (PHY). This can be simple & low bandwidth. Optimally, low-speed TCP/IP could be used.
    – It should come in a form-factor that allows easy setup: fitting into a junction box, DIN-Rail etc. (looking & wiring like a normal switch/outlet/…)
    – enough I/O to have some fun.

    If you had e.g. an arduino/raspi… that looks like a switch, with a relay inside and some power-metering circuitry, home automation is only a simple step away. (failsafe multiway-switching, current/voltage sensing, light control, broken bulb detection) etc.

    There are several problems (size constrains, challenging PSU/circuit design, regulations, etc.) but I could imagine that a community driven approach could evolve into something useful.

    Or maybe someone did this already, and I just didn’t learn about it, yet? :)

Leave a Reply

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