A day in the life of a net meter

Ever wonder what the energy flow through a net meter looks like when you have solar on your house?  No?  Well, if you did, take a gander at the graphs above  (click for a bigger version).  First off – if it looks complicated, in real life, it’s not.  I turn on a light switch and the light comes on, just like every other house – clouds or no clouds.  The details of the energy flow simply don’t matter in daily life.  Sometimes the meter spins backwards, sometimes forwards, and I pay (or get paid for) the net result at the end of the month.

But being the measurement geek that I am, I think it’s cool to see the ins and outs of the energy flow through a day.  And this was a “complicated” day, with clouds, and variable loads making a messy scribble on the graphs.  Here’s an annotated version:

The top graph represents household load (red) and solar generation (green).  The line graphs represent power (watts) at any given time – peaks mean more watts.  The filled in areas of similar color represent cumulative energy use or generation (kWh) as the day progresses.  You can see that the fridges were cycling on and off while we slept, and see that I made my espresso fairly late, and that we cooked pizza this evening – those are the big and little spikes in the use graph.  And you can see it was a semi-cloudy day – rather than a nice bell curve on a cloudless day, it’s got all kinds of dropouts as clouds roll in and out.

The bottom graph shows the net energy and power flow a bit better.  The dark blue line shows the net power, negative (net power use) or positive (net production).  The filled in lighter blue shows accumulating energy net use, or net generation.  At around noon, production caught up with use for the day, and we were net-zero on energy for the day.  We built up a bit of a surplus, and then it started declining again in the evening as the sun went down and the stove turned on.  By midnight we’ll probably be about even again.

After all that, all it really means in the end is that for this day, we will pay (and get paid) approximately $0.00 for our energy.  But we took an interesting path to get there!

By the way, these graphs come from the fantastic and free pvoutput.org.

8 thoughts on “A day in the life of a net meter

  1. Great stuff! What are you actually using to get all of the data into pvoutput?

    If you’re interested…
    I have an arduino interfacing via serial to my Xantrex inverter – http://pvoutput.org/intraday.jsp?sid=1066

    At the other end of the house, I have gone the openenergymonitor approach and am measuring my net home import/export, but can’t get access to the right location to put a CT clamp around my home consumption without solar – https://pachube.com/feeds/24201

    So I need to combine the two somehow – was thinking of accessing the pachube stream with the Xantrex arduino to do the calcs – but finding it complicated!


    • Hi Brad!

      Our setups seem eerily similar. :) My solar data comes from a data API provided by my inverter manufacturer, Enphase. But I also have an openenergymonitor-based Arduino using CTs to measure the mains and the solar feed. I use the household draw data from the arduino, and the solar production data from Enphase. I guess it’s a little Rube-Goldberg-y but it’s working.

      Anyway: there likely is no place to put CTs to measure just household draw – for that you need a bit of math!

      As I said, I’ve got CTs on the mains (i.e. the main breaker, out to the pole) and on the backfed breaker from the solar circuit. To compute the household load, and assuming that the mains are measured as energy in positive, energy out negative, and solar is measured as positive numbers, I just do:

      if (powerSolar > 0)
      powerHome = powerSolar + powerMains;
      powerHome = powerMains;

      The test for powerSolar > 0 is just because I thought I might possibly see some draw from the inverter at night, but in practice I don’t; I have a bit of a calibration issue and I apparently produce 20W from the moonlight all night long. ;)

  2. Eric,

    I’m currently in the last bit of waiting before my solar PV system is activated. I have been reading through your old posts about your setup and this post specifically caught my attention.

    Do you have another post where you go into more information about the data pipeline that you’ve put together? I’m trying to figure out what data I can get from my enphase directly and what I will need to add sensors to measure.

    Have you updated your setup since these posts or ideas of how you would do it differently now?



    • Hi Jon –

      With only Enphase, you’ll only be able to monitor production, although that is very easy to get into pvoutput. So to get the actual ebb & flow of energy through the mains, you’ll need something separate. (I think it’s too bad Enphase doesn’t offer this, it seems like it’d be easy to integrate into their system).

      When I wrote this post I had a custom built Arduino-based monitor, but now I’m using a pair of Brultech ECM-1240s to measure mains as well as various individual circuits. The Brultech seems like the most bang for your buck, although currently use a couple of Linux boxes (a raspberry Pi and my main home server) to suck data out of them and push them to various endpoints, so there was a little work on my end. I’m using the btmon.py script from http://lancet.mit.edu/mwall/projects/power/ so it’s fairly easy. There are other devices out there as well, as long as you get one with accessible data you could surely hook it up to pvoutput.org if you have a little scripting skills. Let me know if you have any specific questions!


      • Thanks for your advice.
        You have confirmed my concerns regarding the limitations of the Enphase gateway.
        At least I’m hoping that some of the missing information is provided on the utility net meter bill so that I can maintain a manual sheet for break-even tracking.

        It turns out that I do have a TED5000 that I had installed into my circuit breaker well before beginning the solar project. It sounds like I may want to re-locate the sensor though since, if I understand the likely wiring performed by the solar installers, in the current location it will only be measuring consumption from the utility.

        If you had only the Enphase gateway and one TED5000 MTU, what would be the most useful line to have the TED sensor collect data on? I realize that you don’t have my electrical schematic but at least in general?

        Given that I already have the TED5000 gateway, would there be value in obtaining a second or third MTU sensor component for it? If so, what lines would you suggest I place those on?



        • You probably have your TED on the mains, and your solar feed is probably on a breaker on the house side of the mains, yes? If you keep them there, you should be able to see net flow (in and/or out) of energy; that plus th eproduction data from the Enphase system should be enough to figure out production, actual home consumption, and net energy use.


          • That is exactly right. That is great news that I could leave my install as it is.

            I’m not sure how familiar you are with the TED specifically, but if you are, can you recommend any resources, scripts, or tutorials that will help me interface with the TED to extract the information?

            Also, if I was willing to spend $85 for an additional TED MTU, is there a data item that would be better to measure directly than compute?



Leave a Reply

Your email address will not be published. Required fields are marked *

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