Announcement

Collapse
No announcement yet.

Solar Power monitoring system calculations

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Solar Power monitoring system calculations

    I have recently installed a solar edge monitoring system through Vivint Solar on my home and have received access to the SolarEdge website to look at the data. I have concerns about how the power production is calculated. When I calculate the energy production from the raw numbers provided, I consistently get an average of 3% lower numbers than the daily production solar edge reports when I run a simple trapezoidal integration routine on the instantaneous reported production.
    https://docs.google.com/spreadsheets...gid=1495326679

    I have even tried some spline interpolation routines to reduce the numerical error, but still get a large over reported error. When I run the same calculations on friends data that I have received, I get less than a 0.1% error. I'd be happy to run similar calculations for others to see if you have issues. It isn't a big error, but it does seem somewhat fishy that Vivint or SolarEdge is telling me that they are producing more energy than they actually are. I have purchased my system from Vivint instead of renting the system and buying the power, but it does make me wonder if Vivint is charging people more money somewhere with this little math bug. However, before I go down that path, I'd like to know if a.) my math is correct and b.) if anyone else has a similar issue.

    Thanks,

    Matt








  • #2
    I lose more than 3% energy in the wires connecting everything together. Have you a way to include that in your calculations? Bruce Roe

    Comment


    • #3
      I would believe that Vivint tried to account for this if the loss was always 3%, but it fluctuates every day. Here is the histogram of the error between what I calucate for energy production and what SolarEdge reports.

      https://docs.google.com/spreadsheets...at=interactive

      The range is between -9% to a +0.6% but the average is 3%.

      Like I said before, the % error in all other data I get from friends is less than 0.01%.

      Thoughts?

      Matt

      Comment


      • #4
        Originally posted by Assumption View Post
        I have purchased my system from Vivint instead of renting the system and buying the power, but it does make me wonder if Vivint is charging people more money somewhere with this little math bug.
        The kwh reported by solaredge inverter is only accurate to +/- 5%.
        Unless you pay for the upgrade to a more accurate meter. (A "revenue grade" meter which is at least within 2% and probably even better than that.)

        For people that lease I think the installer typically will install a separate meter socket and a separate meter. And that meter will be within 2%, and probably even more accurate than that. It'll be basically same quality as what the POCO uses for billing you.


        Comment


        • #5
          Originally posted by Assumption View Post
          I would believe that Vivint tried to account for this if the loss was always 3%, but it fluctuates every day. Here is the histogram of the error between what I calucate for energy production and what SolarEdge reports.

          https://docs.google.com/spreadsheets...at=interactive

          The range is between -9% to a +0.6% but the average is 3%.

          Like I said before, the % error in all other data I get from friends is less than 0.01%.

          Thoughts?

          Matt
          Yea, 0.01% ( 1 part in 10,000) seems kind of tight. 1.0% would be tighter than most.

          Comment


          • #6
            It would help if posters could site their location, grid tied or not, utility company, governing agencies etc. to help respond.
            In NJ where I am, the inverter based meter is not authorized for reporting kwh.
            A separate revenue grade meter is required by NJ BPU and Atlantic City Electric.
            Accuracy required varies depending on load and power factor.

            http://www.njcleanenergy.com/renewab...projects-srecs
            http://www.dps.ny.gov/Part92-Operating-Manual.pdf
            Dennis
            SE5000 18 each SW185

            Comment


            • #7
              A "revenue grade" meter is the only way I would really trust. Several of those are used here. Bruce Roe

              Comment


              • #8
                I use an E-Mon revenue grade sub meter which did not require rewiring my solar setup. This meter uses clamp on CT's but is very accurate. Had to install revenue grade meter after the fact to sell SREC's to Ohio and Illinois. Meter is digital and constantly flashes the exact production from my inverters every two seconds. I can look at the inverter screens and know exactly the difference between the inverter calculations of watts produced and the actual watts going out the line. Then I compare those two calcs with with my TED system calcs. I have adjusted the software in the TED system to match the E-Mon meter so that my reporting to PVOutput is accurate.

                http://www.emon.com/en/products/mete...e-kwh-submeter

                Comment


                • #9
                  Originally posted by dennis461 View Post
                  In NJ where I am, the inverter based meter is not authorized for reporting kwh.
                  I think that's only required for getting SRECs. (And probably required for leased systems)
                  So you can still look at the inverter's data and you can report it on forums like here or report it on pvoutput.org

                  Comment

                  Working...
                  X