Announcement

Collapse
No announcement yet.

real word test on my roof, 12 Sunpower 327 e-20 and 12 LG Neon 2 330w panels

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

  • #16
    Originally posted by solarix View Post
    Actually, the best, front row is connected to the SolarEdge... Collins.jpg
    24 - 230w Trina panels, 6 rows of 4, 12 on each inverter. Panels were purchased together all from same pallet.
    Solarix:

    Thank you for the information and the photo. I believe that clears up my confusion. Looks like aside from the first row, all other rows probably have quite similar irradiance, with wind exposure maybe a bit different for the first and last rows.

    Always good to have real data from actual field use. Truth to power sort of thing.

    I appreciate your time is valuable. If/when you get time, any info on online data availability for the array ?

    Thanx again,

    J.P.M.

    Comment


    • #17
      My production numbers for the first 3 days are 50.0 kwhr for Sunpower and 52.7 kwhr for LG/Enphase. These numbers are taken from monitor.sunpower and enlighten. My power company smart meter read 98.47 khwr and the total produced was 102.7 kwhr. The smart meter is probably the most accurate. It is a combined production for sunpower and LG/enphase.
      Last edited by Floydturbo; 03-06-2018, 02:24 AM.

      Comment


      • #18
        Originally posted by Floydturbo View Post
        My production numbers for the first 3 days are 50.0 kwhr for Sunpower and 52.7 kwhr for LG/Enphase. These numbers are taken from monitor.sunpower and enlighten. My power company smart meter read 98.47 khwr and the total produced was 102.7 kwhr. The smart meter is probably the most accurate but is total for both systems.
        Thank you. Whether the most accurate of the metering devices or not - and agreeing with you my guess is that's probably the case - since the POCO meter is the one used for billing, it is the meter against which all the others are calibrated. As best as I can figure, my Sunpower monitor reports my system output to be about 0.5% higher than the POCO smart meter would have me believe.

        That the output from each portion of your array looks mostly like a dead heat doesn't surprise me. Do you notice any difference in output, S.P. vs. LG for sunny vs. cloudy days ?

        Comment


        • #19
          A small part of the inverter generated power is lost as voltage drop in the AC wiring, so
          the PoCo meter should read slightly lower. Just what the loss is, you can measure.
          Bruce Roe

          Comment


          • #20
            Originally posted by bcroe View Post
            A small part of the inverter generated power is lost as voltage drop in the AC wiring, so
            the PoCo meter should read slightly lower. Just what the loss is, you can measure.
            Bruce Roe
            Probably, but my inverter is about 2 m from my electrical panel w/~ 16 ft. of wire with very wide conduit radii. carrying max. of ~ 20 A. I haven't measured it. I suppose it could be 0.005 of the output from the inverter but that would seem kind of high. Maybe I'm wrong on that one however.

            While vacationing I've killed all the power to the house circuits except for array input and those two circuits that power the fridge and the solar water heater pump, and fitted those devices' circuits with kill-a-watt meters. After 17 day and 20 day absences, and after some careful and immediate before leaving/on return meter readings, and some math, for one trial, I figured the Sunpower monitor was reporting 1.00522 times higher than the SDG & E meter for the first trial, and 1.00514 times higher for the second trial. The two (vacation) trials were about 9 months apart and the array was about 18 months old at the time of the first trial. A appreciate that I have no business claiming 5 significant figures, but I can claim 3 and that's where my (perhaps) slightly better than SWAG of 0.005 downward adjustment in my daily S.P. monitor output comes from. I may do a third trial some day, maybe as a method or backup to other methods I use to SWAG panel performance and thus some gauge of possible panel deterioration as f(time).

            So, regardless of whether the discrepancy is from meter variation, or unaccounted for voltage drop, or anything else, I use 0.005 output reduction in what the Sunpower monitor shows.

            As for the D.C. voltage drop between the array and the inverter, after ~ 4+years of measuring array input at the array and output as reported by the inverter display, My guess as to voltage drop is about 3.1% of array voltage under clear sky conditions. After many readings, I've pretty much convinced myself that the per deg. voltage drop is pretty close to what the Sunpower spec sheets publish.

            That 3.1% is a back calculation derived from several hundred comparisons of array voltage vs. measured individual panel array temps. using the average of the 16 panel temps. as described on this forum in the past. My SWAG on that drop per deg. C. is -1.477 V/deg. C. for 8 panels on a string (2 strings) or ~ 0.185 V/deg. C. per panel. The spec. sheet lists -0.177 V/deg. C., so I'm calcing ~ 4% higher than spec sheet but the spec. sheet number probably doesn't include anything but the panel voltage drop. My calc. also uses the spec sheet Vmpp of 54.7 V = 437.6 Vmpp.for a string of 8 panels.

            The string voltages, at least as reported by the inverter screen, seem to run pretty close to the same voltage, usually within +/- a couple % or so of one another. Perhaps interestingly but maybe not surprisingly, a lot of the variation in string voltage (or lack of it), one string to the other seems to be due to the wind vector. The array is a 4 X 4 in landscape with sting 1 being the 8 lower panels and string 2 the 8 upper panels. A north wind increases the upper string voltage relative to the lower sting voltage indicating a lower string temp. A south wind tends to increase the lower string voltage relative to the upper screen. An east or west wind tends to equalize the two voltages, and strong and sustained winds from most any direction will also tend to make the voltages, and the panel temps. close(r) to one another.

            Comment


            • #21
              J. P. M., I would not expect 3.1% DC loss for your situation. I am running 2 450 foot loops of 6 ga to
              the inverters (at nominally 400VDC), plus up to 250 feet more of 10 gauge to individual strings. This
              gets me a little over 1%, which varies quite a bit with position of the sun and temp.

              I think you are estimating the difference between panel voltage and inverter voltage to get lost
              voltage. But a small error in one of those will give a big error in the difference. Here I generally
              know the resistance of the wire loop, so a readout of current will allow easily calculating what the
              losses are. Throwing in a correction of the copper temp might get another decimal place. In my
              AC loop a Kelvin connection was made to measure the resistance before it went live.

              In the DC side I know the lengths and the wire gauge which allows calculating resistance, the
              inverter tells me current. A 500 foot extension could be added to one volt meter lead to directly
              measure voltage lost on each string feed, I have not done that. Bruce Roe

              Comment


              • #22
                Originally posted by bcroe View Post
                J. P. M., I would not expect 3.1% DC loss for your situation. I am running 2 450 foot loops of 6 ga to
                the inverters (at nominally 400VDC), plus up to 250 feet more of 10 gauge to individual strings. This
                gets me a little over 1%, which varies quite a bit with position of the sun and temp.

                I think you are estimating the difference between panel voltage and inverter voltage to get lost
                voltage. But a small error in one of those will give a big error in the difference. Here I generally
                know the resistance of the wire loop, so a readout of current will allow easily calculating what the
                losses are. Throwing in a correction of the copper temp might get another decimal place. In my
                AC loop a Kelvin connection was made to measure the resistance before it went live.

                In the DC side I know the lengths and the wire gauge which allows calculating resistance, the
                inverter tells me current. A 500 foot extension could be added to one volt meter lead to directly
                measure voltage lost on each string feed, I have not done that. Bruce Roe
                Understood. Thank you. I'm defining the Vmpp at the array as the spec sheet value. It may well be slightly less than that which would make the line loss a lot lower. Whether the line loss is higher or the actual Vmpp is lower and/or probably somewhat variable, the voltage as read/reported at the inverter is what I use. If the Vmpp is different, that will change the line loss but won't affect the temp. gradient of the voltage drop.

                Comment


                • #23
                  After one week my Sunpower panels produced 142 kwhr and the LG Neons with IQ6+ were 147 kwhr. I used the production CT (current transformer) data for the Envoy IQ. It appears that the energy readings from the individual IQ6+ inverters is about 8% higher that the CT. So far, the panels are making about the same power.

                  Comment


                  • #24
                    Originally posted by Floydturbo View Post
                    After one week my Sunpower panels produced 142 kwhr and the LG Neons with IQ6+ were 147 kwhr. I used the production CT (current transformer) data for the Envoy IQ. It appears that the energy readings from the individual IQ6+ inverters is about 8% higher that the CT. So far, the panels are making about the same power.
                    Thanx for the update. Looks like still pretty much of a dead heat panel wise. I'm not surprised.

                    Comment


                    • #25
                      One month totals are 528kwhr for the LG Neons and 512 kwhr for the Sunpower 327 panels.

                      Comment


                      • #26
                        Originally posted by solarix View Post
                        Actually, the best, front row is connected to the SolarEdge...
                        24 - 230w Trina panels, 6 rows of 4, 12 on each inverter. Panels were purchased together all from same pallet.
                        Thanks for the picture. From here it looks like there is no significant shading, but there probably
                        would be if raised to the most advantageous angle. Bruce Roe

                        Comment


                        • #27
                          Almost 4 months now, (6/26/18) LG at 2.52 mwhr and Sunpower at 2.47 mwhr. Statistically a draw. On Friday 6/22/18 with a 104degF day, the LG did a little better with 25.90 kwhr and Sunpower at 24.30 kwhr.

                          Comment


                          • #28
                            Originally posted by Floydturbo View Post
                            Almost 4 months now, (6/26/18) LG at 2.52 mwhr and Sunpower at 2.47 mwhr. Statistically a draw. On Friday 6/22/18 with a 104degF day, the LG did a little better with 25.90 kwhr and Sunpower at 24.30 kwhr.
                            Floyd:

                            Thanx for the info. Looks like the essentially identical performance for the 2 sets of panels continues.

                            So much for Sunpower's "most efficient" claims.

                            Comment

                            Working...
                            X