• Filter
  • Time
  • Show
Clear All
new posts

  • #16
    To answer a different question, I see virtually the same output per panel, regardless of location. I'm using APsystems microinverters and their apps, but it gives roughly the same amount of info as the Enphase tools.

    Here's a snapshot from late morning on a sunny day, with all panels producing roughly 300 watts. You could argue that the 318-watts fromn one panel is significantly more than the rest, but I'm not sure of the measurement accuracy inside the microinverters. Without doing some controlled measurements, I don't trust them to better than 5%.
    Attached Files
    Last edited by bob-n; 03-20-2020, 02:47 PM.


    • #17
      Originally posted by bob-n View Post
      White paint seems like the next experiment. How can I devise a controlled experiment to tell me how much better gloss white paint would be than my current dull off-white stained wood?
      As I mentioned earlier, if you can see individual panel output on Enphase you can try a gloss white. If you have a light weight panel you can put up there on a sunny day to see if there is any instantaneous increase. Otherwise it might be hard to account for day to day differences for an expected benefit of 5%.
      9 kW solar. Driving EVs since 2012


      • #18
        Originally posted by bob-n View Post
        Bruce had a great thought. I never considered putting panels vertical on that wall.

        I just ran PVWatts two ways. With the same ideal assumptions and my location:
        East at 30 degrees gives 4236 kWh per year - My current array configuration.
        South at 90 degrees gives 3824 kWh per year - Vertical panels facing south.
        Not much difference, but the current array placement is slightly better than a vertical array on that south wall, so I'm not going to move my panels.
        It's not clear whether I face E or ESE, but that's a minor difference, and any additional south would benefit the current array over a vertical placement.

        White paint seems like the next experiment. How can I devise a controlled experiment to tell me how much better gloss white paint would be than my current dull off-white stained wood?
        Given our latitude, I'm not surprised at the similarity in outputs.
        However, one question: Were the arrays you ran the same STC size ?

        On the idea of glossy paint being a better reflector than matte finish:

        If you paint a non specular surface such as wood siding or any surface that doesn't produce a reflection - such as slate/brick/fibreboard, etc., the reflectivity difference between glossy paint and flat paint of the same color will not be noticeable. Silver surfacing a rough surface won't produce a specular reflection but the reflectivity will be about the same - maybe a tad - a % or 2 less - than the highest available specular reflective surface - a piece of glass that's front silvered, i.e., a front silvered (but to be clear, the silver exposed to the air will begin to tarnish almost immediately). In the case of the rough surfaces, the reflectivity will be diffuse but still reflective of the irradiance hitting it. A surface does not need to be a specular reflector to be highly reflective of light.

        However, since your asking: One possible experiment:
        1.) Do this on consecutive, sunny days.
        2.) Measure/record panel/array output for the day.
        3.) Paint the side of the house next to the array flat white.
        4.) Measure the panel/array output as above.
        5.) Paint the side of the house next to the array gloss white.
        6.) Measure record the panel/array output as above.
        7.) Compare the results for the 3 days.

        You'll have the best chances of less daily output variation due to diurnal changes solar zenith angle if the above is done near(er) the solstices, but doing so at the solstices and equinoxes will give you a SWAG of annual output changes.

        The above method is probably not practical, realistic or even possible for any number of reasons, but my opinion is it will give you a decent chance of getting some 1st approx. numbers for differences in the reflective value of different surfacesgt.

        But before I did all that , or anything else, and for as easy it is to get a pretty hard number, I'd make sure of, or at least get closer to your array's and other surfaces actual azimuth.

        If you're interested in your array's orientation (and also the walls' orientations) with respect to azimuth, shoot an azimuth (look at shadows cast) at solar noon and guesstimate how far off south the array is. Get a reasonable first approx. of solar noon as halfway between sunrise and sunset. Solar noon and civil noon only agree (at most) 4 times/year.
        Or, see if the vendor has the info used to estimate array output.
        Or use Google maps.
        Or a phone app.

        Why array and surface azimuth are important for this application: POA irradiance and insolation on devices and surfaces and so outputs from devices that are more vertically oriented are more sensitive to azimuth than devices that have a closer to horizontal tilt. For this application however, my guess is that shading from the wall will reduce output and reflection from the wall will help, but the reflection off the wall won't help as much as the shading by the wall hurts.
        Last edited by J.P.M.; 03-21-2020, 01:25 AM.


        • #19
          I believe you will find that your PV panel manufacturer discourages adding reflectors and even prohibits it in the warranty. PV panels are not designed to handle more than one sun of radiance and the back sheet can get damaged from excess temps, the internal wiring and solder joints can come undone from excess amps, and the cells themselves can become scorched and discolored.
          I realize that many people in the solar game are independent minded cusses, but really, do try to color within the lines....
          BSEE, R11, NABCEP, Chevy BoltEV, >2500kW installed


          • #20
            I just checked and the warranty says not to add reflectors. Solarix is right. Oh well.