Announcement

Collapse
No announcement yet.

Panel output loss due to temperature as well as loss due to AC to DC conversion etc?

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

  • Panel output loss due to temperature as well as loss due to AC to DC conversion etc?

    Hi all I have Panasonic 330w panels (x15). When temps are around 25C (best temp for the panels), I am seeing about 316W from each panel during peak. I am wondering if this (~4.3% efficiency loss) typical? If thats what you guys are seeing as well?

    Ive yet to see my panels overproduce. And at this rate I dont think they will?

    Today is our first day in the 90's F (~34C or so). They are producing 7% less due to the heat. Today I am seeing around 295W per panel, my particular panels say -0.258%/ 1C. So at 34C that is is today 34-25 = 9 degree over ideal temp 9 * 0.258 = ~2.3% efficiency loss. Which is a much smaller drop than what I am seeing at ~7%.

    So I have three questions.

    1) Should my base (at 25C) be higher? Or is ~4.3% efficiency loss due to AC -> DC conversion and other things typical?
    2) The heat efficiency loss, is that talking about actual panel temp? Or ambient temp? Because I imagine the actual panels are much much hotter than the ambient temp of 34C? If we do the math assuming the 7% drop, the panels themselves must be around ~52C or so (assuming efficiency loss per 1C is correct). Anyhow, what is your guy's experience with this?
    3) It seems like panels in the summer are less efficient than fall/spring, but this is made up for by having more hours of daylight? Is this your guy's experience?

    Thanks!

  • #2
    Your measurements seem reasonable.

    You're right that panels work better cold, worse hot. If you got below 25C, you'd get even more out of them. Panels love winter.

    I assume that your testing was at the brightest time of day. That is not the laboratory test condition for these panels, however. Panels are tested at something called "standard test conditions" aka STC. STC includes a specific temperature, a specific air mass, and a specific light intensity. If you increase light intensity beyond that of STC, the panel will put out more. If you get colder, more power. If you have less air mass, more power. Here's an article that gives the exact definition of STC.
    https://www.siliconsolar.com/what-ar...onditions-stc/

    So it's even possible to get more than 335 watts from your 335 watt panels in the right conditions.

    I assume that you're measuring power by the numbers on the inverter's measuring system. That means that you're seeing power out of the inverter. Some energy is lost in the cabling and more is lost in the electronics in the inverter. Typical inverters are rated at 97% efficient at their sweet spot, but less efficient at higher or lower power. So you are losing an additional 3% to 6% power due to things other than the panel.

    One last consideration is that measuring power accurately is difficult. The voltmeter and ammeter built into the inverter are not engineered for laboratory precision. You could be seeing 1% to 3% of error just due to inaccuracy in the current-sense transformer. But if you're using lab equipment, you may be getting very accurate measurements.

    You didn't describe the inverter you're using. All of the inverters that I've studied have a self-limiting maximum output power. If the panel produces more than the inverter can handle, the inverter adjusts it's operating point to get only the rated output. In other words, if you try to put 400 watts into a 300 watt inverter, you'll only get 300 watts out. People call this "clipping".

    Does this help?
    7kW Roof PV, APsystems QS1 micros, Nissan Leaf EV

    Comment


    • #3
      Originally posted by bob-n View Post
      Your measurements seem reasonable.

      You're right that panels work better cold, worse hot. If you got below 25C, you'd get even more out of them. Panels love winter.

      I assume that your testing was at the brightest time of day. That is not the laboratory test condition for these panels, however. Panels are tested at something called "standard test conditions" aka STC. STC includes a specific temperature, a specific air mass, and a specific light intensity. If you increase light intensity beyond that of STC, the panel will put out more. If you get colder, more power. If you have less air mass, more power. Here's an article that gives the exact definition of STC.
      https://www.siliconsolar.com/what-ar...onditions-stc/

      So it's even possible to get more than 335 watts from your 335 watt panels in the right conditions.

      I assume that you're measuring power by the numbers on the inverter's measuring system. That means that you're seeing power out of the inverter. Some energy is lost in the cabling and more is lost in the electronics in the inverter. Typical inverters are rated at 97% efficient at their sweet spot, but less efficient at higher or lower power. So you are losing an additional 3% to 6% power due to things other than the panel.

      One last consideration is that measuring power accurately is difficult. The voltmeter and ammeter built into the inverter are not engineered for laboratory precision. You could be seeing 1% to 3% of error just due to inaccuracy in the current-sense transformer. But if you're using lab equipment, you may be getting very accurate measurements.

      You didn't describe the inverter you're using. All of the inverters that I've studied have a self-limiting maximum output power. If the panel produces more than the inverter can handle, the inverter adjusts it's operating point to get only the rated output. In other words, if you try to put 400 watts into a 300 watt inverter, you'll only get 300 watts out. People call this "clipping".

      Does this help?
      Thanks for the reply. The inverters are Enphase IQ7X, one under each panel. Two trunks run into an Enphase IQ combiner, then to my breaker box. The measurements I am talking about are all from Enphase's tracking system.

      One thing I should mention is that weather app did have a "Bad Air Quality" advisory today, so I am not sure if like smog in the air would potentially reduce production even though to the naked eye it looks like a perfectly clear sky.
      Last edited by Duxa; 04-24-2020, 10:21 PM.

      Comment


      • #4
        [QUOTE=Duxa;n414005]

        The measurements I am talking about are all from Enphase's tracking system.



        Is this a physical tracker? Sounds like a catchy name for data collection. Reason I ask is that unless your panels are oriented exactly facing the sun your current will be lower than it could be. Typically this can only be achieved with a dual axis tracker.

        2.2kw Suntech mono, Classic 200, NEW Trace SW4024

        Comment


        • #5
          [QUOTE=littleharbor;n414006]
          Originally posted by Duxa View Post

          The measurements I am talking about are all from Enphase's tracking system.



          Is this a physical tracker? Sounds like a catchy name for data collection. Reason I ask is that unless your panels are oriented exactly facing the sun your current will be lower than it could be. Typically this can only be achieved with a dual axis tracker.
          Woops, I guess poor choice of words on my part. All I meant to say is that I use the data that is sent from the microinverters to their monitoring website. This here in case you havent seen it before - https://www.youtube.com/watch?v=d4xMdWOZDBk

          and yeah you are right, my panels are stationary... I totally forgot that the 330W rating is for perfect orientation....

          Comment


          • #6
            [QUOTE=Duxa;n414007]
            Originally posted by littleharbor View Post

            I totally forgot that the 330W rating is for perfect orientation....

            Actually that 330 watt rating is for "Standard Test Conditions" which are at 25 degrees Celsius under a 1000 watt per square meter test light. Real world conditions rarely match this although, at times, conditions can cause panels to exceed their rated output. Typical solar panel output will vary throughout the day, as well as the time of year. Some people can get their best performance in the winter although not necessarily best total output.
            2.2kw Suntech mono, Classic 200, NEW Trace SW4024

            Comment


            • #7
              Quick answer to your questions:
              1.) Your "Base" for comparing panel % efficiency measurements should be 25 C. for the cell temp. That's the cell temp. when STC measurements are taken. It has no relation to ambient temp. or cell temp. under operating conditions.What you're probably looking for is a representative instantaneous (snapshot in time) array cell temp. to use for array output efficiency estimates.
              2.) The cell efficiency loss (decrease) is measured either as a % loss of STC efficiency per deg. C. or a per W per panel loss per deg. C. The temp, diff. used is the diff. between the operating cell temp. and the STC cell temp. (25 C.)
              3.) My array's historical output as measured by output per day per installed STC kW is highest for the month of June. Second highest period is from 04/18 to 05/10. Longer days help June some, but the lower amb. temps in spring make that time of year almost as productive. Clearness index for both periods is about equal so that tends to not be a factor.

              If you, or anyone reading this mental spoor is curious or want more particulars, read on.
              What follows is mostly applicable for working with arrays under bright sun at high solar incidence angles.

              1.) When the sun is shining, your panels will operate at a temp. above the ambient temp. The panel cell operating temp. can be estimated in several ways. The standard ways to estimate the cell temp. start with an energy balance on a panel (or cell, or array). For that you will need to know or estimate the P.O.A. (Plane of Array) irradiance. For a reasonably accurate estimate of irradiance you will need something called a pyranometer.

              An energy balance on an element (a panel for example) accounts for all the energy that comes into, goes out of and gets stored in, or released from the element. Assuming what's called "steady state" conditions, where no energy is stored in or released from the panel (and, the panel temp. stays the same), and things don't change much from minute to minute for a few minutes (such as around solar noon on a cloudless, windless day, for example), energy input == energy output.

              Under such conditions, for all practical purposes for rooftop arrays anyway, the input is all the solar radiation (irradiance). The output is divvied up four ways:
              a.) Electrical output to the grid, load or batteries.
              b.) Thermal losses via thermal convection and thermal radiation to the ambient environment.
              c.) Reflection losses from the glazing as f(angle of incidence of the irradiance).
              d.) A small amount of thermal conduction loss to the array racking, which can usually be ignored.

              The sum of those losses (or output) will, by definition, under steady state conditions, equal the input (the solar P.O.A. irradiance). The task can be somewhat simplified with what is usually an acceptable loss of accuracy by ignoring the fourth term and combing the second and third terms into one term that is the rate of energy transfer from an array per deg. C of temp. diff. between the array's ave. cell temp. and the ambient environment air temp. at the array.
              I'll cut to the chase for this: At solar angles of incidence ~ < 30 deg. or so, h = the combined convection, radiation and reflection loss term = (5.38 * W) + 24.4 where W == the wind velocity at the array in m/sec. The units of "h" are W/(m^2*deg. C). That correlation compares well with lots of heat transfer correlations for flow over a flat plate, and also with what I've measured several hundred times by matching and curve fitting with several hundred array measurements I've done over the last 6 1/2 years. Long, boring story.

              Your goal is to first find, or estimate, an average cell temp. If you know from a pyranometer or can reasonably estimate the P.O.A. irradiance and the average wind speed at the array, a couple of highly empirical correlations may serve as a 1st rough estimate of the cell temp.

              a.) Under bright sun at near normal solar angle of incidence: Cell temp. ~ = (0.943 * T,amb.) + (0.028* I) - (1.528 *Wind) + 4.3

              (For origin, Google: "NREL/CD-520-33586")

              Where: T, amb = ambient temp., deg. C
              I = P.O.A. irradiance , W/m^2
              Wind = wind velocity, m/sec.

              So, w/an air temp. of, say 20 C, I = 800 W/m^2 and wind = 1 m/sec.:

              Cell temp. ~ (0.943*20)+(0.028*800) - (1.528*1.0) = (18.86) + (22.4) - (1.528) + 4.3 = 44.0 C.

              Which, BTW, given that those input parameters are close yo the measurement conditions for NOCT, that 44 C may be close to your panel's published NOCT (Normal Operating Circuit Temp.). Check it out if your curious.

              Similarly, assuming your conditions of 34 C. ambient, 1,000 W/m^2 P.O.A and, say, a 2 m/sec. wind velocity (?) :
              Cell temp. ~ (0.943*34) + (0.028**1,000) - (1.528*2) + 4.3) ~ = ~ 61.3 C.

              Coincidentally and no more than anecdotally, I'm not too far away from you in 92026 and on 04/24, under what are probably similar conditions to yours, with a clean array, I calc'd my array's cell temp. from a voltage correlation with cell temp. I derived to be ~ 63.0 C. at the time of minimum angle of incidence on my array. A slightly modified correlation similar to the above using different constants specific to my array gave 63.35 C.

              For an efficiency loss as f(cell temp) for your panels at your approx. conditions: % loss ~ (0.258*(61.3-25) ~ = 9.4% lower than the STC eff. That 7 % you use is probably as good as any number, especially since it seems that today's irradiance was probably ~ 3 % above STC irradiance from noon til ~ 1400 hrs. solar time, your wind velocity, P.O.A irradiance and amb. temp. on the roof is unknown, and also, your panels are relatively new and perhaps not completely burned in yet so they're operating at maybe a bit higher than STC eff. at this time.

              b.) Another empirical correlation from Sandia labs that I've found to give pretty good match to what I've measured:
              Cell temp ~ = {EXP[(-0.054*wind)-3.56]}*I + [(I * 3)/1,000] + T,ambient

              Where: Wind = wind velocity, m/sec.
              I = P.O.A. = Plane of Array irradiance as above, W/m^2
              T, ambient = ambient air temp. at the array.

              Using your numbers, and the same (estimated) wind speed of 2 m/sec. and 1,000 W/m^2 gives:

              {EXP[(-0.054*2.0)-3.56]}*1,000 + [(1,000*3)/1000] + 34 ~ = 62.5 C.

              I'd suggest the agreement between the two, 61.3 C, vs. 62.5 C. is reasonable.

              2.) The efficiency loss is as a % of the STC efficiency measured from the STC temp. of 25 C. Another way to calc it is as watts lost per deg. C. of cell temp. increase. That is as -0.258 % of 330 W (= -0.00258*330) = -0.85 W/panel per deg. C cell temp. increase. SOOOO, if you are measuring per panel output of 295 W on data from 04/24/2020, and you want a rough est. of cell temp.:

              330 W - 295 W = 35 W.
              35 W/(0.85 W/deg.*deg.C) = 41.2 C above STC temp.
              STC temp. = 25 C.
              So, Approx. cell temp. = 41.2 + 25 = 67.2 C.
              Believe it or not, that's probably not bad agreement between empirical correlations and estimates using voltage drop correlations. Also, an array's temp. varies by ~ 3 - 8 C. or so over the surface of an array due to wind effects, lower temps at the leading edges (upwind) as I've measured and any heat transfer text will explain.

              Still, I wouldn't be too surprised if the coef. of power was closer to the -0.30%/degree C as published for the Panasonic VBHN330SA16 panel in 2015 as different from the -0.258%/deg.C as published on the 2017 spec shhet of the same model #. If so, that would give a cell temp. of ~ 60.4 C.

              Using that -0.30% temp. coeff. of power, and averaging the results of the three correlations gives a cell temp. of (61.3 + 62.5 + 60.4)/3 = 61.4 C.

              3.) Depending on array orientation, spring and summer around here and in many parts of the U.S. are the best seasons. Since you ask, over the last 6 1/2 years, my array's most productive time in terms of kW/per day per installed STC kW is the month of June with the ave. daily output of 6.07 kWh/STC kW.. The period from 04/18 to 05/10 is the second best period with a daily average output of 5.73 kWh/STC kW. Summer days are longer, but the warmer ambient temps. reduce system efficiency. Spring temps. are also a bit lower making for lower atmos. dew points which tends to make for less H2O scattering of the beam irradiance which increases overall array irradiance.

              Take what you want of the above. Scrap the rest.
              Last edited by J.P.M.; 04-25-2020, 03:58 PM.

              Comment


              • #8
                JPM: Super-great analysis and explanation. Thank you very much!
                7kW Roof PV, APsystems QS1 micros, Nissan Leaf EV

                Comment


                • #9
                  Originally posted by bob-n View Post
                  JPM: Super-great analysis and explanation. Thank you very much!
                  You're welcome. Opinions vary.

                  Like our service ? Tell your friends.

                  Comment

                  Working...
                  X