Announcement

Collapse
No announcement yet.

New System - PVOutput/Monitoring Questions

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

  • #16
    Originally posted by sensij View Post
    1) The insolation curve is a clear-sky model of what your system might produce. It takes into account your location, array rating, and panel orientation. If you included a temperature coefficient in your system setup, and provide temperature data from Wunderground (or wherever), it makes an adjustment for temperature. However, this adjustment is poorly implemented and not very accurate. The insolation curve is a nice way to roughly see how your system might perform with an ideal clear sky, useful (in my opinion) for estimating the consequences of shade. If you want more accurate modeling, it will take a lot more work. But yes, temperature effects are a big reason why the curves do not match in the middle of the day. When we get into winter, some of the other weaknesses of the "insolation" model will be apparent.

    2) Panel data is accessible through the SolarEdge portal. If your installer gave you "reports" access, you can see the historical data. Figuring out year to year degradation is difficult... changes in environmental conditions are responsible for larger swings in production, making the degradation a signal that is hard to tease out from the noise. Simple fouling (dust, etc) on the array is probably larger in magnitude at first. Over enough enough time, the degradation will be easier to pick up, but if you are really interested, you may want to invest some time to learn more about PV modeling and ways to measure and account for the other inputs into the system.

    3) I think the screen has been changed, I can't figure out how to get the details on the bottom now either. In the upper right of the layout view, there is a "information" button that shows you the data from the selected panel, and a "charts" button that let's you see the optimizer data over time. If you don't see those, it is possible your installer did not give full report access, which is kind of silly.

    4) Panel status updates at random intervals (by design), and the data freshness of any one panel may vary between 1-15 minutes, typically, maybe more for larger systems. Summing up all the panels is not giving an accurate measurement of power, since the data are not from the same time and conditions change (the sun moves, if nothing else). The inverter data updates at more regular intervals. There are also calibration differences that exist as well, but yes, since the optimizers are producing DC and are unaware of the inverter efficiency, that should also be considered.


    For me, the panel data is not so useful in an absolute sense, but relationships between panels should be pretty consistent over time. If one panel typically outperforms another on a clear day, it should stay that way within some margin that accounts for difference in cooling that may occur from changes in wind direction, etc. Comparing "instantaneous" readings is almost useless, since there is no control over freshness. Comparing daily totals, especially on clear days, is more informative. If you take a series of panel voltage and current measurements, plot them into a curve (as the charts view does), and compare those, some insight can be gained there as well.
    I'm mostly ignorant about how SolarEdge works. On the temperature correction, does it use the given (supplied) temp. as the panel (array) temp. or does it adjust/estimate a panel/array temp. from the given ambient temp./wind/irradiance ?

    Thanx.

    Comment


    • #17
      Originally posted by J.P.M. View Post

      I'm mostly ignorant about how SolarEdge works. On the temperature correction, does it use the given (supplied) temp. as the panel (array) temp. or does it adjust/estimate a panel/array temp. from the given ambient temp./wind/irradiance ?

      Thanx.
      SolarEdge does not have an estimator that I'm aware of, although an installer can supply estimated values from any source for the sake of monthly comparisons.

      I think NYHeel's questions were about the "Insolation" view offered by PVOutput.org. It has been discussed a few times in the forum, but I'm not going to go digging though a search engine at the moment for links. They are simply taking the ambient temperature and using it to adjust the expect output by simple proportion based on the coefficient entered by the user. In addition to the substantive omissions to their model that you've questioned, they also incorrectly apply the coefficient as a percent of power estimated, instead of the more linear V/deg or W/deg relationship.

      Again, as a rough cut of what the unshaded generation curve might look like for a given location and array orientation, I think it isn't bad. As a finely tuned model to evaluate array performance, it stumbles (to put it gently).

      CS6P-260P/SE3000 - http://tiny.cc/ed5ozx

      Comment


      • ButchDeal
        ButchDeal commented
        Editing a comment
        It is a nice ruler upon which to reference. It also handles two arrays nicely (but only two, more can be done clumsily with groups ). I generally put in the two largest arrays and associated any other modules with the closest array. This gives reasonable results.
        I also find that compass headings are often off (or used magnetic instead of true). But one can easily adjust the array heading based on the insulation curves.

    • #18
      Originally posted by sensij View Post
      Unfortunately, you can't get optimizer data through the API, or change the level of access you have in the portal. You'll need your installer to bump up your portal access level, and I see no reason why they wouldn't do that for you.
      Why do SolarEdge owners have to go through the installer to get better portal data access? The system doesn't belong to the installer, it is the property of the owner.

      I have a SolarEdge system and my installer hasn't responded to a request for greater access to the data and API key for auto uploading to PVOutPut. It's very frustrating.

      SolarEdge needs to step up in these cases where the installer isn't interested in doing anything else for their clients who rightfully should have full access to the data being collected by their inverter. When I signed the contract with the installer, I definitely didn't sign anything that said the installer had sole discretion regarding how much data they would make available to me via the SolarEdge monitoring portal.

      I hope SolarEdge is reading these forums and notices this post, and takes steps to correct the problem it has with data access by system owners.

      Comment


      • ButchDeal
        ButchDeal commented
        Editing a comment
        try opening a case with SolarEdge support, you will need your monitoring site name, and the serial number from the inverter.
        send a note to support@solaredge.com

      • ButchDeal
        ButchDeal commented
        Editing a comment
        send a note to support @ solaredge. com
        you will need your monitoring site ID and the serial number of your inverter

    • #19
      Originally posted by sensij View Post

      SolarEdge does not have an estimator that I'm aware of, although an installer can supply estimated values from any source for the sake of monthly comparisons.

      I think NYHeel's questions were about the "Insolation" view offered by PVOutput.org. It has been discussed a few times in the forum, but I'm not going to go digging though a search engine at the moment for links. They are simply taking the ambient temperature and using it to adjust the expect output by simple proportion based on the coefficient entered by the user. In addition to the substantive omissions to their model that you've questioned, they also incorrectly apply the coefficient as a percent of power estimated, instead of the more linear V/deg or W/deg relationship.

      Again, as a rough cut of what the unshaded generation curve might look like for a given location and array orientation, I think it isn't bad. As a finely tuned model to evaluate array performance, it stumbles (to put it gently).
      Understood. Thanx Sensij/Butch.

      Comment


      • #20
        So I've been tracking my system's performance pretty closely for the last 2 months and I'm trying to learn as much as possible about what impacts performance. What I've noticed on PVOutput is that my system typically outperforms the PVOutput insolation curve in the morning but then falls behind starting around 10:30 AM, even on a fully sunny day. I have a 10.85 kW system on a 10 kW SolarEdge inverter. What I've noticed is that on a sunny day the system maxes out at around 8.2 or 8.3 kW. However, when there are occasional clouds and I get downward spikes in performance from the cloud cover, there are higher highs between the clouds, even sometimes before the downward spikes. I can often tell when a downward spike is coming because it's sometimes preceded by an uptick. I've gotten as high as slightly over 10 kW in between clouds and can get performance of about 9-9.5 kW for about 15-20 minutes between the clouds.

        Do you guys know the reason for the better performance before and (more often) after clouds? Is it just imperfect data from the too frequent reporting (every 5 minutes) or is there an underlying reason for performance to uptick between clouds? Are the panels a little cooler after getting a small break from the direct sun? Also, should I be at all concerned about my mid day performance being consistently under the insolation curve on PVOutput (when there aren't cloud spikes) or is the insolation curve just wrong? I have no midday shading and I did put my 2 arrays into PVOutput based on the measurements my installer gave me for azimuth and tilt.

        If any of you want to take a look, my system name on PVOutput is BergenSolar. (I did have to have my system turned off for a few days due to waiting on POCO inspection during that time.) 6/25 has been my only perfectly sunny day with no spikes. Today's graph is actually showing some weird results so far. I've been a little below PVWatts (set with 10% loss and the local SolarAnywhere weather data) but I know that can vary based on actual weather. The only thing that worries me a tiny bit is that my maximum daily production thus far has been 64-65 kWh yet PVWatts shows 13 days during that time period with results above 65 kWh. Should that worry me?
        Last edited by NYHeel; 08-09-2016, 12:21 PM.

        Comment


        • #21
          Originally posted by NYHeel View Post
          So I've been tracking my system's performance pretty closely for the last 2 months and I'm trying to learn as much as possible about what impacts performance. What I've noticed on PVOutput is that my system typically outperforms the PVOutput insolation curve in the morning but then falls behind starting around 10:30 AM, even on a fully sunny day. I have a 10.85 kW system on a 10 kW SolarEdge inverter. What I've noticed is that on a sunny day the system maxes out at around 8.2 or 8.3 kW. However, when there are occasional clouds and I get downward spikes in performance from the cloud cover, there are higher highs between the clouds, even sometimes before the downward spikes. I can often tell when a downward spike is coming because it's sometimes preceded by an uptick. I've gotten as high as slightly over 10 kW in between clouds and can get performance of about 9-9.5 kW for about 15-20 minutes between the clouds.

          Do you guys know the reason for the better performance before and (more often) after clouds? Is it just imperfect data from the too frequent reporting (every 5 minutes) or is there an underlying reason for performance to uptick between clouds? Are the panels a little cooler after getting a small break from the direct sun? Also, should I be at all concerned about my mid day performance being consistently under the insolation curve on PVOutput (when there aren't cloud spikes) or is the insolation curve just wrong? I have no midday shading and I did put my 2 arrays into PVOutput based on the measurements my installer gave me for azimuth and tilt.

          If any of you want to take a look, my system name on PVOutput is BergenSolar. (I did have to have my system turned off for a few days due to waiting on POCO inspection during that time.) 6/25 has been my only perfectly sunny day with no spikes. Today's graph is actually showing some weird results so far. I've been a little below PVWatts (set with 10% loss and the local SolarAnywhere weather data) but I know that can vary based on actual weather. The only thing that worries me a tiny bit is that my maximum daily production thus far has been 64-65 kWh yet PVWatts shows 13 days during that time period with results above 65 kWh. Should that worry me?
          Probably the biggest reason for the spikes, both up and down on pt. cloudy days is most likely due to clouds. Upward spikes can be due to forward reflection of irradiance by clouds. Sharp downward spikes can be due to fast moving clouds with sharp edges that are in the process of interceding between your array and the sun. Sometimes the two effects happen over a relatively short time period making the % chance over such a short time period more drastic.

          As for PVWatts daily output, comparisons between actual output from your array and the historical day from the TMY data that PVWatts uses is meaningless. PVWatts usually is set to use TMY 2 or TMY3 data, which is historical. Comparing actual data for any current day to PVWatts data for the same day is like expecting today's weather to be identical to the same calendar day in prior years.

          Also, high irradiance after a cloudy period of, say 1-2 hrs. will benefit the high irradiance output from the temporary lower array temp. that was due to the extended cloudy period, further boosting the temporary output spike.

          Also, I don't know where or how PVOutput gets its insolation data, and not a knock, but I'd not put a high amount of faith in its accuracy, particularly on pt. cloudy days.

          Normally, a well designed, mostly equator facing array operating at 20-30 deg. above amb. air temp. in the summertime near solar noon will produce, very roughly about 80 - 85% of the nameplate output rating.

          If you've not cleaned the array since installation, expect a several % or more penalty in performance due to accumulated dirt.

          Comment


          • #22
            Originally posted by J.P.M. View Post

            Probably the biggest reason for the spikes, both up and down on pt. cloudy days is most likely due to clouds. Upward spikes can be due to forward reflection of irradiance by clouds. Sharp downward spikes can be due to fast moving clouds with sharp edges that are in the process of interceding between your array and the sun. Sometimes the two effects happen over a relatively short time period making the % chance over such a short time period more drastic.

            As for PVWatts daily output, comparisons between actual output from your array and the historical day from the TMY data that PVWatts uses is meaningless. PVWatts usually is set to use TMY 2 or TMY3 data, which is historical. Comparing actual data for any current day to PVWatts data for the same day is like expecting today's weather to be identical to the same calendar day in prior years.

            Also, high irradiance after a cloudy period of, say 1-2 hrs. will benefit the high irradiance output from the temporary lower array temp. that was due to the extended cloudy period, further boosting the temporary output spike.

            Also, I don't know where or how PVOutput gets its insolation data, and not a knock, but I'd not put a high amount of faith in its accuracy, particularly on pt. cloudy days.

            Normally, a well designed, mostly equator facing array operating at 20-30 deg. above amb. air temp. in the summertime near solar noon will produce, very roughly about 80 - 85% of the nameplate output rating.

            If you've not cleaned the array since installation, expect a several % or more penalty in performance due to accumulated dirt.
            Thanks, that's very helpful and interesting.

            Also, I understand that comparing a specific day to PVWatts is pointless. I was looking at the PVWatts daily output to see what a perfect or really good day would produce. I saw that there were 10 days between June 15th and July 23rd that PVWatts had over 66 kWh. I've had zero days of 66 kWh or above since I turned my system on June 15th. Of course it's hard to determine the reason for that, if it truly is more than just a fluke. I do have a small amount of late day shading so that could also be a factor. I was just slightly concerned because I seem to be producing a little below the peak in the middle of perfectly sunny days. But it's hard to say for sure without more data.

            Comment


            • #23
              Originally posted by NYHeel View Post
              Thanks, that's very helpful and interesting.

              Also, I understand that comparing a specific day to PVWatts is pointless. I was looking at the PVWatts daily output to see what a perfect or really good day would produce. I saw that there were 10 days between June 15th and July 23rd that PVWatts had over 66 kWh. I've had zero days of 66 kWh or above since I turned my system on June 15th. Of course it's hard to determine the reason for that, if it truly is more than just a fluke. I do have a small amount of late day shading so that could also be a factor. I was just slightly concerned because I seem to be producing a little below the peak in the middle of perfectly sunny days. But it's hard to say for sure without more data.
              You're welcome. Understood. How close did yoiur system get, on average ? Are the azimuth & tilt correct vs. actual ? AC to DC ratio ?, etc. This is not an exact science. What were the temp. & wind like for actual vs. the values PVWatts used ?

              Comment


              • #24
                Originally posted by J.P.M. View Post

                You're welcome. Understood. How close did yoiur system get, on average ? Are the azimuth & tilt correct vs. actual ? AC to DC ratio ?, etc. This is not an exact science. What were the temp. & wind like for actual vs. the values PVWatts used ?
                I'm not sure how to see the weather data PVWatts uses.

                As an example, the last three days have been interesting. Sunday evening brought some rain and then on Monday, we had the first clear day in quite awhile and we had relatively low temperatures. We had a high of 79 and my system produced 63 kWh for the day (10.85 kW system). Peak production was about 8.31 kW and the best hour produced 8.27 kWh. There was one downward spike but it was only one 5 minute increment. Yesterday, on Tuesday, was very similar but with no downward spike. The system produced 62 kWh and peak production was a little lower. The high power amount for the day was 8.08 kW and the best hour period produced 8.01 kWh so production was starting to decline from Monday. Today so far has shown another dropoff. Still a perfectly clear day but each period is consistently worse than the prior day. Peak power production is at 7.79 kW and the best hour so far is 7.74 kWh. Temperature is a little hotter than the last two days with today being 78 right now (at peak production time) vs. 74 and 73 the last two days.

                So what causes the dropoff of 3-4% each day? It almost seems like the equipment is getting tired from operating at the max each day. Could it be the sun moving each day now that days are getting shorter? But even at the pretty fast pace, days are ending only 1.5 minutes earlier each day. I get that it's not a major change but the movement on consecutive days has me wondering what the cause is. Also, I don't think there's anything wrong with my system. I'm just trying to get a better grasp on how production is impacted.

                Comment


                • #25
                  Originally posted by NYHeel View Post
                  I'm not sure how to see the weather data PVWatts uses.

                  As an example, the last three days have been interesting. Sunday evening brought some rain and then on Monday, we had the first clear day in quite awhile and we had relatively low temperatures. We had a high of 79 and my system produced 63 kWh for the day (10.85 kW system). Peak production was about 8.31 kW and the best hour produced 8.27 kWh. There was one downward spike but it was only one 5 minute increment. Yesterday, on Tuesday, was very similar but with no downward spike. The system produced 62 kWh and peak production was a little lower. The high power amount for the day was 8.08 kW and the best hour period produced 8.01 kWh so production was starting to decline from Monday. Today so far has shown another dropoff. Still a perfectly clear day but each period is consistently worse than the prior day. Peak power production is at 7.79 kW and the best hour so far is 7.74 kWh. Temperature is a little hotter than the last two days with today being 78 right now (at peak production time) vs. 74 and 73 the last two days.

                  So what causes the dropoff of 3-4% each day? It almost seems like the equipment is getting tired from operating at the max each day. Could it be the sun moving each day now that days are getting shorter? But even at the pretty fast pace, days are ending only 1.5 minutes earlier each day. I get that it's not a major change but the movement on consecutive days has me wondering what the cause is. Also, I don't think there's anything wrong with my system. I'm just trying to get a better grasp on how production is impacted.
                  1.) To see the weather data PVWatts uses, choose the hourly output option. Also, see the TMY manual (--->> Google: NREL --->> "TMY") for details of where PVWatts gets its weather data. Short answer is it's somewhat historic and probably mostly synthetic. See the TMY manual for the gory details. In any case, any daily output that PVWatts estimates will, in all likelihood be a lot different than the output for any actual day.

                  2.) The actual, "live", real; time solar irradiance at any time at your, or any, site can/will change from min./min. - even a bit under apparently clear and what appear to be steady and consistent atmospheric conditions. That will change the day/day total insolation an array sees. Example: Two days ago, and like most days around here, things were cloudless all day. My array output one day to the next for those two days the varied by ~ 5 % day/day. The daylong insolation was 3.9% higher on the first of those days than the second, while the ave. daytime amb. air temp. at the array was about 0.5 C. cooler for that 1st day. Things that can go unnoticed can have an influence on array output. All "clear" days are not identical.

                  3.) The daily change in average incidence angle at the same time every day as the earth spins around the sun, and thus the Plane of Array irradiance which is f(incidence angle), is changing something like ~~ 0.2%/day this time of year, depending also, to some degree, on array orientation.

                  4.) Under clear skies, for the same approximate irradiance, expect array output to drop off approx. 0.5 % or so for every deg C. increase in array temp., which is itself a f(irradiance, wind, ambient air temp.) and a few other things.

                  5.) I'd guess that the variation you report is probably within normal +/- weather and insolation caused variation. The variation of the 1st two days as you report it doesn't seem that different day/day. It may also be that shading is beginning to have some effect. However, if output continues to drop at a consistent rate of several %/day, that may be an indication that something is amiss and a call to the installer may be in order.

                  Comment


                  • #26
                    +1 on what JPM said, plus you will definitely see a drop with temperature. Silicon cell output decrease by about 2.3mV/C. If you have 60 cell panels, that's about 138mV/C per panel. With your third day being 2.8C warmer than your first day, that's about 0.383v less voltage on each panel on the third day, or about 1.3% difference (assuming nominal 30v output) and about 1.3% less power output. Soiling can be a factor also. I've noticed that after a rain washes my panels they perform better and then get more and more dusty until the next rain.

                    edit:
                    ...just re-read JPM's point #4 above. He already covered my point about temperature effects and the numbers agree.
                    Last edited by DaveDE2; 08-24-2016, 05:43 PM.

                    Comment


                    • #27
                      A friend of mine just had a Sunpower system installed with Sunpower microinverters built in to the panels. He has Sunpower monitoring but wants to use PVOutput. Is there a simple way to get the data fed into PVOutput like there is with a Solaredge system?

                      Comment


                      • #28
                        Originally posted by NYHeel View Post
                        A friend of mine just had a Sunpower system installed with Sunpower microinverters built in to the panels. He has Sunpower monitoring but wants to use PVOutput. Is there a simple way to get the data fed into PVOutput like there is with a Solaredge system?
                        Just wanted to follow up. Is there anyway to get Sunpower production data to be fed into PVOutput? There are Sunpower systems on PVOutput that are updated throughout the day so I figure there has to be some way to do it.

                        Comment

                        Working...
                        X