Announcement

Collapse
No announcement yet.

Strange production behavior

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

  • Strange production behavior

    Hi again,
    Starting to monitor a new system with a 6.4kW array (20 x 32W panels) in two strings, connected to the two MPPT of a 5.5kW inverter. Every panel with a Tigo TS4-R-O.

    Today was a bright sunny day and at mid day the system was producing a nice 5kW. There were zero shadows at that time, the weather was windy and cool, and there was not a cloud in sight in the sky.
    I would think this would lead to a steady production stream , and that was the case for an hour before the event as well as after, but then for about an hour, using the Tigo and inverter monitoring, every 10mn, I observed a sharp drop from 5kW down to 0.5kW and back to 5kW across about 3 to 4 minutes. I had the drops at the following times: 12:41; 12:50; 1:01; 1:21; 1:32;

    Is there any good reason for something like that to happen, and if not would it more likely indicate a problem on the inverter side, or possibly the Tigos?

    Do Inverters usually generate a log that could be looked at and give some indication?

    I will be submitting my questions to the installer and the inverter manufacturer, but I would apreciate any feedback from those with experience of all the things that can go wrong

    Thanks

  • #2
    Originally posted by scrambler View Post
    Hi again,
    Starting to monitor a new system with a 6.4kW array (20 x 32W panels) in two strings, connected to the two MPPT of a 5.5kW inverter. Every panel with a Tigo TS4-R-O.

    Today was a bright sunny day and at mid day the system was producing a nice 5kW. There were zero shadows at that time, the weather was windy and cool, and there was not a cloud in sight in the sky.
    I would think this would lead to a steady production stream , and that was the case for an hour before the event as well as after, but then for about an hour, using the Tigo and inverter monitoring, every 10mn, I observed a sharp drop from 5kW down to 0.5kW and back to 5kW across about 3 to 4 minutes. I had the drops at the following times: 12:41; 12:50; 1:01; 1:21; 1:32;

    Is there any good reason for something like that to happen, and if not would it more likely indicate a problem on the inverter side, or possibly the Tigos?

    Do Inverters usually generate a log that could be looked at and give some indication?

    I will be submitting my questions to the installer and the inverter manufacturer, but I would apreciate any feedback from those with experience of all the things that can go wrong

    Thanks
    Sometimes the grid does some funny things that fall outside the acceptable voltage and frequency that the grid tie inverter is looking for. The software usually has a "wait time" of a few minutes before it checks the grid and decides to reconnect. The event could happen multiple times a day depending on where your are connected and if the grid is seeing a lot of fluctuations.

    I would hope your inverter has a way to tell you why it shut down so that you can determine if there are any options to harden it against grid issues.

    Comment


    • #3
      Originally posted by scrambler
      Today was a bright sunny day and at mid day the system was producing a nice 5kW.

      I would think this would lead to a steady production stream, and that was the case for an hour before the event as well as after, but then for about an hour, using the Tigo and inverter monitoring, every 10mn, I observed a sharp drop from 5kW down to 0.5kW and back to 5kW across about 3 to 4 minutes. I had the drops at the following times: 12:41; 12:50; 1:01; 1:21; 1:32;
      Do check for high AC line voltage, which will be pushed even higher by max output. It
      might be triggering the AC voltage monitor shutdown. Bruce Roe

      Comment


      • #4
        Originally posted by bcroe View Post

        Do check for high AC line voltage, which will be pushed even higher by max output. It
        might be triggering the AC voltage monitor shutdown. Bruce Roe
        Thanks, Can you elaborate on what you mean by that?

        Comment


        • #5
          Originally posted by SunEagle View Post

          Sometimes the grid does some funny things that fall outside the acceptable voltage and frequency that the grid tie inverter is looking for. The software usually has a "wait time" of a few minutes before it checks the grid and decides to reconnect. The event could happen multiple times a day depending on where your are connected and if the grid is seeing a lot of fluctuations.

          I would hope your inverter has a way to tell you why it shut down so that you can determine if there are any options to harden it against grid issues.
          Thanks, interesting to know the grid could cause this. The inverter is just in test right now, and we don't have Right to operate yet. Inspection is next week.

          Comment


          • #6
            If you have it, I'd check what info you can get from the POCO on what passed through their meter during those times.

            I can get SDG & E info on what passes through my meter, +/- kWh at 15 minute intervals.

            Such info might be of some help, particularly to identify if it might be a monitoring problem and not a dip in production.

            Comment


            • #7
              Originally posted by J.P.M. View Post
              If you have it, I'd check what info you can get from the POCO on what passed through their meter during those times.

              I can get SDG & E info on what passes through my meter, +/- kWh at 15 minute intervals.

              Such info might be of some help, particularly to identify if it might be a monitoring problem and not a dip in production.
              I actually checked My PG&E consumption data on their site, and so far it does not show energy I send back to the grid.
              I figured that may be because we have not officially gotten the Right to Operate yet, so full monitoring may not be setup on their side.

              Given both the Tigo Monitoring And the Inverter monitoring show the dips, I would assume they are real.
              I have asked the manufacturer and installer if they have logs from the inverter, we will see what they say.

              Comment


              • #8
                Originally posted by scrambler View Post

                I actually checked My PG&E consumption data on their site, and so far it does not show energy I send back to the grid.
                I figured that may be because we have not officially gotten the Right to Operate yet, so full monitoring may not be setup on their side.

                Given both the Tigo Monitoring And the Inverter monitoring show the dips, I would assume they are real.
                I have asked the manufacturer and installer if they have logs from the inverter, we will see what they say.
                Without approval from PG&E you will not get any useful generating information. For that matter you could be charged for any kWh you think you are sending to the grid because your meter won't differentiate between what is coming or going so all is "coming" from the grid and you are being charged for those kWh.

                Comment


                • #9
                  Originally posted by SunEagle View Post

                  Without approval from PG&E you will not get any useful generating information. For that matter you could be charged for any kWh you think you are sending to the grid because your meter won't differentiate between what is coming or going so all is "coming" from the grid and you are being charged for those kWh.
                  It is a matter of days, as Inspections are scheduled for this week.

                  I already have a Smart Meter and it does show negative number when electricity goes back to the grid. Also when I checked my hourly consumption on the PG&E Website, it shows no consumption while the panels are producing, just no negative consumption.
                  So hopefully they are not recording what is being sent back as usage
                  Last edited by scrambler; 02-05-2020, 02:20 PM.

                  Comment


                  • #10
                    Hi again,
                    A quick update and more question for the good willing expert of this site

                    Drops have been getting more under control, only one or two here and there. The company in charge of the inverter has not been sharing much but I get the feeling they have at least lowered the max output of the inverter, In the early days when I first noticed the drops, it was peaking at 5.5 / 5.86 kW (it is a 5.5kW inverter), and not it is clipping at 5.3kW.
                    It was also unlikely to be an out of range voltage or freq as these produce errors that are logged by the inverter.

                    But I decided to take @bcroe advice, got a Data logger from PG&E, and I am starting to do some logging.

                    I have the logger L1 and L2 and Neutral connected at the arrival of the lines at the top of my House panel, and the Amp CTs on these two lines at the same place ( facing the grid).
                    I am logging the average value every second of Volt, Amp, Power Factor and KVAR

                    Pictures and excel sheet data are in the online folder below
                    https://1drv.ms/u/s!At3vMAQjaOZLkSVK...LaxWN?e=3tKP97

                    With the laptop connected to the logger I can also have a look at the real-time Volt, Amp and KW waveforms, as well as harmonics charts.
                    As most of this is above my pay grade, I am not sure what to make of some of what I see in the pictures below, like: When the inverter is producing heavily, with almost no house load, it is showing me amps in opposite phase of Volts. Would the fact current is flowing opposite the direction of the CTS cause that?
                    When there is a big house load at the same time (like the oven), Amps and Volt reading become more in phase
                    WAV%203-11%2011-04%20inv-on%20solar%20small%20house%20load.jpg?psid=1 .jpg
                    WAV 3-11 11-04 inv-on solar & oven load.jpg

                    I have also created an Excel sheet that takes the 13 hours of data from 7AM to 8PM (46800 samples) when the inverter is running with some solar.

                    The first sheet has all the data plus the solar production data, check out of range, min and max values.
                    Then I have several sheets with scrollable graphs for each parameter over a 16 mn window. Graph Scrollbars are synchronized so that if you are looking at the volts in a specific time window and go to another graph, it shows the same period. I also have solar production graphs underneath to put the graph in perspective with the solar production.
                    This allows me to see what is going on prior to a drop in case it reveals anything.

                    The Excel sheet is downloadable at the link mentioned above, you can’t view it online as it is over 5MB and that is apparently the limit for online viewing…
                    I have extracted a few pictures of the things I am not sure how to interpret. They don’t seem to be related to the drop, but I am still curious if they are indicative of something out of norm.. PF Seems to be at -1 when there is no solar production and no house load
                    PF Seems to be at 1 with no solar and some house load
                    As Solar production becomes significant, it sometimes alternates quickly between 1 and -1 over a few seconds, and sometimes is fairly stable at -1. I read that could be showing the flow from the grid or to the grid. If so I am curious about the fast oscillation between the two.
                    Power factor-1.JPG
                    Power factor-2.JPG
                    Power factor-3.JPG

                    I welcome any insights as usual.
                    I will keep monitoring to see If I can spot some consistent behavior prior to a drop.

                    Comment


                    • #11
                      I thought I would update this thread as I have now identified what was going on.

                      Not hearing back from my provider on the sharp drops I had noticed in my Darfon H5001 hybrid inverter, I used my free time during Shelter in place to closely monitor production vs weather and do some testing.

                      This hybrid inverter has several operating modes such as Backup Mode where the battery is never used except during an outage, Self-Supply Mode or TOU arbitrage where the battery is used after sundown or during specific time slots.
                      Mine was setup to Self-Supply and was showing some “random” sharp drops of production followed by a quick ramp up. I initially noticed them during Sunny clear sky days as they could clearly not be explained by clouds, but realized they also happened (even more so) during part clouds day where the exposure alternates between sun and clouds.

                      One of my test was to switch the inverter to Backup Mode, and in that mode the problem vanished instantly. I tested for a full week in various weather conditions which revealed how bad the problem was in Self Supply mode (see comparison graphs below). To confirm I tried switching back to Self-Supply mode and the problem returned instantly.

                      At that point this means either my inverter is faulty, or there is a bug in the Inverter Software. I passed the info to the provider and am awaiting for their take on it.
                      That said, my inverter was scheduled to be replaced next week for another issue with its disconnect switch, so I will soon enough be able to confirm if this is a one off HW failure or a systemic SW issue.

                      I am including graph comparisons between the two modes for various weather condition, for the benefit of potential users of the Darfon H5001 either from Darfon or from Electriq Power. If you are such a user, you may want to have a close look at your production graphs to see if you spot these symptoms.

                      Clouds and sun days

                      Cloudy day comparison.jpg



                      Sunny with light clouds

                      sunny day with light clouds comparison.jpg





                      Sunny clear sky
                      Sunny day comparison.jpg

                      Comment


                      • #12
                        Q: On your most recent graphs, what are the units of "Solar Radiation" and what method is used to measure it ?

                        Comment


                        • #13
                          I am using the reported values from my Ambient WS2000 weather station. The values are in Watt per square meter
                          I initially calculated a multiplying factor using the surface of the array (34 square meter) x the efficiency (18.8%) which gave a 6.4 square meter equivalent.
                          In the end, I used a multiplying factor of 7 which made the radiation graph closer to the production graph.

                          The reading is only reported every 5mn instead of every minute (like the panel production), and not super precise, nor is the sensor exactly where the panels are or tilted the same way.
                          But I find it gives a good enough sense whether the production is somewhat correlated to the production, or if there are obvious discrepancies

                          Comment


                          • #14
                            Originally posted by scrambler View Post
                            I am using the reported values from my Ambient WS2000 weather station. The values are in Watt per square meter
                            I initially calculated a multiplying factor using the surface of the array (34 square meter) x the efficiency (18.8%) which gave a 6.4 square meter equivalent.
                            In the end, I used a multiplying factor of 7 which made the radiation graph closer to the production graph.

                            The reading is only reported every 5mn instead of every minute (like the panel production), and not super precise, nor is the sensor exactly where the panels are or tilted the same way.
                            But I find it gives a good enough sense whether the production is somewhat correlated to the production, or if there are obvious discrepancies
                            Thank you. I had trouble reconciling the values on the vertical axis of the graph with the usual units used for irradiance.

                            I'd suspect the irradiance sensor on the weather station is in a horizontal orientation. If so, and FWIW, expect the sensor readings to be much lower in the winter with the ratio of horizontal irradiance, which what the sensor measures, to P.O.A. irradiance on the array to be much lower in winter than summer.

                            Comment


                            • #15
                              Originally posted by J.P.M. View Post

                              I'd suspect the irradiance sensor on the weather station is in a horizontal orientation. If so, and FWIW, expect the sensor readings to be much lower in the winter with the ratio of horizontal irradiance, which what the sensor measures, to P.O.A. irradiance on the array to be much lower in winter than summer.
                              This is probably the main reason for the difference between the Surface equivalent of 6.4 square meter and the actual factor or 7 that gives a theoretical production matching the actual array production

                              Comment

                              Working...
                              X