Announcement

Collapse
No announcement yet.

Very weird array behavior

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

  • Very weird array behavior

    Hi everyone - thanks in advance for any insight anyone can offer. This issue has me baffled.

    System: 24 x 305w w/ Enphase m250 microinverters; grid tied; installed in Jan 2015

    My system has been working fine for over two and half years. Three days ago the system-wide Enphase production graph start showing some very weird behavior. Instead of the smooth power production curve I'm normally used to seeing on sunny days, I start seeing a "step ladder" effect. i.e. AC current going up, then down slightly, then up again as the day progresses.

    Coincidentally (or not), around the same time one of my panels starts going offline around 11 AM and stays dark for the remainder of the day. The next day (and every day since) the panel starts producing normally (except for the aforementioned step ladder issue), then again goes offline around the same time.

    Graphs

    Broken panel or inverter? I'm highlighting the moment production goes to zero for this one panel....usually around 11 AM each day.



    System-wide production. This is normally a smooth curve on sunny days. Instead, production goes up, then down slightly, then up again. All panels exhibit this behavior.



    Clearly something needs to be fixed, but I'd love to do as much troubleshooting from the comfort of my home office before I crawl up on my 120* roof (I'm in NC).

    Thanks!

  • #2
    If a turnkey system, what does the installer say ? If it is turmkey, I'm sure the vendor would like to at least know about the problem.

    Comment


    • #3
      One thing to check out is the line voltage. If its pretty high, peak production might push it high enough to trip out
      some micros on high AC line. Once enough micros have dropped out, the line might decrease enough for the rest
      to stay on line. This roughly fits your description of some micros dropping to zero, system power dropping to half.
      Bruce Roe

      Comment


      • #4
        Originally posted by bcroe View Post
        One thing to check out is the line voltage. If its pretty high, peak production might push it high enough to trip out
        some micros on high AC line. Once enough micros have dropped out, the line might decrease enough for the rest
        to stay on line. This roughly fits your description of some micros dropping to zero, system power dropping to half.
        Bruce Roe
        Each micro inverter puts out 240V and they are wired in parallel, so how could AC line voltage be too high? Looks like the graph is showing AC voltage constant.
        Last edited by brucet9; 07-26-2017, 10:38 PM.

        Comment


        • #5
          From the first graph, it looks to me that at the failure point, module current drops to nothing and voltage rises simultaneously, possibly from Vpm to Voc level. Wouldn't that suggest that at that point the module loses load, going to open circuit status. If that is the case, might it not be because heat in the micro inverter causes that open circuit condition? Perhaps the same component that is now "stepping" power output instead of delivering a smooth power curve in response to a smooth DC current input, fails open at about 116deg F?

          Comment


          • #6
            Originally posted by grommet View Post
            Hi everyone - thanks in advance for any insight anyone can offer. This issue has me baffled.

            System: 24 x 305w w/ Enphase m250 microinverters; grid tied; installed in Jan 2015

            My system has been working fine for over two and half years. Three days ago the system-wide Enphase production graph start showing some very weird behavior. Instead of the smooth power production curve I'm normally used to seeing on sunny days, I start seeing a "step ladder" effect. i.e. AC current going up, then down slightly, then up again as the day progresses.

            Coincidentally (or not), around the same time one of my panels starts going offline around 11 AM and stays dark for the remainder of the day. The next day (and every day since) the panel starts producing normally (except for the aforementioned step ladder issue), then again goes offline around the same time.

            Graphs

            Broken panel or inverter? I'm highlighting the moment production goes to zero for this one panel....usually around 11 AM each day.



            System-wide production. This is normally a smooth curve on sunny days. Instead, production goes up, then down slightly, then up again. All panels exhibit this behavior.



            Clearly something needs to be fixed, but I'd love to do as much troubleshooting from the comfort of my home office before I crawl up on my 120* roof (I'm in NC).

            Thanks!
            If you can / willing to go DIY route how difficult it would be to swap panels between inverters? You don't need to move them mechanically, just electrically.
            I'd think that inverter is on its way out or its AC connection is breaking- prob around 11AM it gets hot enough to break and never cools down enough to restore contact until next morning. I'd expect it to stay dead eventually.
            Last edited by max2k; 07-26-2017, 11:14 PM.

            Comment


            • #7
              Originally posted by max2k View Post

              If you can / willing to go DIY route how difficult it would be to swap panels between inverters? You don't need to move them mechanically, just electrically.
              I'd think that inverter is on its way out or its AC connection is breaking- prob around 11AM it gets hot enough to break and never cools down enough to restore contact until next morning. I'd expect it to stay dead eventually.
              Now, you know that troubled inverter will be in the middle of a double row of panels on a 12:12 pitch roof, don't you?

              Comment


              • #8
                Originally posted by brucet9 View Post
                Each micro inverter puts out 240V and they are wired in parallel, so how could AC line voltage be too
                high? Looks like the graph is showing AC voltage constant.
                The inverters put out whatever voltage is required to push power back into the line. 249 VAC is already
                out of range; the more power (inverters) pushing, the higher the voltage will go. Other bobbles in the
                line might help cause some inverters to trip out, letting the voltage drop back some and you running
                at reduced power. Bruce Roe

                Comment


                • #9
                  Originally posted by brucet9 View Post

                  Now, you know that troubled inverter will be in the middle of a double row of panels on a 12:12 pitch roof, don't you?
                  What, is it not always there? .

                  Comment


                  • #10
                    Originally posted by bcroe View Post

                    The inverters put out whatever voltage is required to push power back into the line. 249 VAC is already
                    out of range; the more power (inverters) pushing, the higher the voltage will go. Other bobbles in the
                    line might help cause some inverters to trip out, letting the voltage drop back some and you running
                    at reduced power. Bruce Roe
                    Inverters, as I understand their function, try to operate at some nominal voltage while varying outputted amperage to maintain maximum power output. 249V is high, but at 3.75% over nominal voltage, probably within acceptable variance in residential systems. But if inverters primarily managed power by varying voltage, how would they be able to share a common bus with Grid power?

                    I understand how amperage increases additively as modules/inverters are added to a parallel circuit. But does not voltage in such a circuit settle at the average voltage of the several inverters, much as would happen with several batteries in parallel? If so, adding or subtracting inverters of nearly the same voltage output will change only slightly the voltage of the circuit.

                    Comment


                    • #11
                      OP here - thank you *so much* for the feedback. It's too damn hot to get on the roof here in NC at the moment so I'm going to wait until a cool weekend to check things out.

                      On my to do list:
                      • check AC connections/rewire in the junction box on the roof. I used wire nuts to combine four 240AC array strings into two 240V AC circuits in my house so perhaps those connections have gone bad and are causing voltage issues.
                      • Swap adjacent panels to see if the issue travels along with it. *Thank God* the inverter in question is on the edge of my array. Perhaps troubleshoot with Enphase if we suspect it's going bad.

                      I'll report back to this thread when/if I find out what's going on.Thanks again!

                      Edit: I just checked my dashboard today and lo and behold the panel/inverter is completely dead. @max2k pretty much nailed its fate.
                      Last edited by grommet; 07-27-2017, 10:08 AM.

                      Comment


                      • #12
                        Originally posted by brucet9 View Post

                        Inverters, as I understand their function, try to operate at some nominal voltage while varying outputted amperage to maintain maximum power output. 249V is high, but at 3.75% over nominal voltage, probably within acceptable variance in residential systems. But if inverters primarily managed power by varying voltage, how would they be able to share a common bus with Grid power?

                        I understand how amperage increases additively as modules/inverters are added to a parallel circuit. But does not voltage in such a circuit settle at the average voltage of the several inverters, much as would happen with several batteries in parallel? If so, adding or subtracting inverters of nearly the same voltage output will change only slightly the voltage of the circuit.
                        Your inverters may be "in parallel" with the PoCo. But the PoCo is so big they dominate the situation;
                        your inverters must adjust. Also the voltage at your outlet probably drops when a big load (air cond)
                        surges on. That is because all the wiring, etc between the high line and your outlet have resistance,
                        so the increased current causes increased voltage drop in that resistance. When PV inverters push
                        power back, this happens in reverse; your voltage goes up. Enough current, enough resistance, and
                        a high PoCo line to begin with might reach the typical trip point of 264VAC. I don't know if that is
                        the problem, but its common enough, check it out. Bruce Roe

                        Comment


                        • #13
                          Originally posted by grommet View Post

                          Edit: I just checked my dashboard today and lo and behold the panel/inverter is completely dead. @max2k pretty much nailed its fate.
                          Probably still worth checking the condition of those wire nuts...
                          CS6P-260P/SE3000 - http://tiny.cc/ed5ozx

                          Comment


                          • #14
                            Issue 1 - stairstep: The stairstep is due to a change from Enphase on how they report/use data. Previously you got 5 min data that represented that 5 minutes. Now the 5 minute data is the same for 15 minutes. Many (not all) Enphase systems on Pvoutput are showing it - starting on July 17th. Even your enlighten view will show the same stairstep. In terms of that stair step waveform on the graphs - there is nothing wrong or unique. My guess is a firmware update was pushed not everyone has it so some see it - some don't. Maybe its a bug in their server side or in the firmware or might even be an intentional change - who knows?

                            Issue 2 - Microinverter not producing. Download and use the Enphase installer toolkit for your smart phone (find it on Google play or Apple store - its free). It allows you to see the actual logs for the envoy which has them for every microinverters. That will clearly tell you why the microinverter is shutting down - everything else is a guess.

                            As a side note, I have not logged into my Enphase toolkit in a few weeks and when I just connected it told me it wanted to update the Envoy firmware. The Enlighten www site does not offer this, so maybe the 5 min data replicated for 15 minutes will be fixed by this upgrade. I'll edit this post once I have a days worth of output with the updated firmware.
                            Last edited by tyab; 07-27-2017, 04:30 PM.

                            Comment


                            • #15
                              Originally posted by tyab View Post
                              Issue 1 - stairstep: The stairstep is due to a change from Enphase on how they report/use data. Previously you got 5 min data that represented that 5 minutes. Now the 5 minute data is the same for 15 minutes. Many (not all) Enphase systems on Pvoutput are showing it - starting on July 17th. Even your enlighten view will show the same stairstep. In terms of that stair step waveform on the graphs - there is nothing wrong or unique. My guess is a firmware update was pushed not everyone has it so some see it - some don't. Maybe its a bug in their server side or in the firmware or might even be an intentional change - who knows?

                              Issue 2 - Microinverter not producing. Download and use the Enphase installer toolkit for your smart phone (find it on Google play or Apple store - its free). It allows you to see the actual logs for the envoy which has them for every microinverters. That will clearly tell you why the microinverter is shutting down - everything else is a guess.

                              As a side note, I have not logged into my Enphase toolkit in a few weeks and when I just connected it told me it wanted to update the Envoy firmware. The Enlighten www site does not offer this, so maybe the 5 min data replicated for 15 minutes will be fixed by this upgrade. I'll edit this post once I have a days worth of output with the updated firmware.
                              Thanks @tyab - this is exactly what I was looking for! Glad to hear I'm not the only one with this stair step issue.

                              Re: the bad inverter, I had to order an AP stick from Amazon in order to get connected via the mobile app. Once I get that on Sunday I'll have a better idea what is going on. I'll report back here once I find out something. Thanks again!

                              Comment

                              Working...
                              X