Announcement

Collapse
No announcement yet.

SMA SB7000TL-US-22 Inverter - Accessing individual MPPT performance information?

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

  • SMA SB7000TL-US-22 Inverter - Accessing individual MPPT performance information?

    Hello,
    I apologize for posting this to two different forums but after the first post (here) I realized that this was probably the wrong forum, but I couldn't figure out how to delete the post.

    I would like to be able to monitor the individual performance information for each of the two MPPTs in my SMA SB7000TL-US-22 inverter. It doesn't appear that there is a way to do this with Sunny Portal, Sunny Places, or Sunny Explorer (is there?) since all I see there is overall inverter performance, but I'm not adverse to writing my own TCP/IP application to communicate with the inverter internals directly through its Modbus interface. However, since I haven't yet researched exactly what can be done at this level, do any of you know definitively if individual MPPT performance information would be available using this approach?
    Thanks,
    Ray
    Last edited by MinorDeity; 01-27-2016, 01:05 AM.

  • #2
    There is some technical support for writing your own Modbus interface on the SMA site here. They've updated it recently to include a spreadsheet with the available registers for each inverter type... super helpful if you already understand how Modbus works.

    For your inverter, the following appears to be the only MPPT specific data available:

    SMA Modbus Profile
    Modbus register address Short description Type Format Unit Access
    30769 DC current input [1] S32 FIX3 A RO
    30771 DC voltage input [1] S32 FIX2 V RO
    30773 DC power input [1] S32 FIX0 W RO
    30957 DC current input [2] S32 FIX3 A RO
    30959 DC voltage input [2] S32 FIX2 V RO
    30961 DC power input [2] S32 FIX0 W RO
    CS6P-260P/SE3000 - http://tiny.cc/ed5ozx

    Comment


    • #3
      Thanks Sensij. Those 6 parameters appear to be just what I'm looking for. Once I can get the Installer PW from my installer (or pay SMA to setup a new one ) I plan to enable Modbus TCP/IP and write an application in C# to log individual MPPT history in a way similar to how Sunny Portal logs the combined MPPT power. As a side note, it looks like we have the same kind of PV panels. I have 30ea. CS6P-260P panels in my system and they are working well (except for the twice/day telephone pole shadow during December and January - So much for Opti-Trac ).

      Comment


      • #4
        I finally got the Installer passwords straightened out for both my SB7000TL-US-22 system and my daughter's dual SB6000TL-US-22 system, but now I have another concern. My SB7000TL-US-22 was installed just over a year ago and when I use Sunny Explorer to inspect it the External Communication item on the Settings page shows both Modbus and Webconnect. However, both SB6000TL-US-22s only show Webconnect, and they were installed just a month ago. I can't find anything in the SMA documentation that states that the 6000 and 7000 should be different in this regard. Do you know if maybe SMA now provides support for both Modbus and WebConnect by default right out of the box? I'm wondering if my daughter's installer may have dug out some older inverters that only supported Webconnect. If so I need to hit them up for updating them as necessary. I looked up add-on items for the inverters and I found something called the SMA SWDM-10 Speedwire/Webconnect Interface (for a whopping $191) and I'm wondering if I'm going to have to spend an additional $400 for two of those. Besides, I already have the Webconnect part.

        Comment


        • #5
          My understanding is that modbus is available on all webconnect cards. The item you found is the same as what was installed in your inverter to give you the webconnect access (assuming you are using TCP, not RS-485).
          CS6P-260P/SE3000 - http://tiny.cc/ed5ozx

          Comment


          • #6
            Thanks for all the suggestions. The problem I found with both SB6000TL-US-22s was that even though they were manufactured recently they had the old firmware in them. Once I updated that Modbus became available. I then wrote an application that displays power, current, and voltage graphs for the individual MPPTs in as many inverters as are in the system. Below are some one-day graphs for my single SB7000TL-US-22 system.

            Comment


            • #7
              Nice!
              It looks like the big voltage sag is due to panel temperature. What kind of panels do you have? What software did you write the application with? Tnx.

              Comment


              • #8
                Originally posted by DaveDE2 View Post
                Nice!
                It looks like the big voltage sag is due to panel temperature. What kind of panels do you have? What software did you write the application with? Tnx.
                Hard to tell, but depending on wind, array location and clearance under the array, that voltage drop doesn't look all that unusual to me.

                Comment


                • #9
                  Originally posted by DaveDE2 View Post
                  Nice!
                  It looks like the big voltage sag is due to panel temperature. What kind of panels do you have? What software did you write the application with? Tnx.

                  I have a 7.8kW system containing 30ea Canadian Solar 260w panels. My custom monitoring setup consists of two separate applications written in C#. One runs 24/7 and merely collects data from the inverter every 5 minutes and stores it in daily history files, kind of like what Sunny Portal does even if you're not logged on. The second application can be run whenever you wish and it displays the graphs for whichever history file you select, like what Sunny Portal does when you log on. I've been thinking about something else of value to monitor but at this point it seems like the P/I/V from each MPPT kind of encapsulates it all.

                  Comment


                  • #10
                    Looks like you are running 3 strings total, 10 modules per string, 2 in parallel into "Inverter 1" and the other into "Inverter 2". The panels on inverter 2 appear to be facing a bit more east than the other two strings?

                    At first I thought you could use voltage sag as a measure of average module temperature since for example CS6P-260P voltage output varies by -0.113V/C. On second thought though, that's not an accurate way of measuring module temp because the MPPT has an effect on the module voltage as irradiance varies throughout the day ie; the operating voltage ignoring temperature effects increases with irradiance due to the MPPT. Some of that voltage sag is due to DC line losses too. Perhaps MPPT effects and line losses could be calibrated out but sounds like a bit of a science project. Like you care about any of this. Ha!

                    Comment


                    • #11
                      Originally posted by DaveDE2 View Post
                      At first I thought you could use voltage sag as a measure of average module temperature since for example CS6P-260P voltage output varies by -0.113V/C. On second thought though, that's not an accurate way of measuring module temp because the MPPT has an effect on the module voltage as irradiance varies throughout the day ie; the operating voltage ignoring temperature effects increases with irradiance due to the MPPT. Some of that voltage sag is due to DC line losses too. Perhaps MPPT effects and line losses could be calibrated out but sounds like a bit of a science project. Like you care about any of this. Ha!
                      The change in voltage as a function of irradiance is very small. I've penciled out some calculations in another thread using the voltage data from my SolarEdge system, and it appears to correlate to cell temperature well enough that I don't think the other factors are very significant. Clearly, at startup, shutdown, and during cloud/shade events the MPPT becomes dominant, but under clear sky operation, using voltage as a proxy for temperature has held up well enough to the limited testing I've put at it so far.
                      CS6P-260P/SE3000 - http://tiny.cc/ed5ozx

                      Comment


                      • #12
                        Originally posted by sensij View Post

                        The change in voltage as a function of irradiance is very small. I've penciled out some calculations in another thread using the voltage data from my SolarEdge system, and it appears to correlate to cell temperature well enough that I don't think the other factors are very significant. Clearly, at startup, shutdown, and during cloud/shade events the MPPT becomes dominant, but under clear sky operation, using voltage as a proxy for temperature has held up well enough to the limited testing I've put at it so far.
                        From the cell temps I've measured and voltages measured at my inverter display taken at approximately simultaneous and approximately "instantaneously" - 16 panel temps taken at 4 random locations/panel 2X - ~ 6 min. on either side of min. incid. angle time and under clear sky conditions, while also using displayed voltages at the inverter to est. array temp. seems to produce array temp. est. that are reasonably close to one another.

                        Simply put, the two methods I use seem to be good to an ESTIMATED +/- 1 to 2 degrees C.or so of a representative array average temp., and seem to head in the same direction, and agree with one another reasonably well.

                        As required my my attempts to find out how fast my array fouls and the nature of the fouling and how cleanin/rain affects the fouling rate, I did that side/side comparison about 35 times (Note to Sensij: still farting around w/ mean/std. dev./etc. on all the data - film at 11) on very clear days over the course of about 18 months or so, all data taken at the time of those days min. array incid. angle, and convinced myself that measuring voltages at the inverter was about as accurate as measuring temps. under the entire array - and yes - I can get under it. That was a design consideration. There's more to it than that but that's my story and I'm sticking to it.
                        Last edited by J.P.M.; 02-23-2016, 12:33 AM.

                        Comment


                        • #13
                          Originally posted by DaveDE2 View Post
                          Looks like you are running 3 strings total, 10 modules per string, 2 in parallel into "Inverter 1" and the other into "Inverter 2". The panels on inverter 2 appear to be facing a bit more east than the other two strings?
                          Actually, the 10-panel inverter 2 single-string is facing approximately Southwest and the double 10-panel inverter 1 paralleled strings are facing approximately Southeast. It seems a little strange to me that the Southwest array peaks slightly earlier than the Southeast array, but the whole system has been in operation for over a year and has exceeded its estimated production. I suppose it's possible the installers may have erroneously paralleled the SW string with one of the SE strings. Do my graphs give you any indication that this might be the case and if so, would there be anything to gain by correcting the issue? I suppose an empirical way to find out would be to through a blanket over part of the SW array and see if both power curves dip.

                          Comment


                          • #14
                            Originally posted by MinorDeity View Post

                            Actually, the 10-panel inverter 2 single-string is facing approximately Southwest and the double 10-panel inverter 1 paralleled strings are facing approximately Southeast. It seems a little strange to me that the Southwest array peaks slightly earlier than the Southeast array, but the whole system has been in operation for over a year and has exceeded its estimated production. I suppose it's possible the installers may have erroneously paralleled the SW string with one of the SE strings. Do my graphs give you any indication that this might be the case and if so, would there be anything to gain by correcting the issue? I suppose an empirical way to find out would be to through a blanket over part of the SW array and see if both power curves dip.
                            If you have a total of 3 strings on this inverter, 10 panels each, the data are clearly showing that a SE and SW string are in parallel on MPPT 1, and a SE string is by itself on MPPT 2. With no shading (it doesn't look like you have any), this is probably a better design than putting the SE strings in parallel and the SW string by itself, since that configuration would lead to less balanced power between the MPPT's.
                            CS6P-260P/SE3000 - http://tiny.cc/ed5ozx

                            Comment


                            • #15
                              Originally posted by sensij View Post

                              this is probably a better design than putting the SE strings in parallel and the SW string by itself, since that configuration would lead to less balanced power between the MPPT's.
                              My main concern was that the voltages might be significantly different on the SE and SW arrays when the sun first comes up and goes down. Although I assume a higher voltage from one array wouldn't cause reverse current flow in the lower voltage array due to protection diodes, couldn't it potentially exceed their reverse breakdown voltage? As you can see from my voltage plot for the 2-string MPPT, there is a significant downward spike near both sides, but there is also a similar spike on the AM side of the 1-string MPPT. I don't know the cause or significance of these, however.

                              Also, you seem to indicate that an MPPT power balance is good. My daughter has a 2-SB6000TL-US-22 inverter system with 3ea 13-panel strings. Thus, only one MPPT is being used on one of the inverters. Is there a problem with doing this?
                              Last edited by MinorDeity; 02-24-2016, 04:19 PM.

                              Comment

                              Working...
                              X