Announcement

Collapse
No announcement yet.

Help Needed with Complicated Solar Layout

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

  • Help Needed with Complicated Solar Layout

    I am building a home in a sunny, dry part of Hawaii where electricity is very expensive. I will have significant electrical loads including a 6 ton central AC system (4 zones), 10x60' pool with pool pump and HPHW pool heater, two 80 gal HPHW tanks, stove, oven, refrigerator, washer, dryer, dishwasher, and 2 EVs. We originally designed the home to have a large south (S) facing roof pitch but my HOA made me chop up my roof to make it more "interesting", which markedly reduced my S facing roof area. Four of my neighbors had their solar systems professionally installed on their new homes (without any consumption records) with more than 50 panels each but they don't have as many loads as I will have. Two of them were lucky enough to have NEMs but the other 2 are now pricing batteries. I hope to install 2 PW2 or 3 LG Chem batteries (RESU10) to store energy for evenings and night use. I have already purchased two new inverters, SE10000H-US and SE7600H-US (at half price). I plan to design the solar layout for maximum use of both inverters but may not install that many panels. I think that it will be cheaper to downsize than to upsize my installation.

    My roof is completed but doesn't show up yet on google maps, so I can't use the Solaredge solar design tool.

    Based on my roof dimensions and pitch, my architect calculates that I can only fit 29 panels (Panasonic 325w) on my S facing pitches, see attached layout. At the maximum oversize rate of 1.55 for SE HD wave inverters, I theoretically could connect 47 325w panels to my SE10000H-US inverter and 36 325w panels to my SE7600H-US inverter. I need to distribute these panels on my S, W and E facing pitches to minimize clipping. I think that I would minimize clipping by connecting ~57% (10/1 7.6) of panels on each pitch to my SE10000H-US inverter and 43% (7.6/17.6) of panels on each pitch to my SE7600H-US inverter but that get complicated when building strings. Each pitch will have a different time of peak production, which helps to avoid clipping.

    Do I need to try to distribute S, W and E facing panels somewhat evenly between the 2 strings of each inverter or should I be more concerned with not having a string with less than a given number of E or W facing panels. I recall reading that with E and W facing panels that you should avoid having less than a certain number in a string for low irradiation times.

    Also, I read in one place that the min, max # panels on a string was 8, 25. Is that correct or is the maximum number of panels in a string calculated with a formula?
    Thanks in advance for any assistance.
    190624 MKW Solar 29S 32W 23E.jpg
    Attached Files

  • #2
    You are mixing string suggestions with SolarEdge design ideas. Also in HI, the sun will be very high and hit all faces mid day.
    you REALLY need to read the installation instructions for the equipment. There are other limitations BESIDES 8-25 modules in a string. The minimum depends on the optimizers you use but on a large system like yours, you should not be close to the minimum. The max is limited by wattage! With Panasonic 325w modules you should be using P400 which would give yo a limitation of 8 min to 6kw max (or 18 with 325w).
    https://www.solaredge.com/sites/defa...tasheet-na.pdf

    Your inverters are not capable of working with the RESU10H battery as well, you would need the StoreEdge SE7600A inverter for that.

    You also need to make sure that the inverters are configured for HI operation, should add the consumption meter, and need to find out if you can have net metering. If you can not get net metering then you will have to limit the inverters to not feed in which requires the consumption meter.
    Last edited by ButchDeal; 06-26-2019, 07:16 AM.
    OutBack FP1 w/ CS6P-250P http://bit.ly/1Sg5VNH

    Comment


    • #3
      According to datasheet, the formula is 6kW/325 = 18 panels/string as maximum number. That max wattage suggests that three input strings would be desirable and the manual confirms that 3 strings can be attached to SE10000H-US.
      What is the best way to distribute the panels from the arrays with 3 different azimuths among the 3 strings? I recall that a string won't activate unless a certain wattage or voltage is achieved. Since my east facing panels get lit up 1st, is there a min # of east facing modules that I should put in a string to activate the string as early as possible? Likewise, the west facing panels will receive the latest irradiance. What is the min # of west facing panels that I want to put in a string to keep it activated as late as possible? I realize that not all 3 strings may have panels from all 3 azimuths.

      Comment


      • #4
        Originally posted by Steeler.Fan View Post
        What is the best way to distribute the panels from the arrays with 3 different azimuths among the 3 strings? I recall that a string won't activate unless a certain wattage or voltage is achieved. Since my east facing panels get lit up 1st, is there a min # of east facing modules that I should put in a string to activate the string as early as possible? Likewise, the west facing panels will receive the latest irradiance. What is the min # of west facing panels that I want to put in a string to keep it activated as late as possible? I realize that not all 3 strings may have panels from all 3 azimuths.
        Again you are mixing string inverter information that you read and does not apply to optimized systems.
        With solaredge there is no certain wattage or voltage to turn on. You just need the minimum number and even slight light will do it. So even the west facing modules will come on in the morning. The optimizers are very efficient and it does not take much to turn them on.
        OutBack FP1 w/ CS6P-250P http://bit.ly/1Sg5VNH

        Comment


        • #5
          A solaredge support person suggested that I find a nearby home with desired roof azimuths and model my panel layout on their roof. I did find such a home and eventually figured out how to prepare the roof model and add panels to the south facing pitch on that roof. That worked well for my one inverter model. I love the monthly bar graphs of estimated output along with percent of each month's output that would be clipped and the overall annual estimated percent clipped, see attached.

          I have add my 2nd inverter and the right number of panels to the S, W and E facing pitches. Now I need to experiment with different panel distributions on my 5 strings to find the optimal string distribution of my S, W & E facing panels that will result in the least estimated amount of annual clipping.

          I could have loaded a image of my roof into their online program and worked with it but using a large enough roof with appropriate azimuth pitches and close to my home seems to be the easiest solution.
          image_12552.jpg
          Last edited by Steeler.Fan; 07-09-2019, 10:38 PM.

          Comment


          • #6
            We decided to change our layout to 83 panels with 26S, 22E and 35W facing panels. I have placed those number of panels on my neighbor's roof and next I will experiment with different distributions of panels from the various azimuths onto my 5 inverter strings, to find the distribution which will minimize my clipping and yet allow me maximum oversizeing (1.55) of my SE10000H-US and SE7600H-US HD Wave inverters.
            Neighbor House w 26S 22E 35W panels.jpg

            Comment


            • #7
              It does not look like it meets the ridge, hip, and valley setback requirements.
              i would skip the east roof and out those on the south.
              you could always add another inverter to avoid clipping or use two 10kw units.
              OutBack FP1 w/ CS6P-250P http://bit.ly/1Sg5VNH

              Comment


              • #8
                My county permit reviewer has been vague about setback. The County guidelines only specify a 10" setback from roof ridge. When I asked him about valley and roof edge setbacks, his only answer was "it depends". In order to place panels on my overhangs (roof portions beyond walls), I would need to hire a structural engineer and then try to convince the county reviewer and the HOA that it is permissible to mount panels on my overhangs. They may force me to do that if they invoke setbacks on valleys, hips and edges.

                Comment


                • #9
                  Here is my first generation of estimated output from SE solar designer site with the following string distributions:
                  SE10K String#1 16S; #2 18W; #3 13E
                  SE7.6K String#1 10S, 8W; #2 9W, 9E
                  How crazy is this?
                  MKW 26S 22E 35W Est Mo Energy w clipping.jpg

                  Comment


                  • #10
                    Here is the system loss diagram. I am happy with 5% annual clipping.
                    Do these numbers look reasonable?
                    MKW 26S 22E 35W System Loss Diagram.jpg

                    Comment


                    • #11
                      I am not going to try to tweak the panel distribution on the strings to minimize clipping until I run the panel layout diagram by my county permit reviewer. If he says that I am missing unofficially required valley, hip and roof edge set backs and have to reduce the number of panels on my roof, I may need to try to place panels on my roof overhangs.

                      Comment


                      • #12
                        I couldn't resist tweaking the distribution of some panels to see how it affected clipping. I thought that my 1st trial had too many S facing panels on the SE10K inverter, which led to more clipping.
                        1st Trial
                        SE10K String#1 16S; #2 18W; #3 13E; SE7.6K String#1 10S, 8W; #2 9W, 9E
                        2nd Trial
                        SE10K String#1 15S; #2 18W; #3 14E; SE7.6K String#1 11S, 7W; #2 10W, 8E
                        3rd Trial
                        SE10K String#1 14S; #2 18W; #3 15E; SE7.6K String#1 12S, 6W; #2 11W, 7E

                        Surprisingly, the estimated output was identical for all 3 of my trials. I would have expected less clipping with SE10K inverter and more with SE7.6K but not identical. Maybe it just calculates clipping by combining all panels and comparing it to AC limit of 17.6 kW.
                        Last edited by Steeler.Fan; 07-10-2019, 09:44 PM.

                        Comment


                        • #13
                          Originally posted by Steeler.Fan View Post
                          I couldn't resist tweaking the distribution of some panels to see if it affected clipping. I thought that my 1st distribution may have had too many S facing panels on the SE10K inverter, which might lead to more clipping.
                          1st Trial
                          SE10K String#1 16S; #2 18W; #3 13E; SE7.6K String#1 10S, 8W; #2 9W, 9E
                          2nd Trial
                          SE10K String#1 15S; #2 18W; #3 14E; SE7.6K String#1 11S, 7W; #2 10W, 8E
                          3rd Trial
                          SE10K String#1 14S; #2 18W; #3 15E; SE7.6K String#1 12S, 6W; #2 11W, 7E

                          Surprisingly, the estimated output was identical for all 3 of my trials. I would have expected less clipping with SE10K inverter and more with SE7.6K but not identical. Maybe it just calculates clipping by combining all panels and comparing it to AC limit of 17.6 kW.
                          The difference is going to be small in Hawaii as the sun is quite high at noon, striking all your azimuths and that is where the clipping will be. Unless you do something like put all one azimuth on the smaller inverter then you will increase clipping...
                          OutBack FP1 w/ CS6P-250P http://bit.ly/1Sg5VNH

                          Comment


                          • #14
                            A SE support engineer told me that if I connect a battery DC-DC via a Storedge interface to an appropriate SE inverter, it can store excess energy produced above my AC output limits (10 & 7.6 kWh) and reduce my clipping.

                            From the above bar graph, I calculated the following amount of clipped energy for the 2 months with the greatest amounts of clipping;
                            For May with 3770 kWh production and 13.32% clipping = 502 kWh/mo ~ 16.2 kWh/day.
                            For July with 4090 kWh production and 8.45% clipping = 345 kWh/mo ~ 11.2 kWh/day.
                            If I connected 2 LG Chem RESU10 DC-DC to my inverters, I should be able to store rather than clip all excess energy. Is this logic correct?

                            If so, I hope that SE will soon release a StorEdge Interface for the SE10000H-US and SE7600H-US inverters, similar to the SESTI-3 and SESTI-4 interfaces that work with their SE7600 and SE10000 HD Wave inverters in Asia, Australia, New Zealand and Europe!
                            Last edited by Steeler.Fan; 07-10-2019, 09:30 PM.

                            Comment


                            • #15
                              I thought about tilting my E and W facing panels, which could further separate their times of peak production and reduce clipping. However, that would require greater spacing between panels to avoid shadowing. My roof space seems to be too small to allow greater spacing between rows of panels.

                              Comment

                              Working...
                              X