Determining array wiring without removing panels - tricks?

bill von novakbill von novak Solar Expert Posts: 891 ✭✭✭✭
When we purchased our most recent home, it came with a 9.68KW solar power system.  It was our first home with solar we didn't install ourselves.  Recently though I've seen a pretty consistent 20-30% loss of power, and I am suspecting a bad string.

Problem is that while the plans the homeowner gave us match the placards, neither matches the physical design of the system or the voltages I see.  (The company that installed it is also out of business.)  Through some experimenting and wire tracing I have reconstructed the basics of the system topology:

44 panels Sunpower 220 watt

Inverter 1 - Sunpower 5200 (really Fronius)
4 strings of 7 panels, all facing southeast
Strings paralleled down to 2 feed wires from roof
Feed wires switched by 2 poles of a 3 phase disconnect
Total power 6160 watts

Inverter 2 - Sunpower 3500 (really Fronius)
1 string of 8 panels facing southeast
1 string of 8 panels facing southwest
Strings paralleled down to 1 feed wire from roof
Feed wire switched by 1 pole of a 3 phase disconnect
Total power 3520 watts

So one problem that became immediately evident was that I have two orientations paralleled.  That won't be easy to fix so I am going to leave that for now.

The southwest array is easy to figure out (8 panels 1 string.)  But the southeast array is 36 panels - 1 string of 8 and 4 strings of 7 - that aren't laid out in any logical pattern.  Before starting to test individual strings I'd really like to understand how they are wired on the roof so I can be more methodical about debugging - but I also don't want to pull off a dozen panels to get at the wiring beneath.

So does anyone have any good tricks to use to determine string topology without pulling panels off the rack?  (This is sort of the Christmas light problem in reverse - and with much higher voltages/currents.)

Comments

  • mcgivormcgivor Solar Expert Posts: 3,600 ✭✭✭✭✭✭
    Seems whatever method you attempt, it will involve a fair amount of work. If you were to isolate a string at the combination point, then using a sheet of cardboard cover the panels one by one and when the output collapses, you will know that that particular panel is part of that string. Draw a diagram of the panel layout and assign a number to each string, the layout should be logical to some degree, this would let you know which panels are in which string. Get an assistant and a case of beer, for afterwards of course. 
     
    1500W, 6× Schutten 250W Poly panels , Schneider MPPT 60 150 CC, Schneider SW 2524 inverter, 400Ah LFP 24V nominal battery bank 
    Second system 1890W  3 × 300W No name brand poly, 3×330 Sunsolar Poly panels, Morningstar TS 60 PWM controller, no name 2000W inverter 400Ah FLA 24V nominal used for water pumping and day time air conditioning.  
    5Kw Yanmar clone single cylinder air cooled diesel generator for rare emergency charging and welding.
  • PhotowhitPhotowhit Solar Expert Posts: 5,410 ✭✭✭✭✭
    edited January 2017 #3
    You've got a problem and it's on the roof, I guess.

    Not unusual to have some slow days this time of year, but if you're pretty comfortable that you have a problem. I would first determine which sting is having issues. This can be done simply with a DC clamp meter. Clamp each string in full sun (easiest way to know they are all exposed to the same amount of sun). Write down the measurements, The string with a large difference will be your faulty string. If all measurements are close, it's just the time of year and the short days and winter storms.

    To figure out which panels are in  which strings. cover a good sized portion of a panel and look at your measurements. There will be a huge different in the string that has that panel. The voltage may be very close, but the amperage will drop off considerably. Now just move what ever you have covering the panel from one to another. I think you could do this with some string, there would be enough light that got around the string so that you will not see as much difference between panels covered by the string and those with several square feet covered.

    For my ground array, I could see devising a board 10" x 2' and make a handle sort of on top. I would put the rubberized shelf paper type stuff on the bottom to add friction. I could move it around with the pole I use to brush the panels.
    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Prosine 1800 and Exeltech 1100, 660 ah 24v ForkLift battery. Off grid for @16 of last 17 years. Assorted other systems, and to many panels in the closet to not do more...lol
  • mcgivormcgivor Solar Expert Posts: 3,600 ✭✭✭✭✭✭
    Another way to determine which panels are in which string is by isolating a sting and connect a telco tone generator between  one leg and ground, attach the tone amplifier to a long stick/pole and pass over each panel, the tone will let you know which are connected. If you know someone in the security, telecommunications or IT industry they more than likely have one you could borrow. See link.
    https://www.anixter.com/en_us/products/26000900/FLUKE-NETWORKS/Telecom-Test-Equipment/p/290695
    1500W, 6× Schutten 250W Poly panels , Schneider MPPT 60 150 CC, Schneider SW 2524 inverter, 400Ah LFP 24V nominal battery bank 
    Second system 1890W  3 × 300W No name brand poly, 3×330 Sunsolar Poly panels, Morningstar TS 60 PWM controller, no name 2000W inverter 400Ah FLA 24V nominal used for water pumping and day time air conditioning.  
    5Kw Yanmar clone single cylinder air cooled diesel generator for rare emergency charging and welding.
  • bill von novakbill von novak Solar Expert Posts: 891 ✭✭✭✭
    Thanks for the suggestions mcgivor and photo.  I have an old wire tracer (intended to find buried power lines) that I can try, and the shading trick sounds effective.  Both should at least give me a list of which panel is in which string, which should be half the battle.
Sign In or Register to comment.