Announcement

Collapse
No announcement yet.

Sanity Check for Combining multiple inverters

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

  • Sanity Check for Combining multiple inverters

    I have a project where I am combing 6 inverters for a grid-tie system:
    (4) SMA SB8000TL's
    (2) SMA SB6000TL-22's

    The Grid is 208Y 3-phase, 225A

    I am planning on using an Eaton load center (3-phase), using 2-pole breakers for the inverters. 50A breakers for the 8's and 40A for the 6's. Output for the 8's is 38.5A and 25A for the 6's, which gives me a total of 204A, upped to 255A (125%). The panel I am looking at is a 225A rated CH series, which would allow me 270A (120%).

    Does this seem feasible?

    Oh, regarding existing loads on the utility, there is only an old 225A panel that has a pump attached (was all to be removed / demolished before the client decided they wanted to have solar). So we are going to simply tap at the meter (or splice box) and leave the existing alone.

  • #2
    Originally posted by JohnP View Post
    I have a project where I am combing 6 inverters for a grid-tie system:
    (4) SMA SB8000TL's
    (2) SMA SB6000TL-22's

    The Grid is 208Y 3-phase, 225A

    I am planning on using an Eaton load center (3-phase), using 2-pole breakers for the inverters. 50A breakers for the 8's and 40A for the 6's. Output for the 8's is 38.5A and 25A for the 6's, which gives me a total of 204A, upped to 255A (125%). The panel I am looking at is a 225A rated CH series, which would allow me 270A (120%).

    Does this seem feasible?

    Oh, regarding existing loads on the utility, there is only an old 225A panel that has a pump attached (was all to be removed / demolished before the client decided they wanted to have solar). So we are going to simply tap at the meter (or splice box) and leave the existing alone.
    Your numbers are mostly being applied logically but incorrectly. <sigh>

    1. If you do what is called a line side (or service side) tap, then you can almost make what you want work.
    If the panel is one which is listed for suitable for use as service equipment (SUSE) then you can connect it at the meter.
    But even though there are no loads on the panel, if you put a 225A main breaker on the combiner panel, you will not be able to get away with the 120% rule with a picky AHJ.
    2. The output of each inverter will need to go to a breaker which is at least 125% of the rated output. So the breakers for the 8's will have to be larger than 48.1 (so 50A, as you note.) The breaker for each 6 will need to be greater than 31.5, hence 40A.
    But dealing with a three-phase system is not simple, even for an electrician who does not do it regularly.
    If you had a matched set of all 8's and the panels to drive them, you would have a total of four breakers with one end connected to each of the three line wires.
    But the current in the wires would NOT be four times 50 (=200) or even 4 x 38.5 (= 154) because of vector addition of the line-to-line currents. Instead each line would carry 38.5 x sqrt(3) x 2 amps. That comes out to 133A. That will be the real, actual current.
    If you are under an NEC version before 2014, you will have to instead take the 50A breaker size figure to start with and get 173A. You do NOT need to multiply that again by 125%.
    So, summing it all up, you would have a real current of 133A [2014] or a calculated current of 173A [earlier] on each of the three phase wires.
    Not a big deal for a three phase breaker panel with a 225A bus rating, you might think.
    But you could not put a main breaker of 173A or more into that panel without blowing the 120% rule completely out the window.
    You also could not even put a 150A main in the panel since 133 + 150 is still greater than 270. And a 125A main will be too small.
    If you are lucky, the AHJ will have you apply the 100% rule (173 < 225) instead since there are no loads in the panel and you will be OK.
    Or, if you get a picky AHJ, you are allowed to use a panel without a main breaker and use the six branch breakers as your service disconnect. The problem there is that adding in the existing panel with one main breaker takes the handle count for a complete disconnect of the service to seven. You lose unless you get rid of the existing pump panel and do not add that pump load in anywhere else except maybe on one of the branch circuits that is also used for an inverter. (That one makes my head spin, so do not rely on that idea.)

    Bottom line: This is a complicated installation which is beyond your math abilities at the moment. And even beyond some residential electricians.
    Get a licensed electrician with either three-phase or solar experience or both to help you, and involve the inspector early in the planning process since he may interpret the NEC differently than you or I do.

    Note: putting in your actual mix of 4 x 8000 and 2 x 6000 makes the resulting current values a bit more favorable for you, but the math gets to where I do not feel like doing it at this time. You will still end up with three of the 8000s and one of the 6000s connected to one phase line and the bus calculation may still fail based on the current on the most heavily loaded line conductor(s). If you are on an NEC version before 2014 you still do not have a prayer of meeting the 120% rule.
    Note 2: You appear to have put the 125% factor into your calculations one more time than is required, so my numbers are more favorable than yours for that reason too. Your figure of 255A would not be allowed on the 225A bus, since the 100% rule must be applied to that sum before you get to look at the 120% rule.
    Note 3: What is the nominal size of the POCO service? It had better be at least 200A. And with your numbers instead of mine, it would have to be a 250A (not generally available) or a 300A service (maybe POCO would give you that) or 400A service.

    Bottom bottom line:
    Sanity check results: You are insane and so am I. Get help or scale back the project.
    Last edited by inetdog; 08-20-2014, 10:07 PM.
    SunnyBoy 3000 US, 18 BP Solar 175B panels.

    Comment


    • #3
      Sanity Check...

      Well, Thanks a million Inetdog, this response is exactly what I was hoping for (in twisted way). 3-Phase is not easy, nor are the rules used to utilize it. I have more homework to do and quite honestly, so does my electrician.

      Comment


      • #4
        Why not use Fronius or ABB tripower inverters they are mad for 3 phase
        NABCEP certified Technical Sales Professional

        [URL="http://www.solarpaneltalk.com/showthread.php?5334-Solar-Off-Grid-Battery-Design"]http://www.solarpaneltalk.com/showth...Battery-Design[/URL]

        [URL]http://www.calculator.net/voltage-drop-calculator.html[/URL] (Voltage drop Calculator among others)

        [URL="http://www.gaisma.com"]www.gaisma.com[/URL]

        Comment


        • #5
          hmmmm,

          We we always design around SMA. SMA also makes tripower inverters, but they are for higher voltages. perhaps we need to consider other manufacturers...

          Comment

          Working...
          X