Announcement

Collapse
No announcement yet.

SMA "secure power" and rapid-shutdown

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

    SMA "secure power" and rapid-shutdown

    Interested in using one of SMA's new "-41" series inverters. After searching SMA's webpages, I am still mightily confused about when and how SPS "secure power" can peacefully co-exist with 2017 rapid-shutdown requirements.

    Is it even possible ? What specifically is required ? TS4-R-O optimizers on every panel ? An explicit disconnect for the high-voltage DC ?

    #2
    The SPS does not start automatically when the main power is cut, and also does not supply power to the normal building wiring. It must be manually initiated by by flipping a switch after the POCO power is gone (and presumably after rapid shutdown has occurred. And it supplies power only to a dedicated output which is not powered during periods of normal grid interactive operation.
    But to allow it to operate in conjunction with panel-level shutdown would appear to require a way to disable the panel level shutdown once it has automatically occurred.

    The 2017 rapid shutdown requirements seem to be more of a problem for a hybrid inverter system with batteries which is designed to automatically fall back using a transfer switch to battery powered operation with battery charging from the PV.

    Comment


      #3
      Talked to a helpful and knowledgable guy at Tigo, and he comfirmed: It is impossible to support 2017 rapid-shutdown requirements and still have SPS functionality. I believe this means that to have SPS in a legal system, the system must be exempt from rapid-shutdown requirements. This means ground-mount (or possibly carport-mount, a grey area).

      Comment


        #4
        It should not be impossible from a code and engineering standpoint. But if they (SMA and Tigo) have not engineered it that way, I doubt anyone else has, yet.

        For systems with any kind of backup capability, complying with rapid shutdown from a code point of view is mainly a matter of adequate signage and a suitable location for a switch. There may be design constraints and extra material and labor costs, but it usually wouldn't require switching technologies.

        Comment


          #5
          Originally posted by jaggedben View Post
          For systems with any kind of backup capability, complying with rapid shutdown from a code point of view is mainly a matter of adequate signage and a suitable location for a switch.
          I asked the guy about that - simply having a switch on the high-voltage DC - and he said it still wouldn't comply because there'd still be high voltage between the panels and the switch. Unless the panels are less than 10ft from the house, there might not be a suitable location for the switch: either too far from the house, or live conductors too far from the panels. Only problem with that is, I thought he told me the way the RS works that the inverter stops letting DC current flow and then the TS4 MLPEs shut down.

          Anyhow, as I discuss in my other thread (on ground-mount) I'm just gonna punt and put the inverter with the panels, and run AC to the house.
          In general, I'll have greater wiring losses, because the DC voltages are typically higher than 240; but since I have significant shading issues, there may be times that at least one of my strings is down towards the inverter's minimum, more like 100.

          Comment


            #6
            I would hope that there is some provision for hybrid systems with a transfer switch and batteries. If not, then the code has managed to paint itself into a corner.
            Or is it now only legal to have a standalone backup power system if it does not have the ability for grid interactive operation?

            Comment


              #7
              Originally posted by GoldDigger View Post
              I would hope that there is some provision for hybrid systems with a transfer switch and batteries. If not, then the code has managed to paint itself into a corner.
              Or is it now only legal to have a standalone backup power system if it does not have the ability for grid interactive operation?
              I don't think the code has painted itself into a corner at all, although manufacturers aren't necessarily keeping up.

              ​​​​​​​Keep in mind that the 2017 basically requires MLPE for modules. So whether you the MLPE meets rapid shutdown because it is inherently engineered to (micro-inverters) or hard programmed to (optimizers), a switch in the conduit coming down from the roof could meet the requirements. Note that while a lot of manufacturers have now engineered their stuff to shut down when the grid connection goes dead, the code allows another option. Meanwhile, there is no rapid shutdown requirement for conductors supplied by batteries or other energy storage systems.

              Comment


                #8
                Originally posted by RustyShackleford View Post
                I asked the guy about that - simply having a switch on the high-voltage DC - and he said it still wouldn't comply because there'd still be high voltage between the panels and the switch. Unless the panels are less than 10ft from the house, there might not be a suitable location for the switch: either too far from the house, or live conductors too far from the panels.
                You know, you've confused this discussion by initially referring to 2017 requirements and then throwing in the 10ft rule which only applies to 2014 requirements. Also I can't tell if you're talking about the general situation or your particular ground mount situation. Under 2014 requirements, you are fine having up to 10ft between the array and the switch, period. Under 2017, with your ground mount situation, you're fine having a switch at the place where the conductors meet the building. Having high voltage between the array and the switch doesn't run afoul of requirements in either of those situations.

                Only problem with that is, I thought he told me the way the RS works that the inverter stops letting DC current flow and then the TS4 MLPEs shut down.
                I don't know how Tigo works these days, and it's been years since the one time I worked with it. But back then they had a little box with a red button on it that essentially performed the rapid shutdown. Evidently they've done something else now to make rapid shutdown be initiated by the grid side disconnect, is that right? Anyway the code does not require you to do it that way. Just a product limitation.

                With SolarEdge optimizers, if they're disconnected from the inverter they stop producing voltage, period. So that's one basically one way to achieve rapid shutdown, unless the AHJ starts to question how long it takes for conductors between inverter and switch to drop voltage. In any case, even if that method isn't available, you could still label the the AC breaker for the SolarEdge inverter as the rapid shutdown switch, or find a way to bring the inverter AC wiring outside to a disconnect.

                For DC coupled SolarEdge systems, I just label the DC switch on the inverter as the Rapid Shutdown Switch. Since both the optimizers and the batteries shut down voltage automatically when disconnected from the inverter, this generally works fine. Only problem is that under 2017 the inverter would need to be outside.

                Comment


                  #9
                  Originally posted by jaggedben View Post

                  I don't know how Tigo works these days, and it's been years since the one time I worked with it. But back then they had a little box with a red button on it that essentially performed the rapid shutdown.
                  The Tigo boxes are one per module and they disconnect them individually from the DC wiring. They monitor a "stay alive" signal and disconnect the modules when it goes away.

                  Originally posted by jaggedben View Post
                  With SolarEdge optimizers, if they're disconnected from the inverter they stop producing voltage, period.
                  Not exactly. They reduce the voltage to 1V per module when the inverter shuts down, and be advised that with the larger SolarEdge inverters it is possible to have a string long enough that that voltage is enough to violate RS.

                  Comment


                    #10
                    Originally posted by jaggedben View Post
                    You know, you've confused this discussion by initially referring to 2017 requirements and then throwing in the 10ft rule which only applies to 2014 requirements. Also I can't tell if you're talking about the general situation or your particular ground mount situation. Under 2014 requirements, you are fine having up to 10ft between the array and the switch, period. Under 2017, with your ground mount situation, you're fine having a switch at the place where the conductors meet the building. Having high voltage between the array and the switch doesn't run afoul of requirements in either of those situations.
                    Thanks for clarifying, I didn't realize that. I'm definitely under 2017. And I'm talking about my particular installation (but hopefully the thread can be of general interest to others at some point).

                    So I guess my choices are:
                    1. Manage to have no more than 5ft of high-voltage DC wiring inside the house, or encase some portion of it in concrete (as suggested in other thread).
                    2. Put shutoff switch where high-voltage DC enters the house. This will be conveniently (for firefighters) very close to the meter.
                    3. Put the inverter with the panels, and bring 240 vac to the building.

                    In any case, I'm going to have to sign the waiver authorizing SMA to turn off inverter feature where loss of the grid causes the optimizers to be shut off.

                    I'm leaning towards #3. Generally lower voltages, so more wiring loss; except for the fact that there are significant shading issues, so some of the strings may at times be generating voltages well below 240. At least bringing 240vac in, I know what I'm dealing with, and it'll be 50ft max, so 8awg (and perhaps 10awg) will be fine.

                    Sorry for the way these two threads have blended together; I guess I should've just made one thread.



                    Comment


                      #11
                      It's convenient for grid-tied PV to have rapid shutdown initiated by loss of grid AC, no extra switch and it happens automatically. I would guess the waiver is just to make sure you know you have to install the RSS switch yourself. There is no code requirement that the inverter controls rapid shutdown or that loss of grid AC initiates rapid shutdown.

                      Comment


                        #12
                        Originally posted by RustyShackleford View Post
                        Interested in using one of SMA's new "-41" series inverters. After searching SMA's webpages, I am still mightily confused about when and how SPS "secure power" can peacefully co-exist with 2017 rapid-shutdown requirements.

                        Is it even possible ? What specifically is required ? TS4-R-O optimizers on every panel ? An explicit disconnect for the high-voltage DC ?
                        We install lots of SMA inverters I think they are the -40 ones, the rapid shutdown is a DC relay box that generates its own 12VDC 'stay alive signal' on the roof and as far as I can tell has nothing to do with the inverter. Secure power works just fine. The rapid shutdown has failed on me 2 times leaving 480VDC ON after the button is pressed.
                        I dislike having the DC disconnect built into inverters as it makes a inverter swap out either 'hot work' or a roof access job.
                        For your system I would go with 2 or 3, I really like having an old fashioned knife switch DC disconnect like you would on a service.
                        Comments based on 2017 NEC unless otherwise noted.

                        Comment

                        Working...
                        X