Yeah, and the code was amended to say "between any two wires" instead of "between any wire and ground". I designed a couple of bipolar arrays back in the day, but that was several companies ago for me. I don't know if the systems survived until now, but that company did not.There is, in fact, in inverter topology that uses bipolar arrays with a positive, negative, and a (grounded) neutral conductor. The purpose was essentially to have a 1200V array that didn't exceed 600V to ground. I think it became a failed idea economically and I haven't heard of it for a while, but I presume there are a number of installed systems still operating. In that case the neutral would fit the Article 100 definition and would not count as a CCC.
It was added in NEC2017, and unfortunately, it refers to two wildly different kinds of systems. It's a good term, but we really could use another term to tell them apart. The vast majority of any PV the way we've done it for current century, is likely a functionally-grounded system in some form or another, even if not called that at the time.Not sure what "functionally grounded" actually means in terms of what the inverter is doing, but if each inverter functionally grounds one conductor, you might need to be able to configure one inverter to functionally ground its -, and the other to functionally ground its +.
Does anyone even make transformer coupled PV inverters anymore?It was added in NEC2017, and unfortunately, it refers to two wildly different kinds of systems. It's a good term, but we really could use another term to tell them apart. The vast majority of any PV the way we've done it for current century, is likely a functionally-grounded system in some form or another, even if not called that at the time.
Type 1 is where a polarity is normally grounded thru a GFCI device, and could become ungrounded in the event of a fault. Prior to 2017, we treated this as if it were a grounded system, with the normally-grounded polarity treated just like a neutral is normally treated, except with a warning label about what happens if during a ground fault. The AC waveform is initially monopolar, but becomes centered on ground after passing through a (usually) built-in transformer.
Type 2 is where the AC side is connected to a grounded grid, and the DC side is non-isolated. The two polarities are at equal and opposite voltages to ground, in order to each generate their respective halves of the AC waveform. I'd prefer to call this a symmetry-grounded system. Prior to 2017, we called these ungrounded systems, even though the voltage-to-ground is constrained, rather than free to drift like a true ungrounded system.
Only in the scale of central inverters. And even then, you may also see the isolation externalized to the job of your medium-voltage transformer, rather than an integratted transformer in the inverter, since the expectation is that you'll already have a medium voltage transformer. The drawback of a design like this, is that it makes it much harder to retrofit a new inverter, if such a transformer required an exotic AC voltage.Does anyone even make transformer coupled PV inverters anymore?
I gather than modern string inverters are of all the above type? This would preclude the DC MWBC configuration I asked about, with two inverters.Type 2 is where the AC side is connected to a grounded grid, and the DC side is non-isolated. The two polarities are at equal and opposite voltages to ground, in order to each generate their respective halves of the AC waveform. I'd prefer to call this a symmetry-grounded system. Prior to 2017, we called these ungrounded systems, even though the voltage-to-ground is constrained, rather than free to drift like a true ungrounded system.
Some will and some won't.I gather than modern string inverters are of all the above type? This would preclude the DC MWBC configuration I asked about, with two inverters.
But does this also mean that modern string inverters won't work on an ungrounded AC system, e.g. an ungrounded 480V delta at an old factory?
Thanks,
Wayne
I can't say all for certain, but certainly the vast majority of string inverters I've seen within about the last 10 years, will run with what I've called a Type 2 functionally grounded system on the DC side. Or as I'd also call, a symmetry-grounded system.I gather than modern string inverters are of all the above type? This would preclude the DC MWBC configuration I asked about, with two inverters.
But does this also mean that modern string inverters won't work on an ungrounded AC system, e.g. an ungrounded 480V delta at an old factory?
Thanks,
Wayne
Yeah, and they are weird. It's still used in multi-MW central inverters where the inverter and an MV transformer come preassembled as a unit on a PCS skid. They often still use a grounded DC conductor too. The AC voltages are not standardized, I see a number of voltages used in the 600-700V range.Does anyone even make transformer coupled PV inverters anymore?
This was just another example of the dodgy things that go into the NEC sometimes to make things more palatable for AHJs. AHJs demanded grounded DC and this is what was given them. Everyone assured them it was a grounded system, until there was an alternative and then it was not considered grounded and outside of some large ground mounted installations it is no longer used.Type 1 is where a polarity is normally grounded thru a GFCI device, and could become ungrounded in the event of a fault. Prior to 2017, we treated this as if it were a grounded system, with the normally-grounded polarity treated just like a neutral is normally treated, except with a warning label about what happens if during a ground fault. The AC waveform is initially monopolar, but becomes centered on ground after passing through a (usually) built-in transformer.
