interconnecting different voltage systems without galvanic isolation ok per NESC NEC?

[Electric-Light]

I was reading in another thread that this is one of the methods used by some solar installers to interconnect 208/120 public utility network with 480/277 energy source using shared winding transformer that do not provide galvanic isolation between two sides.

If due to accidents or equipment failure and one of the winding is separated from the neutral, is whatever safeguard built into the inverter considered acceptably sound safeguard to positively prevent high legging of the broken points of the 208/120 side system to 277v in the view of NESC and NEC? I know that islanding protection is supposed to stop exporting power but I am not sure how it would respond to something like this when the neutral and other two phases remain solidly grounded and the break off of the affected winding from the neutral can be complete or intermittent.

Wires getting ripped out and transformer getting thrown pulled is reasonably probable during a storm where a power surge could also take place. Could this leave open the possibility of one of the insulated gate transistors shorting and back feeding the islanded broken wire floating at 277v?

 

[SolarPro]

A listed interactive inverter will simply cease to export power in this scenario.

 

[jaggedben]

I was reading in another thread that this is one of the methods used by some solar installers to interconnect 208/120 public utility network with 480/277 energy source using shared winding transformer that do not provide galvanic isolation between two sides.

What other thread are you referring to? I've never heard of anything like the transformer configuration you have in your drawing. (Is the XF supposed to be part of the inverter or a separate component?)

...

Wires getting ripped out and transformer getting thrown pulled is reasonably probable during a storm where a power surge could also take place. Could this leave open the possibility of one of the insulated gate transistors shorting and back feeding the islanded broken wire floating at 277v?

Seems to me like an unlikely sequences of events for equipment that has passed a UL listing process. You seem to be positing that inverter electronics which are designed to export X amount of current will be catastrophically damaged by a surge that carries less than X*1.25 current, such that the OCPD won't open the circuit, thus leaving raw DC voltage exposed the the grid. I think the safety factors in inverter design are probably not so tight. Also, you know, many inverters have relays that physically disconnect the inverter electronics from the grid when grid voltage or frequency is out of range.

 

[Electric-Light]

What other thread are you referring to? I've never heard of anything like the transformer configuration you have in your drawing. (Is the XF supposed to be part of the inverter or a separate component?)

...

Seems to me like an unlikely sequences of events for equipment that has passed a UL listing process. You seem to be positing that inverter electronics which are designed to export X amount of current will be catastrophically damaged by a surge that carries less than X*1.25 current, such that the OCPD won't open the circuit, thus leaving raw DC voltage exposed the the grid. I think the safety factors in inverter design are probably not so tight. Also, you know, many inverters have relays that physically disconnect the inverter electronics from the grid when grid voltage or frequency is out of range.

The thread was called Inverter Step-down transformer. I thought it wasn't wise or allowed for the issues I mentioned here and someone mentioned some solar companies use the installation practice of using an auto transformer fed on ends and neutral from an inverter designed for 480Y/277v power system and connect to the 208Y/120 power system through the knuckles and the neutral (see picture). I am positing inverter was not intended to be used in this manner and would present the hazard that certain conditions can make 277v appear on 120v as a result of doing something beyond its intended purpose.

Seems to me like an unlikely sequences of events for equipment that has passed a UL listing process.

It's uncertain if the components are UL listed or code compliant to use in that matter. The pieces are not separate components of a UL listed product. The thread that I can't seem to find seemed to say its a 480/277 inverter used with an unisolated transformer to interconnect to the utility's 208/120 system

 

[GoldDigger]

The thread was accidentally deleted. Feel free to recreate or continue it.
The main concern raised was that lifting the neutral or braking the lower section of the winding will allow the GTI to put 277 on one of the 120V phases.
The developing answer was that two things would happen at more or less the same time.
1. The GTI matched the grid voltage and delivers reverse current at that voltage.
Rather than drive the grid at 277V the inverter would deliver its maximum current at 120V.
2. The anti islanding mechanism of the GTI will detect that the "grid" voltage at its terminals is out of tolerance and shut down the entire inverter.
There is no way that it could put high voltage on the line, just as there is no way that it could put normal voltage on an open grid connection. The answer might be different if a non interactive generating system (such as a separate utility source) were involved.

mobile

 

[Electric-Light]

The thread was accidentally deleted. Feel free to recreate or continue it.
The main concern raised was that lifting the neutral or braking the lower section of the winding will allow the GTI to put 277 on one of the 120V phases.
The developing answer was that two things would happen at more or less the same time.
1. The GTI matched the grid voltage and delivers reverse current at that voltage.
Rather than drive the grid at 277V the inverter would deliver its maximum current at 120V.
2. The anti islanding mechanism of the GTI will detect that the "grid" voltage at its terminals is out of tolerance and shut down the entire inverter.
There is no way that it could put high voltage on the line, just as there is no way that it could put normal voltage on an open grid connection. The answer might be different if a non interactive generating system (such as a separate utility source) were involved.

mobile

N, L1, L2 are still solidly connected. Somehow the knuckle/middle tap on L3 and N and L3 common side develops loose connection.

Since the inverter is 480/277, it expects to see 277 between N and L3 (and N is still connected, as are the other two phases). Winding 3 becomes a series inductor. If you had a one winding transformer tapped at G/N, 6.6, 277v, you get 6.6v between N and 6.6 but that is conditional to it connected to N. If you lift the N, the open circuit voltage at either N or 6.6 is 277v to ground.

It won't have a lot of amps but its basically 277v as far as shock hazard is concerned.

Do any manuals show instructions supporting the connections in which the higher voltage is pushed down into a lower system outside your system? as opposed to within the system into your transformer.

 

[GoldDigger]

You are totally ignoring my point that the GTI does not independently generate a 277 volt waveform. It will instead match the 277 or 120 which is present at its output, coming from the grid.

mobile

 

[Electric-Light]

You are totally ignoring my point that the GTI does not independently generate a 277 volt waveform. It will instead match the 277 or 120 which is present at its output, coming from the grid.

mobile

Oh so you'd have to have like a 3 phase motor on the 3 phase side alongside the inverter for this to become an issue.

 

[GoldDigger]

Oh so you'd have to have like a 3 phase motor on the 3 phase side alongside the inverter for this to become an issue.

Also an interesting thought experiment.
That might get interesting if you list just that phase from the grid, although I think anti islanding would still shut the inverter down because of the phase and amplitude anomalies.
If you lost the autotransformer winding I would expect that OCPD would open the motor to grid connection. And the motor case would be just as bad without the inverter.
Putting a motor on the inverter side of an autotransformer would be a design mistake.

mobile