Delta to Delta Connection

[Steven Sallwasser]

I have an installation where there is a wye-delta 480V to 2500V transformer connected to a delta-wye 2500V to 480V transformer. None of the phase conductors at either end of the 2500V line are connected to ground, so they are effectively floating from what I can tell. What is to prevent one of the phase conductors from floating above the cables rating? What is the easiest way to prevent a inadvertent fault?

 

[ggunn]

I have an installation where there is a wye-delta 480V to 2500V transformer connected to a delta-wye 2500V to 480V transformer. None of the phase conductors at either end of the 2500V line are connected to ground, so they are effectively floating from what I can tell. What is to prevent one of the phase conductors from floating above the cables rating?

Relative to what?

 

[GoldDigger]

I have an installation where there is a wye-delta 480V to 2500V transformer connected to a delta-wye 2500V to 480V transformer. None of the phase conductors at either end of the 2500V line are connected to ground, so they are effectively floating from what I can tell. What is to prevent one of the phase conductors from floating above the cables rating? What is the easiest way to prevent a inadvertent fault?

Is the cable shielded, in open air or in a racway?

The voltage on the phase conductors can easily (via intermittent arc faults, for example) get pumped up relative to ground to many times the 2500 Volt line-to-line voltage. But in the absence of something like that it will take only a very large leakage resistance to ground to keep the line to ground voltage within control.
The NEC simply states that for insulation purposes, in an ungrounded delta, the line to ground voltage shall be assumed to be equal to the line-to-line voltage, as it would be in the event of a ground fault.

But take a look at 250.21 and tell me whether you think that you are allowed to operate an ungrounded system subject to the NEC at 2500 volts.
I would expect that at a minimum you would need to corner ground or use a high impedance ground.

 

[david luchini]

But take a look at 250.21 and tell me whether you think that you are allowed to operate an ungrounded system subject to the NEC at 2500 volts.
I would expect that at a minimum you would need to corner ground or use a high impedance ground.

250.21 doesn't apply to 2500V systems. There is no NEC requirement for grounding 2500V systems unless it is supplying mobile or portable equipment. 250.20(C) PERMITS grounding of AC systems 1kV and over.

 

[Steven Sallwasser]

Delta to Delta Connection

Is the cable shielded, in open air or in a racway?

The voltage on the phase conductors can easily (via intermittent arc faults, for example) get pumped up relative to ground to many times the 2500 Volt line-to-line voltage. But in the absence of something like that it will take only a very large leakage resistance to ground to keep the line to ground voltage within control.
The NEC simply states that for insulation purposes, in an ungrounded delta, the line to ground voltage shall be assumed to be equal to the line-to-line voltage, as it would be in the event of a ground fault.

But take a look at 250.21 and tell me whether you think that you are allowed to operate an ungrounded system subject to the NEC at 2500 volts.
I would expect that at a minimum you would need to corner ground or use a high impedance ground.

Based on your statements, I am correct in saying there is nothing to prevent the phase conductors from floating and the cheapest solution would be to ground one of the phase conductors which I believe is typically the B phase.

I did not design the installation in question. I discovered it when collecting data for an Arc-Fault study and want to make a recommendation to the owner, but want to make sure I am on solid ground.

I also found in the facility breakers that were not rated for the available fault current with fuses in series with the breakers that were downstream of the breaker but less than a foot of cable downstream. I suppose the fuses will limit the fault current through the breaker, but the fuse should really be ahead of the breaker in order to protect it. I told the client the fuse is providing protection.

 

[Smart $]

I have an installation where there is a wye-delta 480V to 2500V transformer connected to a delta-wye 2500V to 480V transformer. None of the phase conductors at either end of the 2500V line are connected to ground, so they are effectively floating from what I can tell. What is to prevent one of the phase conductors from floating above the cables rating? What is the easiest way to prevent a inadvertent fault?

Based on your statements, I am correct in saying there is nothing to prevent the phase conductors from floating and the cheapest solution would be to ground one of the phase conductors which I believe is typically the B phase.

...

G is correct, though implied.

Relative to what?

The voltage (difference in potential) will be between conductors. There will be some reactive coupling to the proximate environment, but not enough to stress the insulation properties of the cable.

 

[beanland]

Coupling

Coupling

Under normal circumstances, the capacitance to ground of transformers and wire will attempt to keep the phase voltages relatively the same in reference to ground/earth. It would be normal to use 5kV conductors on this system so that an unintentional ground will not exceed the conductor rating. However, ungrounded systems of this type are subject to arcing faults because of the leakage current through system capacitance. Such arcing faults, if the current is high enough, will lead to cable damage and fires. It is good practice, to provide some kind of ground detection (not the same as ground fault). There are many commercially available systems that monitor the resistance to ground of the conductors and set alarms if resistance is low, indicating an insulation failure. There is no requirement to ground these systems using corner (IEEE recommends against) or impedance (grounding transformer) systems. Also, if there are surge arresters on the 2500V system, make sure they are rated to handle the line-line voltage continuously.