establishing an un-grounded system

[triplstep]

This is a hypothetical scenario.

What method would be the most effective in establishing and un-grounded 480 system? The build (ing) is being fed with a 480 wye, and the kva is plenty capable of handling the load. Would a transformer, or a generator, be the best choice?

Can this be accomplished with the use of a transformer fed with a grounded system?

Are there any Ed MacLaren drawings out there to help me wrap my head around the answer?

 

[jim dungar]

Do you want the entire facility to be ungrounded or just a portion?

Why do you want to go ungrounded instead of high-resistance grounded?

 

[winnie]

An ungrounded system is simply _any_ galvanically isolated electrical system in which _no_ intentional connection has been made between any system conductor and earth or any ground bonded conductor.

Most commonly this is simply a transformer secondary with no bonding jumper connected to the system, but it _could_ be a generator or even a battery bank. In other words, if you have _any_ electrically isolated transformer secondary, and simply do not connect your bonding jumper, then you have an ungrounded system. Most commonly this would use a delta secondary. You have additional requirements, eg. ground fault detection, no line-neutral loads, etc.

With that said, I would strongly urge you to consider using a conventional 480/277V wye secondary and 'high resistance grounding' rather than an ungrounded system. A high resistance grounded system provides the same benefits of continuity of service and (single) fault current reduction as an ungrounded system, can serve the same loads, and provides benefits in reducing transient system overvoltages. Additionally you have much more flexibility should you decide to ground the system in the future.

-Jon

 

[LarryFine]

Can this be accomplished with the use of a transformer fed with a grounded system?

The direct answer to your question is yes, a 480-D/480-D is the simplest way to derive ungrounded 480 from grounded 480.

 

[triplstep]

Thanks of the replies guys.

What I am trying to understand is the path of fault current, in an un-grounded system, derived from a grounded primary. I'm sure I could do a better job of presenting the question.

I'm not trying to engineer a system, just looking to understand some theory.

Doesn't 250.4 (B) require a path that eventually leads to ground of the primary, electrically connecting the primary and secondary?
Therefore giving a fault in the secondary a path to source.

Does a fault in the secondary look at either the primary or the secondary as source?

I hope the query is understandable

 

[winnie]

The concept of 'electricity seeking to return to its source' is talking about the fact that electric current must always flow in a closed circuit. The current flowing from one terminal of an energized transformer coil is trying to find its way back to the other end of that transformer coil. The primary has nothing to do with the path that the current takes; though of course it is necessary to provide the changing magnetic field which drives the current through the secondary.

The intent of grounding is not to connect the secondary of the system to the primary of the system; instead it is to connect the secondary of the system to all the various bits of 'non-current carrying' metal, eg. the conduit and boxes that enclose the conductors. As a side effect, if the primary is also a grounded system, the grounding of the secondary will establish some continuity, but this is incidental to the desired fault current path.

You can supply fully grounded system from an ungrounded system, and the fault current path of the grounded system would not be compromised. (For example, you could have a plant with 480V ungrounded distribution, supplying transformers with 480V delta primary and 208/120V wye secondaries, and ground the wye secondaries.)

A fault on the circuits connected to the secondary of a transformer 'look to' the secondary as the source.

In an ungrounded system, you explicitly and intentionally do not have a fault current path back to the transformer. This is done intentionally so that a fault will not cause high current flow and will not operate OCPD. The whole goal is 'continuity of service', in that a _single_ ground fault won't cause circuits or systems to shut down. Instead you 'quietly' detect the ground fault, shut down when safe and convenient, fix the problem.

-Jon

 

[LarryFine]

Trip, if one conductor of an ungrounded system becomes grounded, it fixes the voltage at zero to earth on that conductor, and the voltage to earth of the others at whatever the voltage difference is between that conductor and the others.

For example, an ungrounded 480v 3ph transformer secondary will measure approximately 277v to earth on each conductor (let's use 'phase') with a high-impednace voltmeter because of capacitance, which is the lowest impedance on such a system.

If one phase becomes grounded, whether intentionally or not, the voltage to earth on each of the other two instantly becomes 480v. Note that the relative voltages between conductors does not change with grounding; only the voltage to earth does.

 

[triplstep]

Jon and Larry, thanks for the replies.

The concept of 'electricity seeking to return to its source' is talking about the fact that electric current must always flow in a closed circuit. The current flowing from one terminal of an energized transformer coil is trying to find its way back to the other end of that transformer coil. The primary has nothing to do with the path that the current takes;

Using Kirchoff's current law, an electron leaving the secondary winding will never look to return on the primary winding.

As a side effect, if the primary is also a grounded system, the grounding of the secondary will establish some continuity, but this is incidental to the desired fault current path.

The reason being, the current only want to return to the secondary winding.

Trip, if one conductor of an ungrounded system becomes grounded, it fixes the voltage at zero to earth on that conductor, and the voltage to earth of the others at whatever the voltage difference is between that conductor and the others.

Got it.

If one phase becomes grounded, whether intentionally or not, the voltage to earth on each of the other two instantly becomes 480v. Note that the relative voltages between conductors does not change with grounding; only the voltage to earth does.

Though no danger to equipment or personal under this circumstance.



So if I've got this correctly, primary and secondary windings of transformers won't co-mingle electrons, whether both systems are grounded, ungrounded, or one of each.

And furthermore, to have an ungrounded system derived from a grounded system, I must not bond secondary equipment to the transformer cabinet, or any conductive parts that the primary is bonded to.

 

[tryinghard]

...And furthermore, to have an ungrounded system derived from a grounded system, I must not bond secondary equipment to the transformer cabinet, or any conductive parts that the primary is bonded to.

Well you don't bond one of the transformer windings (usually one of three) to frame [this is exactly why it's an ungrounded system]. The equipment grounding throughout is all the more important to have in an ungrounded system. If a short occurs it needs the complete equipment grounding path to enable the second short to open the circuit. Notice 250.4(B), 250.21(B). All the metallic conductive parts should be connected (effective fault path 250.4(3)-(5)) regardless of primary or secondary.

 

[LarryFine]

So if I've got this correctly, primary and secondary windings of transformers won't co-mingle electrons, whether both systems are grounded, ungrounded, or one of each.

Correct, except for an accidental contact between primary and secondary system conductors.

And furthermore, to have an ungrounded system derived from a grounded system, I must not bond secondary equipment to the transformer cabinet, or any conductive parts that the primary is bonded to.

Well, you won't bond any secondary conductor to the EGC/GEC system. Basically, treat every conductor like a nongrounded (i.e., hot) conductor.

Everything else, such as transformers, enclosures, conduits, etc., still gets bonded and grounded as usual. Only the main bonding jumper is omitted.



Added: Or, what Mr. hard said above.

 

[augie47]

Note in post 9 the statement that equipment grounding is very important with such a system. If all metallic points are not bonded, one phase "going to ground" on a 480 volt ungrounded system creates a voltage potential of 480 to other phases, especially dangerous if a second piece of unbonded equipment goes to ground on a separate phase.
As mentioned, one of the main advantages of such a system is to allow troubleshooting when there is a ground fault. Note that 250.21(B) requires such ungrounded systems to be monitors to advise on ground faults.