Grounded C.T. Delta Transformer

[Headdog]

On the system in question there will be a 480 delta to 240 delta transformer (with one center-tapped winding). Naturally, there are 230 VAC 3 phase loads on the secondary side and a few 230 VAC single phase loads spread around the two windings that are not center tapped (phases A to B and B to C).

If my jargon is correct these loads include the "bastard leg" (phase B in this case) which should be marked as orange since it is the highest voltage above ground.

That leaves the final secondary winding (C to A) which includes the center tapped winding which is also grounded via the ground rod. All of the 115 VAC loads are attached here (phase A or C to ground) and maybe even some 230 single phase loads (C to A) as needed to "balance" out the overall loading. (I also realize the 115 VAC loads are practically limited to about 30% of the transformer rating to keep the unbalanced loading from overheating the transformer; that part I think I understand)

That leaves my stupid question(s) . . . Why is it OK to ground this center tap when it is not really the theoretical neutral point (which should really exist at the center of all three "delta" phases)? Are there practical limitations on this set up? For example it is only ok for transformers up to a certain size, or is it always OK and the unbalanced currents just flow through the earth? Hope this all makes some sense and someone can explain it to me! Thank you. -- HD

 

[winnie]

Remember that current isn't trying to get to 'earth', and it isn't trying to get to 'neutral'. Current simply flows in closed circuits from the source and back to it.

You can take _any_ single terminal of a transformer and ground it, and have a functioning system. You can even leave the system totally ungrounded, and it will function just fine. Grounding and bonding do nothing the alter the voltage of one system conductor to another, and do nothing to alter the current flowing in the system conductors (unless there is a fault). The only thing that grounding alters is the voltage of the system conductors relative to ground, and the only thing that bonding alters is the current flow during a fault condition.

In the case of your high-leg system, current leaving terminal A goes through different loads, arriving at terminal C, B or the AB centerpoint. Each different terminal is at a different relative voltage and phase angle, but each is a possible destination for the current flowing through A. The bonding of the centerpoint doesn't change this. Very little current should be flowing along the GEC/bond conductor from the centerpoint; just a small amount of capacitively coupled current.

Code requires that the terminal conductor selected for grounding be the one that minimizes the voltage to ground of the other conductors; but in the case of three wire three phase systems, the grounded conductor can be one of your phases; this is called a 'corner grounded' system.

-Jon

 

[jim dungar]

Why is it OK to ground this center tap when it is not really the theoretical neutral point...?

Per code, we bond a reference point to ground.
The fact that it is a neutral in some circuits is coincidence, because in other cases it may be a phase conductor.

 

[LarryFine]

If my jargon is correct these loads include the "bastard leg" (phase B in this case) which should be marked as orange since it is the highest voltage above ground.

That is correct, and not only in the service, but every place such a conductor is accessible: every disco, every J-box, every load wiring connection, etc.

That leaves my stupid question(s) . . . Why is it OK to ground this center tap when it is not really the theoretical neutral point (which should really exist at the center of all three "delta" phases)?

Because a real bonding has a much lower impedance than a theoretical bond does. An ungrounded Delta will measure as your theory suggests with a high-impedance voltmeter because the capacitive impedance is lower.

As soon as you use a low-impedance tester, that 'phantom' voltage will collapse. Every line will test at or near zero, yet your line-to-line voltages will be correct. The greatest voltage is dropped across the greatest impedance.

Think of a high-leg Delta from the center-tapped secondary's perspective, which is exactly the same thing as a 120/240v 1ph supply; the neutral point is obvious. Adding either one (open Delta) or two more transformers doesn't change that.

Are there practical limitations on this set up? For example it is only ok for transformers up to a certain size, or is it always OK and the unbalanced currents just flow through the earth?

Can't answer the first part with any real authority, but I believe the answer is "no;" but I can respond to the second part:

Why does the current want to flow into the earth, generally speaking? Only because we opt to ground one of the circuit conductors. The voltage to ground is whatever the difference is between the terminal we grounded and the one we're measuring.

That current is only striving to return to the grounded supply terminal through whichever ground connections we (and the POCO) make: the electrodes. A power source only becomes a wire-to-ground shock hazard because we ground another wire.

Every transformer (again, both ours and the POCO's) is a separate source, and the current only attempts to return to that source's grounded terminal. Because we interconnect our system neutrals through grounds, the earth as a whole acts as a return conductor of sorts.

But, non-grounded supplies, like the POCO's ungrounded Delta services and our ungrounded SDS secondaries, don't really create a reliable line-to-ground shock hazard. They also don't trip breakers when the first ground-fault occurs. The line-to-line hazard is the same.

Added: The only reason the high leg has a higher voltage to ground is because that leg has a higher voltage to the terminal we ground.

 

[Headdog]

Remember that current isn't trying to get to 'earth', and it isn't trying to get to 'neutral'. Current simply flows in closed circuits from the source and back to it.

You can take _any_ single terminal of a transformer and ground it, and have a functioning system. You can even leave the system totally ungrounded, and it will function just fine. Grounding and bonding do nothing the alter the voltage of one system conductor to another, and do nothing to alter the current flowing in the system conductors (unless there is a fault). The only thing that grounding alters is the voltage of the system conductors relative to ground, and the only thing that bonding alters is the current flow during a fault condition.

In the case of your high-leg system, current leaving terminal A goes through different loads, arriving at terminal C, B or the AB centerpoint. Each different terminal is at a different relative voltage and phase angle, but each is a possible destination for the current flowing through A. The bonding of the centerpoint doesn't change this. Very little current should be flowing along the GEC/bond conductor from the centerpoint; just a small amount of capacitively coupled current.

Code requires that the terminal conductor selected for grounding be the one that minimizes the voltage to ground of the other conductors; but in the case of three wire three phase systems, the grounded conductor can be one of your phases; this is called a 'corner grounded' system.

-Jon

Thank you Winnie for the information. I guess I am "fretting" about nothing. When you are working on your first installation you probably tend to be a bit of a "worry wort"! I just am trying to avoid doing something obviously dangerous.

 

[Headdog]

Per code, we bond a reference point to ground.
The fact that it is a neutral in some circuits is coincidence, because in other cases it may be a phase conductor.

Thank you Jim, I am starting to see the "light"!

 

[Headdog]

Because a real bonding has a much lower impedance than a theoretical bond does. An ungrounded Delta will measure as your theory suggests with a high-impedance voltmeter because the capacitive impedance is lower.

As soon as you use a low-impedance tester, that 'phantom' voltage will collapse. Every line will test at or near zero, yet your line-to-line voltages will be correct. The greatest voltage is dropped across the greatest impedance.

Think of a high-leg Delta from the center-tapped secondary's perspective, which is exactly the same thing as a 120/240v 1ph supply; the neutral point is obvious. Adding either one (open Delta) or two more transformers doesn't change that.

Larry, thank you for your message. I needed a few days to think about it!

I have to confess I get a little lost trying to understand this section. I think I understand that a high impedence "voltmeter" would see the "phantom" voltage (from the center tap to ground; when not grounded) but that the voltage would "collapse" under any current demand (since no closed circuit current path really exists) in this ungrounded (example) system. The "because the capacitive impedence is lower" part I don't follow. Why is it lower (because of the voltmeter, the grounded secondary CT, or something else)? Also, "the greatest voltage is dropped across the greatest impedence" statement confuses me since I don't understand it as well. If the secondary CT is grounded, the impedence (at least to ground) is zero (the lowest possible impedence) so where are you talking about dropping "the greatest voltage"?

I freely admit I don't understand these two points while at the same time I believe you know what you are telling me. I just need a little more detailed explaination so I can understand it!

Every transformer (again, both ours and the POCO's) is a separate source, and the current only attempts to return to that source's grounded terminal. Because we interconnect our system neutrals through grounds, the earth as a whole acts as a return conductor of sorts.

But, non-grounded supplies, like the POCO's ungrounded Delta services and our ungrounded SDS secondaries, don't really create a reliable line-to-ground shock hazard. They also don't trip breakers when the first ground-fault occurs. The line-to-line hazard is the same.

Ok for this comment string I think I understand that the first "ground-fault" or line in contact with ground does not trip breakers becuase there is no closed circuit (i.e. it is "ungrounded"). It takes a 2nd "ground-fault" to actually trip the breakers. I guess I had not really throught about that case. Seems kind of scary to think this should also be the case even for the POCO's transmission lines (in an ungrounded system).

I still wouldn't want to get near the first ground-fault wire! Even though it might not be lethal, it would scare the you know what out of me! -- HD
(2nd answer here)

 

[LarryFine]

The "because the capacitive impedence is lower" part I don't follow. Why is it lower (because of the voltmeter, the grounded secondary CT, or something else)?

Compared to a high-impedance meter, the capacitive coupling is lower (meaning capable of passing more current); otherwise, the meter wouldn't show a phantom voltage.

Also, "the greatest voltage is dropped across the greatest impedence" statement confuses me since I don't understand it as well.

In a series circuit with non-matching impedances, the higher voltage will appear across the higher impedance. For instance:

If you have a 10K resistor and a 1Meg resistor in series and apply a voltage across them, most of the voltage will be measured across the 1Meg resistor. Now, for our discussion:

If the higher impedance is the meter, you'll read the phantom voltage. If the higher impedance is the capacitive coupling, you'll read very little, if any, phantom voltage.

If the secondary CT is grounded, the impedence (at least to ground) is zero (the lowest possible impedence) so where are you talking about dropping "the greatest voltage"?

Okay, back to the original question: If you have no grounded circuit conductor:

~ A high-impedance meter will read the same voltage from each line to ground, close to what it would be if the supply system was a grounded Y.

~ A low-impedance meter will read very little voltage to ground. As you move the test lead from line to line, the normally-equal voltages will shift.

~ If you attach a low-impedance tester between A phase and ground, a high-impedance meter would read the line-to-line voltage on B and C phases.

I freely admit I don't understand these two points while at the same time I believe you know what you are telling me. I just need a little more detailed explaination so I can understand it!

That's what we're here for! :smile: Let me know if this helps, and don't hesitate to ask more questions.



But, wait, there's more!

Ok for this comment string I think I understand that the first "ground-fault" or line in contact with ground does not trip breakers becuase there is no closed circuit (i.e. it is "ungrounded"). It takes a 2nd "ground-fault" to actually trip the breakers.

That's exactly why ungrounded systems are used; to supply loads that should not be unexpectedly shut down if at all possible.

I still wouldn't want to get near the first ground-fault wire! Even though it might not be lethal, it would scare the you know what out of me!

Technically speaking, the faulted conductor is the "safe" one to touch. Grounding one conductor locks the others to a set voltage to ground. In that sense, grounding a supply makes it dangerous.

However, after years of both grounded and non-grounded systems, it was determined that the advantages of grounding outweigh the disadvantages, except where the advantages of not grounding prevail.

 

[Headdog]

Thank you "LarryFine". I don't have any more questions for now.