Ungrounded System Vs Corner Grounded System

[wwhitney]

A 3P3W system can be either corner grounded or ungrounded. For the corner grounded configuration, we are allowed to use distribution equipment with 2 pole OCPD that only disconnects the ungrounded conductors.

So suppose we have such a system, and change the supply from corner grounded to ungrounded, and add the necessary ground fault monitor. What new hazard exists with respect to the conductor without OCPD that didn't exist when that conductor was grounded?

In other words, why aren't we allowed to designate one conductor of a ungrounded 3P3W system as "always wired straight through" and skip the OCPD on that one conductor? The same way we are allowed to treat the grounded conductor of any grounded system.

[We could remove the 3 phase aspect of the question and just consider a 1P2W system with a voltage above 150V.]

Cheers, Wayne

 

[hornetd]

A 3P3W system can be either corner grounded or ungrounded. For the corner grounded configuration, we are allowed to use distribution equipment with 2 pole OCPD that only disconnects the ungrounded conductors.

So suppose we have such a system, and change the supply from corner grounded to ungrounded, and add the necessary ground fault monitor. What new hazard exists with respect to the conductor without OCPD that didn't exist when that conductor was grounded?

In other words, why aren't we allowed to designate one conductor of a ungrounded 3P3W system as "always wired straight through" and skip the OCPD on that one conductor? The same way we are allowed to treat the grounded conductor of any grounded system.

[We could remove the 3 phase aspect of the question and just consider a 1P2W system with a voltage above 150V.]

Cheers, Wayne

Because that former grounded conductor, which had no voltage to ground, now has 240 volts to ground and requires protection against ground fault that it didn't require before.

Tom Horne

 

[wwhitney]

Because that former grounded conductor, which had no voltage to ground, now has 240 volts to ground and requires protection against ground fault that it didn't require before.

It has no (physics-wise) defined voltage to ground, as the system is ungrounded. [The NEC definition of "voltage to ground" does say that the voltage to ground is equal to the L-L voltage for ungrounded systems, but that's a code fiction, not reality.] The ground fault monitor will tell us when this ceases to be true.

What I'm looking for is a fault scenario that will be worse or won't be cleared because on the ungrounded system we designated one conductor be "always wired straight through" without OCPD. I'm not seeing any. Since we do this all the time with grounded systems, the problem scenario would have to hinge on whether or not this circuit conductor without OCPD is grounded or not.

Cheers, Wayne

 

[ptonsparky]

Capacitive value?

 

[hornetd]

Maybe my only problem with it is I have never seen it done that way and having installed ungrounded Delta transformers before the requirement for Ground Detectors was actually written into the NEC quite a few of them ended up with only 3 pole breakers for protection. Even the Air Force didn't have 2 pole breakers with one conductor unprotected. They were always 3 pole. The Ground Electronic Engineering and Installation Agency (GEEIA) inspectors were often EEs. The floating Delta secondaries were often for medical refrigeration were a loss of power would have caused a loss of medication or a loss of medical samples. The Ground detection would not have been hard to make even with the remote monitoring needed so that the facilities unit could respond immediately with the trailer mounted generator already in tow. I never came to understand why the remote monitoring was done with thermostats in the refrigeration equipment instead of monitoring the power. Maybe both would be better. It takes a refrigerator a while to cool down after power is lost and monitoring the power would have given base facilities more time to intervene and restore the power or change out the refrigeration equipment. They never liked that second approach and when some doctor would try to tell them to do that they would tell them they were not qualified to handle medical samples. The fact is that all of use were very reluctant to handle unknown medical samples. Air Force people go all over the world. It's a little like the reason that our children go to school. It's an opportunity to become a carrier for every disease process that is active in the community and bring it right home to Mon and Dad.

Tom Horne

 

[wwhitney]

Maybe my only problem with it is I have never seen it done that way

Well, it's not allowed, as 240.4 requires overcurrent protection of every conductors, which is commonly understood to mean every ungrounded conductor.

So my question was a theory question about whether there is a reason it is not allowed on a ungrounded system. Basically drawing a comparison with the grounded system, and observing that I don't see how the "groundedness" of one conductor plays any role in the basic strategy of putting OCPD into all conductors except one.

Cheers, Wayne

 

[Jraef]

It has no (physics-wise) defined voltage to ground, as the system is ungrounded. [The NEC definition of "voltage to ground" does say that the voltage to ground is equal to the L-L voltage for ungrounded systems, but that's a code fiction, not reality.] The ground fault monitor will tell us when this ceases to be true.

What I'm looking for is a fault scenario that will be worse or won't be cleared because on the ungrounded system we designated one conductor be "always wired straight through" without OCPD. I'm not seeing any. Since we do this all the time with grounded systems, the problem scenario would have to hinge on whether or not this circuit conductor without OCPD is grounded or not.

Cheers, Wayne

You have to look at WHY you would want to go to an ungrounded system. The main reason people want ungrounded delta is to maintain production in the case of a ground fault on the system, basically because as you surmised, you go from ungrounded to corner grounded, so nothing shuts down. Sure, if the "straight through" phase is the one that goes to ground, it's no different than a corner grounded delta system. But in your scenario if you have only two phases with circuit protection and EITHER OF THOSE TWO goes to ground, now you have a corner grounded system, but with only ONE phase that has protection on it! Now things get really complicated as to how and why something might trip, meaning you have essentially defeated the reason you went to ungrounded delta in the first place.

 

[hornetd]

Jraef​

Bladergast and bedraggled you beat me to it. I was in the middle of composing that same question when the notice about your post popped up. That is the reason that the NEC now requires a ground detection system on ungrounded delta circuits. Once the first unintended ground occurs it MUST be cleared before a second one occurs. The first ground fault has made the system a corner grounded delta system. with 3 pole common trip breakers it does not matter on which phase the second fault occurs because the resulting current flow will open all 3 of the transformers output conductors.

Tom Horne

 

[wwhitney]

But in your scenario if you have only two phases with circuit protection and EITHER OF THOSE TWO goes to ground, now you have a corner grounded system, but with only ONE phase that has protection on it! Now things get really complicated as to how and why something might trip

I agree with this part of your post. But it's the complicated part that I'm interested.

With an ungrounded system, OCPD in all the conductors, the first ground fault effectively makes it a grounded system. With the ground fault monitors, the user is supposed to observe this has happened and resolve it forthwith.

But suppose they don't, and we had actually omitted OCPD in just one of the legs (consistently the same one), and one of the other legs faulted to ground. So now we have a grounded system, with OCPD in the grounded conductor, and with one ungrounded conductor without OCPD. What is the subsequent fault that this system could have that would be worse than a normal grounded system, where the conductor without OCPD is the grounded conductor and the ungrounded conductors all have OCPD?

Thanks,
Wayne

 

[LarryFine]

I'm thinking about it from the point of view of why we avoid OCP in the intentionally-grounded conductor unless it opens all conductors.

No conclusions yet, but I feel it's related.

 

[electrofelon]

Interesting question. I cannot come up with a resulting dangerous scenario. Basically we are used to the idea that for a grounded system with N conductors, we only need N-1 OCPD's to clear every type of fault. Doesnt the same apply to an ungrounded system?

 

[Tulsa Electrician]

I would think where we chose to ground the phase would be the difference.

What if it grounded on a 30 amp three phase branch corcuit with an #10 egc. And the service was 2000 amps and the available was 42K. I for one would not want to get in series with that.

Then this branch circuit originated from a 800 amp panel board with a 600 amp connected load.

The back ward GEC( ground faulted phase to the #10 egc) would be way under sized. Then you would have unwanted currents on the egc back to the service main.and then GES.
Then if had a second fault on a different feeder on another line which feeds another 800 amp panel from the 2000 amp mdp. I bet that #10 would not stand a chance.

Then that would clear the 30 amp ocpd and move the gec to the new faulted feeder egc unless the 30 amp so happened to be the non ocpd line. It would smoke the #10 egc and #10 phase wire all the way back if it did not clear the 800 amp feeder ocpd.. Probably catch fire any way. I bet the #10 would not be able to withstand the fault current in time to clear the 800 amp.

If the IA was at 30,000 amps on that 800 , most likely would not trip. Parallel #10 would not carry that long enough for the 800 to open. It would burn clear.

Remember the first faulted line has no over current protection all the way back to thru the 2000 amp MDP to the transformer.

Now we have those conductors burned open, cleared by fault. Some one comes to investigate and repair that 30 amp branch circuit conductors they assume is dead. After all that two pole 30 amp breaker is open. If they become in series with the first fault they become the second fault. Again why would the 800 open. Ooch!

Draw that out and see if that makes since.

On an ungrounded the first fault location is your GEC and the EGC becomes part of the the GES. This fault current will travel on the egc and has to withstand the fault. In the case the #10 from fault to terminal bar in panel where we may have 2- 500 mcm back to the 800 ocpd in the MDP. The terminal bar just so happens to be the first faulted line. The 30 amp two pole will trip. Now we have a corner grounded system by accident.

On a grounded system the GEC for the GES and sizing on the EGC if NEC followed can withstand the fault current and clear the ocpd.

As you read think about this.
If they used the terminal bar and two pole ocpd's and had wire type EGC. How would handle the bonding. Isolation to prevent unwanted currents on the EGC. What if it was as op stated change from corner grounded to ungrounded with no isolation. You would have ground fault every where untill isolated. Imagine in multiple locations where they did not follow phasing. You would would line to line faults every where if not isolated terminal bars.

Think goodness this is just a what if.

 

[Tulsa Electrician]

Kinda long

 

[hornetd]

I'm thinking about it from the point of view of why we avoid OCP in the intentionally-grounded conductor unless it opens all conductors.

No conclusions yet, but I feel it's related.

Leaving the corner grounded delta aside for a bit if a single overcurrent device in a grounded neutral system the Grounded Conductor of the circuit becomes energized by the voltage through the loads from the energized conductors. We expect that conductor to be no threat because it is grounded but now we have a possibly deadly set of white, or in other parts of the world black or blue, conductors that are now a hazard to our very lives.

About 3 decades ago I went to do a service call on a home that had once been served by a trolley car line traction power service. That is 600 volts DC. I cannot imagine how the loads were arranged but the original panel had been built by hand. A cabinet was made, lined with asbestos panels, and fitted with porcelain base double pole single throw knife switches with fuses in each line, The 2 conductors laid in the groove at one end of the switches nearest the fuse holder and the tap for each fuse holder was a copper clip tightened down on a small stripped portion of the wire's insulation. The fuse holder was followed by the hinged points of the switches. That was followed by the contact jaws of the switch. The current went out to the loads through one fuse, knife switch and knob and tube wire and came back through a knob and tube wire a knife switch and fuse.

Picture this then. The fuse is the first thing after the bus contact clip. If you want to change it you do it hot at 600 volts. Then the knife of the switch which remains energized when open because of the direction the mechanism faces. So if you wanted to work on a circuit, or perhaps even change a bulb, you could open the knife switch keeping in mind that the knives are always energized on one leg but not the other with no sign of which was which. If you wanted to change the fuse it did no good to open the knife switch unless you wanted more challenges in your life. Opening the double pole knife switch just brought the energized blade close to your fingers when your trying to turn the fuse out of its Edison base screw shell socket.

I'm supposed to figure out why the lobby lights aren't working. Since the service is now 208/120 the AC neutrals are fused. Someone has printed a helpful note on the inside of the panel cover saying caution fused neutrals. I went and found the disconnect for the 3o ampere feeder wires to the field built panel It's a screw shell fused switch from about the 1940s and actually fully enclosed. It is marked use only slow blow fuse in white wire. I open the switch remove the fuses, lock it closed and off, and go back up the stairs to the panel located above the pie shaped top step. Were was I supposed to clip my safety harness to again?

One of the fuses was blown because the neutral it was installed in was shared with a circuit that supplied the lobby lights. I don a headlamp and open up the box by the front door. I cap of the K&T neutral that has been brought into the box.
I tone out and deenergize the energized conductor from the BX cable.
Replacing the powdered insulation on the GE BX non grounding AC cable's white wire with insulation stripped from a piece of modern wire I carefully put the lighting circuit back together and tie the lighting neutral back into the neutral in the box that had the insulation crumbled off of it. I reenergize the lobby lighting circuit and the muttering amongst the occupants ceases and now I'm a hero.

Back to the ancient panel above the stairs. I shunted out all of the fuse holders and switch poles that are neutrals gauged the wires, Install type S fuse adapters for the wire size of each energized conductor and put in the Type S fuses. Then I close the cabinet that was built by some long dead electrician and say a prayer for him. GO HOME GO HOME.

We didn't have camera phones back then and I no longer have the Polaroid photograph is somewhere in the attic with the rest of my business records.

Tom Horne

 

[electrofelon]

.

As you read think about this.
If they used the terminal bar and two pole ocpd's and had wire type EGC. How would handle the bonding. Isolation to prevent unwanted currents on the EGC. What if it was as op stated change from corner grounded to ungrounded with no isolation. You would have ground fault every where untill isolated. Imagine in multiple locations where they did not follow phasing. You would would line to line faults every where if not isolated terminal bars.

Think goodness this is just a what if.

I don't follow. The "neutral bar" that we are using for an ungrounded conductor would be isolated from the egcs.

 

[electrofelon]

Leaving the corner grounded delta aside for a bit if a single overcurrent device in a grounded neutral system the Grounded Conductor of the circuit becomes energized by the voltage through the loads from the energized conductors. We expect that conductor to be no threat because it is grounded but now we have a possibly deadly set of white, or in other parts of the world black or blue, conductors that are now a hazard to our very lives.

I don't follow. This conductor is ungrounded so we treat it just like any other ungrounded conductor. It would not be white.

 

[LarryFine]

About 3 decades ago I went to do a service call on a home that had once been served by a trolley car line traction power service. That is 600 volts DC. I cannot imagine how the loads were arranged but the original panel had been built by hand. A cabinet was made, lined with asbestos panels, and fitted with porcelain base double pole single throw knife switches with fuses in each line, The 2 conductors laid in the groove at one end of the switches nearest the fuse holder and the tap for each fuse holder was a copper clip tightened down on a small stripped portion of the wire's insulation. The fuse holder was followed by the hinged points of the switches. That was followed by the contact jaws of the switch. The current went out to the loads through one fuse, knife switch and knob and tube wire and came back through a knob and tube wire a knife switch and fuse.

 

[Tulsa Electrician]

I don't follow. The "neutral bar" that we are using for an ungrounded conductor would be isolated from the egcs.

If there was multiple service disconnects. In short the whole system would need reworked to insure isolation and code for identification. I have personally seen a disconnect on the load side of the service disconnect with the bonding screw in it.

Hard to believe I found this on a reasonable new installation here in Tulsa.



Tulsa Electrician said:
As you read think about this.
If they used the terminal bar and two pole ocpd's and had wire type EGC. How would handle the bonding. Isolation to prevent unwanted currents on the EGC. What if it was as op stated change from corner grounded to ungrounded with no isolation. You would have ground fault every where untill isolated. Imagine in multiple locations where they did not follow phasing. You would would line to line faults every where if not isolated terminal bars.

 

[wwhitney]

What if it was as op stated change from corner grounded to ungrounded with no isolation.

Every system needs to keep the EGC isolated from any grounded circuit conductors. There's zero connections in an ungrounded system (as there is no grounded circuit conductor) and one connection in a grounded system, the MBJ (for a service) or SBJ (for an SDS).

Cheers, Wayne

P.S. Sorry I didn't read your long post. But if you can draw (or concisely describe) a fault scenario on a system with N circuit conductors, N-1 of which have OCPD and 1 of which does not, and for which the fault is worse if the circuit conductor without OCPD is ungrounded rather than grounded, then that would address my original question.

Seems like the simplest case is N=2, so say a 240V 2-wire single phase system with a black wire and a red wire, and the black wire has OCPD, the red wire is always wired straight through.