3 phase wiring question

[delfadelfa]

I have worked with single phase and 3 phase 4 wire but not that much with 3 phase 3 wire 240 volt. I was working at a warehouse with 3 phase 3 wire 240volt the other day. They called me out because the maintenance man had changed a 1000w HPS ballast in one of the outside fixtures and he could not get it to work. He said he was getting strange voltages from the 240volt time clock that was feeding the fixture.

Now I have seen a 3 phase wild leg where the voltage to ground is 0 volts but on this one I was getting - 10 volt - 185 volt - 234 volts - to ground on each leg but getting 243 volts between each leg. I checked back at the main and got the same voltage. Checked a 3 phase compressor disconnect and got the same voltage. So now I'm scratching my head. I turned off the double pole breaker, in the 3 phase panel, feeding the outside light. The voltage changes through out the building - 85 volts - 156 volts - 122 volts - on each leg to ground.

Now I'm thinking I have to stop listening to this maintenance man about this voltage thing and find the problem with the outside light HE FIXED. I take it apart and he's got the ground from the building connected to the ballast wire that is clearly marked lamp. One hot connected to the 240 volt marked feed for the ballast which it right. The common which you would hook up the other side of the 240 Volt is hooked to the lamp wire. And the other leg of the 240 volt is capped off and not used.

To make a long story even longer, I hooked everything up right on the ballast and the light works. I just don't understand why having this ballast hooked up wrong would change the voltage through the whole building.

 

[mcclary's electrical]

I have worked with single phase and 3 phase 4 wire but not that much with 3 phase 3 wire 240 volt. I was working at a warehouse with 3 phase 3 wire 240volt the other day. They called me out because the maintenance man had changed a 1000w HPS ballast in one of the outside fixtures and he could not get it to work. He said he was getting strange voltages from the 240volt time clock that was feeding the fixture.

Now I have seen a 3 phase wild leg where the voltage to ground is 0 volts but on this one I was getting - 10 volt - 185 volt - 234 volts - to ground on each leg but getting 243 volts between each leg. I checked back at the main and got the same voltage. Checked a 3 phase compressor disconnect and got the same voltage. So now I'm scratching my head. I turned off the double pole breaker, in the 3 phase panel, feeding the outside light. The voltage changes through out the building - 85 volts - 156 volts - 122 volts - on each leg to ground.

Now I'm thinking I have to stop listening to this maintenance man about this voltage thing and find the problem with the outside light HE FIXED. I take it apart and he's got the ground from the building connected to the ballast wire that is clearly marked lamp. One hot connected to the 240 volt marked feed for the ballast which it right. The common which you would hook up the other side of the 240 Volt is hooked to the lamp wire. And the other leg of the 240 volt is capped off and not used.

To make a long story even longer, I hooked everything up right on the ballast and the light works. I just don't understand why having this ballast hooked up wrong would change the voltage through the whole building.

He had power flowing through the lighting transformer, onto the grounding conductor, which was throwing off your reference point for your meter

 

[ptonsparky]

You most likely were working with a 240 volt ungrounded system. The odd voltages to ground occured when he grounded one of the legs via the transformer on the ballast. An ungrounded delta will have odd voltages to ground even without the problem you encountered.

A three phase wild leg will have a voltage of 208 to ground on one leg and 120 volt to ground on the other two.

A grounded delta will have 240v to ground on two legs and 0v to ground on the third.

 

[nakulak]

if phase to phase voltage is ok (which it seems to be), and you are getting strange voltages to ground, then you probably have no ground (you should check the grounded conductor, apparently the main bonding jumper is missing or loose), if the system has a ground detection system, then it is intentional.

 

[delfadelfa]

if phase to phase voltage is ok (which it seems to be), and you are getting strange voltages to ground, then you probably have no ground (you should check the grounded conductor, apparently the main bonding jumper is missing or loose), if the system has a ground detection system, then it is intentional.

Phase to phase was constant 243 volts. This system did not have a grounded conductor, only a grounding conductor, which is the conduit.

 

[ptonsparky]

Your system is not grounded. Meaning at the transformer, or first disconnect one of the phases is not intentionaly grounded. For a Wye system the center point is intentionaly grounded. Delta systems may or may not have any leg grounded.

Just because you have a conduit does not mean the system is grounded.

 

[delfadelfa]

Your system is not grounded. Meaning at the transformer, or first disconnect one of the phases is not intentionaly grounded. For a Wye system the center point is intentionaly grounded. Delta systems may or may not have any leg grounded.

Just because you have a conduit does not mean the system is grounded.

I see what your saying, all the metal in the building is bonded to ground but the transformed itself is not bonded to ground so when I am testing phase to ground there is no path back to the transformer. Am I thinking about this right?

 

[Smart $]

I see what your saying, all the metal in the building is bonded to ground but the transformed itself is not bonded to ground so when I am testing phase to ground there is no path back to the transformer. Am I thinking about this right?

Yes, you are thinking correct. Does the system have a ground detector? It should. Either that, or the system is grounded but lost it somehow...

 

[kwired]

Yes, you are thinking correct. Does the system have a ground detector? It should. Either that, or the system is grounded but lost it somehow...

As I was reading through posts this was on my mind. OP may have a corner grounded system and has lost the grounded phase someplace.

 

[Smart $]

As I was reading through posts this was on my mind. OP may have a corner grounded system and has lost the grounded phase someplace.

...lost the grounded phase? Not quite the same as lost the [corner] phase ground. You may have meant this, but I felt compelled to clarify. Having ~240V between all three line-to-line combinations is an indication no "phase" has been lost.

 

[kwired]

...lost the grounded phase? Not quite the same as lost the [corner] phase ground. You may have meant this, but I felt compelled to clarify. Having ~240V between all three line-to-line combinations is an indication no "phase" has been lost.

You are right, and I'm not exactly sure what I was thinking other than a lost connection on a conductor that is grounded. More likely the equipment grounding conductor - there may be no bonding jumper at service and or transformer and load is being introduced on the EGC.

His voltages are similar to what you would see on solidly grounded wye system that has an open neutral conductor and there is line to neutral loads connected.

As far as if it were non grounded system having what appears to be abnormal voltages to ground would be normal, but arent they often more abnormal than what was mentioned? I don't really have any experience with ungrounded systems but do understand the theory and many concepts of what is going on there.

 

[Smart $]

...

As far as if it were non grounded system having what appears to be abnormal voltages to ground would be normal, but arent they often more abnormal than what was mentioned? I don't really have any experience with ungrounded systems but do understand the theory and many concepts of what is going on there.

I'm uncertain what you mean by "more abnormal". In any one situation there are many parameters that need to be considered to make that determination to any high degree of certainty.

Electricity has an extreme advantage over human intellect in that it operates by its inherent characteristics AND does so at [near] the speed of light. That said, the measured voltages to ground seem (to me) to be "normal" for an ungrounded system, or a grounded system which has lost its ground, which has been grounded through a load at some distance and taking load impedance and return path resistance into consideration.

 

[delfadelfa]

Yes, you are thinking correct. Does the system have a ground detector? It should. Either that, or the system is grounded but lost it somehow...

This is a 50 year old bottling plant with one single phase and one 3 phase service. The 3 phase service feeds two over head buss systems and one panel that powers this light. I had the 3 phase main panel cover off but only saw the breakers for the buss and panel. Where would the ground detector be located and would this 50 year old main have one?

 

[acrwc10]

DANGER, Will Robinson!!!!!!!!

If this is a corner grounded system you would read phase to ground on 2 of the 3 phases and zero on the third, BUT the corner grounded conductor would need to be colored white or identified as the grounded conductor.
If you have ground detectors, they should have tripped with the HID ballast being grounded ( as this would have grounded a phase conductor )
It sounds as if this system is not corner grounded or protected by ground detectors, this is where the danger occurs, when you have a fault phase to phase the OCD should work normally (assuming that is install correctly) BUT when you have a PHASE to GROUND fault it will not clear (open the OCD) when this happens the voltage can start building up like a capacitor and reach very high levels, and not clear. This can cause a fire, damage equipment, or kill someone.
You need to get someone in there FAST that understands corner grounded and ground detectors on a system.

 

[acrwc10]

Look at code sections; 200.6, 250.26, 250.21, to start with.

 

[Smart $]

...Where would the ground detector be located and would this 50 year old main have one?

Typically on or near the MDP. However, detection and /or indicators could be at a monitoring station somewhere remote to the gear... could be a simple 3-light detector.

I don't know the NEC requirements at the time it was installed, but I surmise a detector was required even back then for ungrounded systems of that capacity (if in fact it is an ungrounded system). I recommend one being installed if there is none... even if grandfathered compliant without one.

 

[Smart $]

.. when this happens the voltage can start building up like a capacitor and reach very high levels ...

Voltage doesn't build up, nor does it exceed source voltage. Where there is only one line to ground fault, the respective line's voltage is reduced to or near zero using the GES as a reference and current is typically minimal. Only on a second line to ground fault does it potentially become dangerous. If to the same line as the first, parallel current paths are established for operating circuits, while the circuits' ocpd's may not trip. If second line to ground fault is of another line, the system functions similar to an intentionally grounded system.

 

[hurk27]

DANGER, Will Robinson!!!!!!!!

If this is a corner grounded system you would read phase to ground on 2 of the 3 phases and zero on the third, BUT the corner grounded conductor would need to be colored white or identified as the grounded conductor.
If you have ground detectors, they should have tripped with the HID ballast being grounded ( as this would have grounded a phase conductor )
It sounds as if this system is not corner grounded or protected by ground detectors, this is where the danger occurs, when you have a fault phase to phase the OCD should work normally (assuming that is install correctly) BUT when you have a PHASE to GROUND fault it will not clear (open the OCD) when this happens the voltage can start building up like a capacitor and reach very high levels, and not clear. This can cause a fire, damage equipment, or kill someone.
You need to get someone in there FAST that understands corner grounded and ground detectors on a system.

I don't think he has a corner grounded system, as has been said.

I was working on many ungrounded deltas back in the "70's" that had no ground monitors installed, and have seen many installations like this as being common back then, like it was said the difference is when a phase is unintentional grounded you now have a corner grounded system, and if a second phase goes to ground then will a OCPD open.

I'm not sure when the code started to require them, but I remember I started seeing them more often about mid "80's"

Ungrounded deltas are used to keep production running, that would or could other wise be a safety hazard if the whole production line was not shut down in an orderly fashion.

Ground monitors in most cases will not open anything, they will sound an alarm and flash a light, they will alert the maintenance department that there has been a ground fault and it will be needing to be fixed at the next safe shutdown.

As for a grounded "B" delta having to have a ground monitor? not sure where you got that, grounded "B" does not need a ground monitor, it has a breaker that will tell you when a ungrounded phase goes to ground.

The other type of service that is used is called a hi resistance grounded system, and is used for the same as above except more common on higher voltages, above 480.

But in a ungrounded delta, your voltages to ground will vary, as high as your phase to phase voltages or as low a 0 volts, or anywhere in between, but using a good loaded meter will usually pull it down to 0 volts.

 

[eprice]

DANGER, Will Robinson!!!!!!!!

If this is a corner grounded system you would read phase to ground on 2 of the 3 phases and zero on the third, BUT the corner grounded conductor would need to be colored white or identified as the grounded conductor.
If you have ground detectors, they should have tripped with the HID ballast being grounded ( as this would have grounded a phase conductor )
It sounds as if this system is not corner grounded or protected by ground detectors, this is where the danger occurs, when you have a fault phase to phase the OCD should work normally (assuming that is install correctly) BUT when you have a PHASE to GROUND fault it will not clear (open the OCD) when this happens the voltage can start building up like a capacitor and reach very high levels, and not clear. This can cause a fire, damage equipment, or kill someone.
You need to get someone in there FAST that understands corner grounded and ground detectors on a system.

The part that I have made bold is an over-reaction. Think about it. If on one hand I have a corner grounded delta system where one phase is intentionally grounded, and on the other hand I have an ungrounded delta system where one phase has become accidentally grounded (phase to ground fault) why would the two behave differently with regard to the build up of voltage?

As I understand it, one application where ungrounded systems are used is an installation with equipment where unexpected shutdown can cause a hazard. If an ungrounded system is used, a ground fault on one phase (essentially making it a corner grounded system) will not cause the tripping of an OCD, so the equipment continues to function. The ground detection device will alert management to the fact that a ground fault has occurred and the equipment can be shut down in an orderly manner without creating a hazard in doing so, and the fault corrected.

 

[Electric-Light]

One issue is that ballast reliability can be compromised with improper grounding.

A ballast maybe rated for 240v across terminals, but it may not be designed to have 240v across its casing and any of the terminals.