Apparent leakage current from phase bus bars to ground bus bar

[Elect117]

Actually there are no line to neutral loads the switchgear so there was no need to have a grounded conductor (grey) from the XFMR center tap. There is however a green grounding conductor (basically the SSBJ) that ties to the neutral to the ground buss in the swgr


Sent from my iPhone using Tapatalk

I feel like I am only getting some of the grounding information. I can't tell if you understand what makes a service grounded or ungrounded.

1) Grounding electrode at transformer is directly connected to transformer neutral winding ?
2) Transformer case is bonded to the SSBJ from the transformer to the switchgear?
3) The green wire from the transformer to the switchgear and it is landed where at each location?
4) The switchgear has a main bonding jumper between the neutral bar and the equipment grounding bar?
5) The building's grounding electrode system is bonded in the switchgear to the ground bus which has a main bonding jumper to the neutral bar?

Not earthing the center tap of the transformer is the definition of having a ungrounded service. Not running a conductor from the transformer center tap to the switchgear's neutral or the switchgear's main equipment grounding bar will create problems. (Or the normal ungrounded stuff is needed).

 

[Dale001289]

Proper terminology is important in this discussion.

Who owns the transformer?
If it belongs to the utility the secondary is considered a service. The NEC requires the wye point of a service transformer to be brought into the service equipment.
If it belongs to the customer the secondary is considered a separatly derived system. The NEC does not require the wye point of a separatly derived system to be brought to secondary overcurrent protective device(s).

This is privately owned facility and a separately derived system
They brought the supply side bonding - SSBJ from the XFMR into the SWGR grd bus


Sent from my iPhone using Tapatalk

 

[Dale001289]

Nothing. Leave it alone. It isn't a "real" voltage that can drive a current.

Coolwill - This is what I plan to do - I agree there isn’t enough there to worry about and will megger the cable just to be sure there’s no leakage
Thanks for your input


Sent from my iPhone using Tapatalk

 

[hillbilly1]

Actually it’s connected at both ends - XFMR-SWGR


Sent from my iPhone using Tapatalk

If it is, you would be getting 277 volts to ground, even if the neutral is not used. The voltages you are getting, shows it is not. It is not common for a wye secondary to not be bonded. As others have said, if it is a delta secondary, if one leg is not bonded, then a ground fault detector must be used.

 

[Dale001289]

If it is, you would be getting 277 volts to ground, even if the neutral is not used. The voltages you are getting, shows it is not. It is not common for a wye secondary to not be bonded. As others have said, if it is a delta secondary, if one leg is not bonded, then a ground fault detector must be used.

Actually I don’t think this system would be considered ungrounded
There are no line to neural loads downstream therefore there is no 277V to ground
There’s only an EGC (or SSBJ)
Code requires that every transformer has fault path for short circuit - it also requires a GEC - it has all of this requirement


Sent from my iPhone using Tapatalk

 

[jim dungar]

This is privately owned facility and a separately derived system
They brought the supply side bonding - SSBJ from the XFMR into the SWGR grd bus

Very important information to have known.

 

[jim dungar]

There are no line to neural loads downstream therefore there is no 277V to ground
There’s only an EGC (or SSBJ)

If the transformer wye point is connected to ground, you have a 480Y/277V system and should be measuring 277V L-G regardless if the neutral conductor is actually used.

If you do not use a neutral and are concerned with this coupling capacitance, you might want to consider converting to a High Resistance Grounded system.

 

[hillbilly1]

Actually I don’t think this system would be considered ungrounded
There are no line to neural loads downstream therefore there is no 277V to ground
There’s only an EGC (or SSBJ)
Code requires that every transformer has fault path for short circuit - it also requires a GEC - it has all of this requirement


Sent from my iPhone using Tapatalk

You will still get 277 volts to GROUND if it is a wye. If you are not, a fault to ground on any phase will not trip the ocp until there is a second fault from a different phase.

 

[Isaiah]

Could be coupling.

Did you measure between the phases or just Phase to ground?

Voltage to ground on a 3 wire service can be misleading.

I would also see if you can get frequency or phase angle of the voltages between each phase. Like A to B vs B to C vs C to A.

Aside from the possible grounding issue what else would cause capacitive coupling on a 480V system ?


Enviado desde mi iPhone utilizando Tapatalk

 

[jim dungar]

Aside from the possible grounding issue what else would cause capacitive coupling on a 480V system ?

It has to due with the laws of physics. This coupling capacitance is partly due to the insulation around and between energized conductors and between them and earth. Moisture around the conductors and their spacing also impact the capacitance.

 

[__dan]

Seems to me he is describing a solidly grounded Y secondary but 3 phase 3 wire loads, no neutral.

Measurements he has would have to be open circuit both ends. With both the feeder breaker and load breakers open.

Just the feeder cables open circuit, that's what's measured floating, so it's induced Voltage not any leakage. The cable induced Voltage would be, it's not even a nuisance, it would have no effect of its own in any scenario, except for electronic noise. It could affect data signals but not power.

You could test this by shunting the open circuit cable Voltage to ground with a 1k resistor and measuring the current flow (resistor Voltage drop), which would give you power capacity from the installation (induction geometry). That should not deliver power. Signal level only.

Not what you asked about, but the cable tray will give you a lot of radiated EMF, noise, when that has regular power current flow,.

In a fault clearing event, higher impedance, more EMF. Possibly slower fault clearing time, more Voltage raise at the fault. That's the difference between the tightly bundled cable tray layout and the conductors more widely spaced. There's more magnetic field geometry, less EMF cancelling between circuit conductors, with the unbundled cable tray, especially compared to conduit, in or at industrial power levels.

 

[hillbilly1]

I have a scenario where a bus bar enclosure containing 3, copper bus bars, A,B and C and 1 copper ground bus bar. The feeder from the upstream switchgear is 480V with an 800A for the feeder so there are two parallel 1/C, x 3 - 500 kCMIL per phase attached to the 3 phases.
When I put a meter across each phase to ground I'm getting readings of leakage current (inductance?) from each phase bus bar to ground, 15V, 30V and 15V respectively. I noticed the parallel feeders are NOT bundled together in triad shape in the cable tray above - they are laying flat. Could this be the cause?

The OP is probably measuring 15V to grd bus because its not energized


Enviado desde mi iPhone utilizando Tapatalk

The original post did not say it was de-energized until

That’s correct - power is off.
The upstream CB is locked out. It’s a 3Ph, 3W system so there is no line to neutral voltage


Sent from my iPhone using Tapatalk

It is just voltage created by other live conductors in the same cable tray, sort of a transformer effect. Normal.

 

[Dale001289]

Seems to me he is describing a solidly grounded Y secondary but 3 phase 3 wire loads, no neutral.

Measurements he has would have to be open circuit both ends. With both the feeder breaker and load breakers open.

Just the feeder cables open circuit, that's what's measured floating, so it's induced Voltage not any leakage. The cable induced Voltage would be, it's not even a nuisance, it would have no effect of its own in any scenario, except for electronic noise. It could affect data signals but not power.

You could test this by shunting the open circuit cable Voltage to ground with a 1k resistor and measuring the current flow (resistor Voltage drop), which would give you power capacity from the installation (induction geometry). That should not deliver power. Signal level only.

Not what you asked about, but the cable tray will give you a lot of radiated EMF, noise, when that has regular power current flow,.

In a fault clearing event, higher impedance, more EMF. Possibly slower fault clearing time, more Voltage raise at the fault. That's the difference between the tightly bundled cable tray layout and the conductors more widely spaced. There's more magnetic field geometry, less EMF cancelling between circuit conductors, with the unbundled cable tray, especially compared to conduit, in or at industrial power levels.

That’s right. It’s 3ph/3W with no true neutral because there’s nothing downstream that requires 277V.
I’m not worried about EMI (noise) since there are no millivolt, analog or discreet circuits anywhere close to power conductors. The induced voltage I think can be rectified by bundling the parallel feeders and keeping them separate from the others in the tray. There is space in the tray to do that


Sent from my iPhone using Tapatalk

 

[__dan]

The induced voltage I think can be rectified by bundling the parallel feeders and keeping them separate from the others in the tray. There is space in the tray to do that


Sent from my iPhone using Tapatalk

No, the induced Voltage would still be there more or less. By itself, that's not something I would see as needing mitigation or elimination, especially relative to the proposed work of rebundling the parallel 500 MCM's.

Rebunding, the feeder cables, if you want to do that, the reasoning is how it performs in regular power and fault current carrying. Two things are happening. Conductor magnetic fields due to current flow are reaching out through space, the air, seeking their matching cancelling counterpart in the other cables. While they reach out they encounter, permeate, anything in between and surrounding, increasing circuit impedance (and EMF radiation). The smaller tight bundle reduces this. Also in a fault or even motor starting those cables will move, whip around. Higher circuit impedance raises the Voltage at the far end in a fault (V drop E = IZ is more than zero, Z is bigger), also reducing peak fault current (possible slower clearing time).

Not sure if it's a code requirement for tight lashed bundles for feeders in cable tray, could be. There's reason for it. But reducing induced Voltage from unrelated nearby sources is not on the list.

 

[Dale001289]

No, the induced Voltage would still be there more or less. By itself, that's not something I would see as needing mitigation or elimination, especially relative to the proposed work of rebundling the parallel 500 MCM's.

Rebunding, the feeder cables, if you want to do that, the reasoning is how it performs in regular power and fault current carrying. Two things are happening. Conductor magnetic fields due to current flow are reaching out through space, the air, seeking their matching cancelling counterpart in the other cables. While they reach out they encounter, permeate, anything in between and surrounding, increasing circuit impedance (and EMF radiation). The smaller tight bundle reduces this. Also in a fault or even motor starting those cables will move, whip around. Higher circuit impedance raises the Voltage at the far end in a fault (V drop E = IZ is more than zero, Z is bigger), also reducing peak fault current (possible slower clearing time).

Not sure if it's a code requirement for tight lashed bundles for feeders in cable tray, could be. There's reason for it. But reducing induced Voltage from unrelated nearby sources is not on the list.

Very nice explanation thank you


Sent from my iPhone using Tapatalk