Neutral Current on a 3-wire system

Status
Not open for further replies.
P

philly

Senior Member
I just got done taking (2) weeks worth of power readings on the secondary of a 2000kVA 480V/277V wye secondary transformer. The neutral point of the transformer is solidly grounded. We do not use a neutral from the transformer and therefore only have 3-wires + gnd feeding two MCC's off off two seperat breakers.

I went back and looked at trends and saw that the average neutral current reading was about 5A. This current was measured directly with a CT on the neutral bus.

Why am I seeing neutral current if I have no L-N loads connected on the secondary of the transformer? Is this a result of capacitive coupling to ground?
 
I'm confused. Where is the neutral wire in which you are seeing this 5 amps? Are you saying there is a neutral wire from the transformer's WYE point (with the grounding electrode conductor connected to this point as well) to the MCC's neutral bar, and no connection from that neutral bar to anything downstream of the MCC? If so, then that wire is open at one end, and cannot be the carrier of current in any other way than by capacitive and inductive coupling.
 
philly said:
I went back and looked at trends and saw that the average neutral current reading was about 5A. This current was measured directly with a CT on the neutral bus.

Why am I seeing neutral current if I have no L-N loads connected on the secondary of the transformer? Is this a result of capacitive coupling to ground?
Click to expand...
Try turning loads off and on and see if this current changes.
 
I echo the question of where exactly the CT was placed.

By 'neutral bus' do you mean something in the transformer, presumably where the wye neutral, the EGCs, and any GECs are connected?

Is the CT placed so that you are only measuring current between the wye neutral and everything else, or could you be measuring current flowing from one 'ground' (EGC or GEC) conductor to another?

If you are measuring current between the transformer neutral and various grounded and bonded metal structures that surround the system conductors, then you _might_ be measuring capacitively coupled current. You also might be measuring a high impedance ground fault, or 277V loads using bonded metal as a bootleg neutral.

If you are feeding only 3 phase loads, then the capacitive coupling should be pretty well matched between phases, meaning net zero current on the neutral.

Do you have any frequency information? Is the 5A measured at 60Hz? VFDs can couple considerable capacitive current at their switching frequency.

-Jon
 
As Charlie B said, where is your neutral CT? And what size is it?

2000KVA could mean at least 2500:5 or even 3000:5 CT's, in either case a 5A neutral current would mean a measurement of .2% which is probably beyond the range of the CT's and therefore is not meaningful.

If this is actually a 50:5 CT being used to monitor residual ground current, it may be measuring capacitive coupling/leakage current.
 
jim dungar said:
2000KVA could mean at least 2500:5 or even 3000:5 CT's, in either case a 5A neutral current would mean a measurement of .2%
Click to expand...


I neglected to notice how tall the KVA number is. I was thinking inductive or capacitive coupling seemed improbable until I actually read the numbers. You're right Jim, it's a small percentage.
 
The neutral current was measured directly on the neutral bar on the secondar of the transformer. The CT was placed on the neutral bar at a point after where all the three phase windings tied into the bar. So in other words the CT was placed between where the neutral point was connected to the GEC and the 3 phase windings.

No neutral wire is carried from the transformer to the MCC. Only phase conductors and EGC's.

There are no single phase L-N loads on this system, thus there is nothing connected with a bootleg neutral.

I'm not exactly sure of the CT ratio, they are CT's that came with the portable fluke 1735 power meter I'm using. The meter has (3) settings for the CT's (which I'm assuming determines input measuring) which are Low, Medium and High. I have it set on high since the 2000KVa output current is considered on the high end of the scale.

I understand that with capacitive coupled current on a balanced system the coupled current should offset each other for each phase and thus sum to zero in the neutral. There are some L-L single phase loads on the system. Is it possible to have this magnigute of neutral current with unbalanced single phase loads, as a result of capacitive coupling? Is this harmful to the transformer?

If there is a high impedence ground fault on the system could i verify this by taking L-G readings on all phases. If there is a L-G fault on one phase wouldn't I read L-L voltages for L-G measurements on other phases?
 
If your meter is set to high and the CT's are effectively 2000A, you are outside the reasonable range of measurements. You are measuring the ground leakage current (and yes it can exceed 5A), you should try setting your meter/CT to a lower range, ideally around 50A.

To check for voltage to ground on an ungrounded system you should use a low impedance voltmeter, otherwise you will get a reading across any coupling capacitance.
 
philly said:
If there is a high impedence ground fault on the system could i verify this by taking L-G readings on all phases. If there is a L-G fault on one phase wouldn't I read L-L voltages for L-G measurements on other phases?
Click to expand...

That is how an ungrounded system would act in the case of a ground fault. But you have a grounded system, and the system ground bond holds the phase voltages in proper relation.

A high impedance fault is one that allows current to flow, but some aspect of the fault limits current flow to a level that doesn't trip OCPD. For example, if you have a heating element that breaks and short to ground somewhere near the middle. The electrical connection to ground permits current flow, but the resistance limits that current flow.

-Jon
 
jim dungar said:
To check for voltage to ground on an ungrounded system you should use a low impedance voltmeter, otherwise you will get a reading across any coupling capacitance.
Click to expand...

Why is that, wont your meter be in parallel with the coupled capacitance (btwn line and ground) and thus read the same votage across the meter as the coupled capacitance?
 
winnie said:
That is how an ungrounded system would act in the case of a ground fault. But you have a grounded system, and the system ground bond holds the phase voltages in proper relation.
Click to expand...

That makes sense. I'm assuming that your referring to the ground bond between the ground and neutral point. I guess the same would be true for a resistance grounded wye system minus any voltage drop across the grounding resistor?

winnie said:
A high impedance fault is one that allows current to flow, but some aspect of the fault limits current flow to a level that doesn't trip OCPD. For example, if you have a heating element that breaks and short to ground somewhere near the middle. The electrical connection to ground permits current flow, but the resistance limits that current flow.
-Jon
Click to expand...

How can you verify this? Can you used a zero sequence CT? I would think not since we have said there can indeed be capacitive coupling current.
 
jim dungar said:
If your meter is set to high and the CT's are effectively 2000A, you are outside the reasonable range of measurements. You are measuring the ground leakage current (and yes it can exceed 5A), you should try setting your meter/CT to a lower range, ideally around 50A.
Click to expand...

Did you just say he should ignore it? :grin:

jim dungar said:
To check for voltage to ground on an ungrounded system you should use a low impedance voltmeter, otherwise you will get a reading across any coupling capacitance.
Click to expand...

I agree with this to some degree, it might only be a couple microamps, but, I rather doubt it.
 
What would you read for L-G voltage readings on the two unfaulted phases of a resistance grounded system after a ground fault on the third phase?
 
philly said:
What would you read for L-G voltage readings on the two unfaulted phases of a resistance grounded system after a ground fault on the third phase?
Click to expand...
The typical high resistance grounded system is designed to alway have enough resistance in the neutral circuit so that the current is limited to about 5A. This means there is nothing like a 'true' short circuit on the first fault, therefore your L-N and L-G voltages should be near 277V (actually there will be some voltage drop).
 
jim dungar said:
The typical high resistance grounded system is designed to alway have enough resistance in the neutral circuit so that the current is limited to about 5A. This means there is nothing like a 'true' short circuit on the first fault, therefore your L-N and L-G voltages should be near 277V (actually there will be some voltage drop).
Click to expand...

So you are saying that if on a HRG system phase A goes to ground then we will read 0V L-G for A and still read 277V L-G for phases B and C. I would think that Phases B and C L-G readings would change to 480V when phase A is faulted. I'm having a hard time seeing this one.

What would happen if the fault was actually on a winding of a delta connected motor. For instance what would happen to all system L-G voltage readings if fault was on on a winding between A & B in a motor?
 
philly said:
So you are saying that if on a HRG system phase A goes to ground then we will read 0V L-G for A and still read 277V L-G for phases B and C. I would think that Phases B and C L-G readings would change to 480V when phase A is faulted. I'm having a hard time seeing this one.
Click to expand...
I may have been misleading.

If "A" goes to ground on an HRG system you will still read about 277V L-N.
If "A" is touching G then Vag is dependent on the resistance of the fault, during a bolted fault Vag = 0V.

The fault circuit is phase to ground to a 55ohm resistor (typical value) to the neutral point of the transformer. This means that the first L-G fault simply connects a heater load from L-N.
 
philly said:
So you are saying that if on a HRG system phase A goes to ground then we will read 0V L-G for A and still read 277V L-G for phases B and C. I would think that Phases B and C L-G readings would change to 480V when phase A is faulted.
Click to expand...
You are correct. The line-to-line and line-to-neutral voltages would not change, but the line-to-ground voltages would read like a corner-grounded Delta. The neutral-to-ground would be 277v across the resistor.

What would happen if the fault was actually on a winding of a delta connected motor. For instance what would happen to all system L-G voltage readings if fault was on on a winding between A & B in a motor?
Click to expand...
If the fault was in the center of a winding, it would act like a grounded center tap: The two lines flanking that fault would read 240v to ground, and the third line would act like a hyigh leg and read 416v to ground.
 
LarryFine said:
If the fault was in the center of a winding, it would act like a grounded center tap: The two lines flanking that fault would read 240v to ground, and the third line would act like a hyigh leg and read 416v to ground.
Click to expand...

From what you said, I would suspect that if the fault was in the center of a motor winding, then I could read the L-G voltages you listed above back at the MCC bus?

I had a case on a HRG system where all three nominal L-G voltages were 277V in an unfaulted condition. When a motor was turned on which had a ground fault all three L-G voltages on the bus jumped to 310V. (HRG detection system indicated a fault for this motor).

What would cause all three voltages to go high to about the same level when there was a fault? Could the fault be somewhere in the motor winding that would cause these voltages all to rise?
 
Status
Not open for further replies.
Top