Stray Ground Currents in Low Voltage Power System

[mlcrk]

I have encountered a situation in which a measurable ground
exists in a low voltage power system. The system is a 277/480 Vac three phase power system derived through a 2000 kva transformer fed from a 13.2kv service. The transformer neutral is solidly grounded. The main breaker has ground fault relaying. The existing measured currents are below the rating for the ground fault sensors. The current has been measured on a number of the equipment ground conductors (with clamp on CT's and Data logging equipment) that are in the conduits with the feeders. Ground currents have also been measured on the ground grid conductors on the bonding conductor to the transformer neutral. The currents vary in magnitude over time but we cannot correlate it to any equipment stopping or starting in the process. This substation is part of an industrial facility that operates 24-7 so troubleshooting is difficult. The majority of the load served is 3 phase consisting of motor loads (some with VFD's), process resistive loads and 480/120 Vac distribution transformers. There is some 277V lighting however this load was isolated and meggared during an earlier phase of the troubleshooting effort. No grounds were detected in the lighting systems. The power distribution feeders are for the most part in underground conduits routed through a manhole system. This condition has been going on for some time since first detected (months).
My experience says that I would suspect that there is obviously a high resistance ground in the system. However, my experience over the last twenty years in the engineering and maintenance business is that ground faults on solidly grounded systems generally develop quickly into damaging phase to phase or three phase faults if not cleared by the ground fault protection. Usually the damage and resultant equipment outage is easy to find.
We plan to ground check all feeders at the next available downtime.
However because I have never experienced a solidly grounded system behave like this I would also like to get some other opinions on what else could possibly explain the situation and what other things we should investigate.

 

[jghrist]

How much ground current is there on the ground bond to the transformer neutral? There will always be some resistive and capacitive leakage currents to ground. There may also be an inadvertant ground on the neutral of a 277 volt lighting circuit. You might locate the latter by measuring the neutral to ground voltage at some lights. If it is near zero, this would indicate a ground on the neutral other than at the source.

 

[dpc]

I agree with jghrist that a mis-wired light fixture is a likely cause.

It would be good to know if this current is fundamental current or third harmonic.

 

[busbar]

NEC 250-6 discusses ‘Objectionable Current Over Grounding Conductors’ raising several points; mainly, what is objectionable current? Electrical neutral-to-ground faults can desensitize feeder ground-fault protection, but is there significant risk of that happening in this case?

If you’ve taken reasonable precautions in mitigating the problem, and tracking its indications, it’s possible that you may have to ride it out, and wait and see what finally fails.

About the “non-stop 24/7” point of view: no allowance for scheduled downtime is a great way to guarantee some unscheduled downtime—very likely at a time when it will may yield delayed restoration and cost the most in manufacturing loss. Production folks occasionally forget that there are no absolutes, and stuff does fail—here again with possibly very bad timing.

 

[alehman]

A small amount of ground current is not unusual. There's even measurable current in a few conduits in my house. I'm guessing your 2000kVA USS might be feeding a fairly large system. If your ground current is relatively low (< a few amps), it may well be capacitive leakage current. If it is a significant fraction of the circuit rating, it probably should be investigated.

 

[jbartos]

Comment: The original posting does not address the age of the power distribution system. The older power distribution system may experience larger surface leakage currents due to deteriorating (lowering) surface resistances, and due to deteriorating insulations (dielectricum) of conductors and other devices. The currents leaking through dielectricum tend to be capacitive or charging current. The 13kV level can cause larger dielectric currents.

 

[PWR]

From what you described, it is doubtful to me that you actually have a ground fault. Current can flow on the grounding conductors for a number of reasons. I once measured 250A on the neutral bushing bonding conductor of a 2500KVA wye connected transformer BEFORE it was energized. In that case, a nearby high capacity open bus system was inducing current on the neutral bus because the bus geometry did not allow for flux cancellation between the phases. While not usually as dramatic, a version of that phenomenon can occur in a typical feeder circuit. Consider: there is a long conductor in your conduits that is grounded on both ends thereby completing a loop. If there is incomplete cancellation of the individual phase conducter flux linkages, the resultant flux will link the loop present and circulate current in it. You have not specified any actual current levels but this may be what you are experiencing.

 

[buzzp]

Any equipment that is drawing reactive power is also returning this power. Perhaps the return path is via your ground lead (especially in an older facility).
I have seen large potentials between points on the ground mat that is related to equipment starting (magnetizing current) and any switching power supplies (VFD's, many computers, etc.)

 

[mlcrk]

Thanks to everyone for all the thoughts and responses on this problem.
I just wanted to provide some clarification to my initial statements and answer some of the questions that were asked. The current at the ground bond at the neutral of the transformer has been measured at as high as 50 A. These measurements were taken with instruments that probably only respond to peak readings not true RMS. We have totally isolated and deenergized the 277 V lighting branch circuits so they shouldn’t be a source of a ground.

I don’t know about the harmonic content. The transformer is a Delta/Wye type, solidly grounded. I didn’t think that this type of connection would be a source of third harmonic current.

This switchgear system is from the late 1960’s / early 1970’s. The cabling is a little older but most of what I have seen appears to be either rubber or thermoplastic type insulation.
I would suspect that the sum of the capacitive and leakage currents should not be in the 50A range for this system. I thought about whether the primary leakage current could be returning through the zero sequence impedance of the transformer? Does anyone have any thoughts on this?

Loop currents were mentioned as a possible source of the problem due to incomplete flux cancellation. I assume this would be due to imbalance between phase currents in individual conduits? If so is there anyway to calculate the magnitude based on the difference in current or is it so related to geometry that no practical methods exist?

Why would the reactive power from inductive loads return through the ground circuit?

If anyone has any further thoughts or ideas please let me know.

 

[Deansharafi]

You mentioned that you have VFD drives and lighting in the circuit. These two (if lighting consist of florosent lamps)constitute major sources of third harmonics, which go back to system through the star point. The level you mentioned 50A doesn't concern me, because if you carry out an analysis you will find that this current is mainly third harmonics produced by some of the load in the system.

 

[busbar]

Expected capacitive charging current on a 2MVA low-voltage system would be less that 10 amperes.

Fifty amperes ground current could be a collection of “bits and pieces,” and does not seem to represent much of a heating source or capable of doing much damage.

 

[jbartos]

Comment on mlcrk (Electrical) Feb 18, 2004 marked ///\\ I just wanted to provide some clarification to my initial statements and answer some of the questions that were asked. The current at the ground bond at the neutral of the transformer has been measured at as high as 50 A.
///Please, clarify &quot;at the ground bond at the neutral&quot; whether is was between the neutral and ground or between the neutral and transformer winding.\\
I don’t know about the harmonic content. The transformer is a Delta/Wye type, solidly grounded.
///The harmonic content can be present from the downstream loads and/or from upstream power supply (exclude triplen harmonics).\\ I didn’t think that this type of connection would be a source of third harmonic current.
///There might be some third harmonic content.\\This switchgear system is from the late 1960’s / early 1970’s. The cabling is a little older but most of what I have seen appears to be either rubber or thermoplastic type insulation.
///Depending on the maintenance and environment, 30+ years may be felt in the ground current return.\\I would suspect that the sum of the capacitive and leakage currents should not be in the 50A range for this system.
///It depends on the power distribution and nature of loads. Capacitors, poor insulations, and rotating machinery tend to provide higher ground current returns.\\ I thought about whether the primary leakage current could be returning through the zero sequence impedance of the transformer?
///Normally, the delta primary winding of good transformer should break the zero sequence current path.\\ Does anyone have any thoughts on this?
///Usually, the primary leakage currents return to the upstream source.\\Loop currents were mentioned as a possible source of the problem due to incomplete flux cancellation. I assume this would be due to imbalance between phase currents in individual conduits?
///Normally, the imbalance load current flows in the neutral in 3phase 4wire power distribution system. The ground return currents can be due to induced currents into the grounded conductive path, e.g. long railing, usage of unshielded cables along long parallel conductive metals, etc.\\ If so is there anyway to calculate the magnitude based on the difference in current or is it so related to geometry that no practical methods exist?
///Normal pencil and paper calculations will be tedious. It may be better to use software to simulated the leakages, stray currents, capacitive currents, etc. using distributed parameter models.\\Why would the reactive power from inductive loads return through the ground circuit?
///The inductive loads tend to have many turns that represent long conductors with somewhat small insulation thickness. This could somewhat justify the return of the reactive power and current flow in the ground circuits.\\

 

[PWR]

It is not practical to calculate the current induced in the grounding conductor by what could be called transformer action from the phase conductors. For the lighting circuits, double check that the neutral bus in the panel(s)is not grounded. Third harmonic currents in particular result from discharge type lighting units and they will be found in the neutral and the ground if the latter is not kept isolated from the neutral beyond the single desired connection at the source transformer or switchgear.