Sizing a zero sequence CT

[mbk2k3]

I have a 480V switchgear with Eaton Magnum DS breakers equipped with Digitrip 1150 units.

The system ground fault is limited to 5A via a high resistance ground on the upstream transformer. As such, the Digitrip 1150 will never see the ground fault since its lowest ground fault pick-up setting is 0.25 x In.

One of the proposed solutions to detect ground fault in the 480V switchgear is to add zero-sequence CTs on each feeder, as this would help identify which feeder the ground fault exists on, and using this zero-sequence CT with the Digitrip 1150 or an external device like an I-Gard, Bender, etc.

How do I go about sizing the zero-sequence CT such that it is sensitive enough to pick up a ground fault current of <5A? Do they come in 5A:1A configurations?

 

[pwrengrds]

Typically a high resistance grounding system wouldn't trip off, just alarm. If you want to determine the feeder it's on with high resistance grounding a pulse system can be used to vary the resistance of the neutral ground resistor. When a ground fault alarm goes off the electrician would turn on the pulse system and start opening buckets and putting a clamp on amp meter around all 3 phases to find the fault which is changing between 3-5 amps.

 

[KuanYau]

To mbk2k3 (Electrical)
(OP)

1. " One of the proposed solutions to detect ground fault in the 480V switchgear is to add zero-sequence CTs on each feeder, as this would help identify which feeder the ground fault exists on, and using this zero-sequence CT with the Digitrip 1150 or an external device like an I-Gard, Bender, etc."

I propose for (Type A.three-phase or Type B.three-phase and neutral systems) each feeder, you may install three CTs for Type A [one on each line]; or four CTs for Type B. [one on each line and one on the neutral].
These CTs primary rating shall be 1.25 x line current. The secondary may be 1A or 5A
These CTs are wired in [parallel] (wire all S1 together and all S2 together) then wire S1 to the interposing CT* P1 where S2 is wired to interposing CT* P2.
Wire interposing CT* S1 and S2 to the protection relay.
The [very small size] interposing CT* can be 5:1 or 1:1 to suit the protection relay input current rating.

2. With this method you are measuring the [spill current] similar to the zero-sequence CT; but avoid requiring a [CT with big window and very low current rating].

Hope this proposal is helpful to you

 

[mbk2k3]

hey pwrengrds,

i started looking into pulsing systems. do most clamp type ammeters do zero sequence measurements as well? i suppose what you implied is that instead of installing zero sequence CTs in the switchgear for every feeder, you can use the clamp type ammeter to determine if there is zero sequence current in that feeder?

 

[davidbeach]

No, a clamp on meter can't measure zero-sequence (unless it is around all 3 phases), but with the pulse system it would see the step changes in current if clamped on between the pulser and the ground fault.

I wouldn't go with KuanYau's suggestion. It's what works well for a solidly grounded, or even low impedance ground system but won't provide the sensitivity necessary for a high impedance grounded system. There also cannot be a neutral in a high impedance grounded system.

Personally, I'd go with 50:5 CBCTs and feed them into a relay that can work with 0.1A secondary. That 5A ground current would be 0.5A secondary, well above the minimum setting.

Alarming would be the typical response to the first ground fault; otherwise most people would avoid the costs associated with high impedance systems, but nothing says you can't trip on the ground fault.

 

[mbk2k3]

Can you elaborate on this?

"No, a clamp on meter can't measure zero-sequence (unless it is around all 3 phases), but with the pulse system it would see the step changes in current if clamped on between the pulser and the ground fault."

Everything I've read so far indicates that a pulse system would cause zero sequence current changes in the feeder that feeds the fault. And that's what you need to detect.

 

[pwrengrds]

Most loads are small enough so you can clamP around all three phases, you shouldn't need anything special just an amp meter that can see the 3-5 amp change. Typically the pulse relay is 1-2 /second.

You can find the fault without any tools. Just start flipping off loads one at a time until the ground clears. That implies that turning off the loads is not a problem. Have someone checking to see if it's cleared. That also doesn't require the pulsing system. Just have an alarm when there is a fault. The concern is that if there is a a grounded phase the other two phases are now at phase to phase voltage equals the phase to ground voltage. If a second ground fault happens it will be a phase to phase fault.

 

[davidbeach]

Zero-sequence is something that exists in a 3-phase circuit. If you have a clamp-on meter on one conductor, all you can measure is the phase current. If you have a pulse system, and you are clamped on between source and fault you will measure the pulse in a change in phase current, but that phase current will include any load. If you can clamp around all three phases, then you can measure zero-sequence current.

 

[mbk2k3]

very enlightening thank you very much for all your comments.

i will be checking with a technical applications rep at littelfuse startco to confirm implementation of their SE-330 product.

they have an excellent technical note (

http://www.littelfuse.com/~/media/f...tes/tn-gf10_pulsing-ground-fault-systems.pdf)

which says in once place "The location of the ground fault can be determined by observing the zero-sequence currents" which lead me to believe we need zero-sequence CTs to measure the pulse change.

But the same document later states "The CT on the faulted feeder will detect the sum of the charging currents of the unfaulted feeders and the resistor current. When the pulsing contactor is turned on, this CT will detect a modulating current that indicates the faulted feeder." This leads me to believe the regular phase CTs will detect the pulse change.

 

[pwrengrds]

A zero sequence CT is one that goes around all three phases. If a clamp-on CT is around all three phases it will read the zero sequence current. Typically you will need to get around all three phases on a pulsing system to find the ground because the loads generally fluctuate enough or are high enough that a pulse is masked if you around only one phase.

To detect the ground fault just put a voltage relay across the neutral grounding resistor. If your going with the pulse system a lot of companies make them that you can buy off the shelf with the ground fault alarms built in. Type in "neutral ground resistor alarm" into a search engine and 5 different companies come up that want to sell a pulsar system.

 

[MatthewDB]

Regarding 3PH 4 wire with HRG or floating: it's absolutely banned here in the US and Canada.
But not so under IEC rules. It's common to see UNgrounded 400Y230 and 690Y400 in Europe and Asia. With LN loads on the system.

I suspect that's where KuanYau is coming from.