Core Balance CT Question 1

[davidbeach]

Does anyone have a good reference on how placement of conductors within a Core Balance (Zero Sequence) CT can affect the accuracy?

We had a situation where a 480V motor failed. The 480V system is fed from a 7200 V system through a Dyn1 transformer. The feed to the transformer passes through a 50:5 CBCT with a 50G set at 2A (20A primary). A solid ground fault on the 480V side puts about 1500A on two of the 7200V side phases. If badly placed in the CBCT, could we get 2A of error current and the trip, or do I need to expand my search?

 

[scottf]

Can you please clarify " puts about 1500A on two of the 7200V side phases" what is the net current through the CBCT you expect during this fault?

 

[davidbeach]

Ignoring load, there was some but I don't know how much, a solid ground fault on the 480V side equates to about 1500A in two phases of the 7200V side, 180 degrees apart from each other. The two should add to zero but apparently didn't.

Fried 480V motor is known, 50G trip at 7200V is known, but not easily explainable.

 

[jghrist]

There will be no net current through the primary CBCT for a ground fault on the secondary of a Dyn1 transformer.

 

[davidbeach]

In theory, I agree completely. But I've also seen it mentioned many times that the conductors should be centered in the window to the extent possible. So, if one conductor is centered and one is against the edge of the window could there be enough error in the coupling to the CT core that there could be 2A of secondary error?

 

[dpc]

There was a similar question a few weeks ago. Maybe this will help?

http://www.gedigitalenergy.com/smartgrid/Mar07/article4.pdf

 

[LiteYear]

That similar thread was here:

http://www.eng-tips.com/viewthread.cfm?qid=364029

And that GE reference did indeed help the OP (me). What we found in my situation was that a large primary current, if positioned nearer to one side of the core than the other, could cause local saturation and thereby a small percentage error in core balance reading. You're talking about a reading of 20A due to an uncentred 1500A primary current (1.3%) in a 50:5 CT. In my case we had a reading of 0.6A due to an uncentred 200A primary current (0.3%) with a split-core 100:1 CT (it was reduced to below noise with a solid core CT).

So I'd suggest that depending on the design of your CT, that a false reading is possible. Modern solid core CTs have effective techniques for distributing flux even in the presence of uncentred conductors, but given such a high primary current compared to CT rating, and if the conductor was very poorly centred, or the CT has some non-symmetry in its winding or magnetic path, that a cancellation error of a percent or two might be possible.

We have an open project to conduct the tests described in the GE paper with our CTs. I'd love to hear the results of a similar test with yours!

 

[randyman]

From A CT manufacturer's perspective...

With low ratio window-type CTs that are used strictly in a core balance scheme (all 3 conductors passing thru window), let the buyer beware!! I encountered the DOs and DON'Ts of these CTs long before that GE article was written and I can tell you that many CT manufacturers do NOT differentiate between a standard 50:5 and one used in core balance scheme. Is there a difference? YES if you want it to operate trouble free. The manufacturers I have been associated with DO make special windings for those used in core balance schemes BUT the manufacturer must know the application. Simply stating zero-sequence is not always enough in a CT spec for this application. When I see that I ask the question.

Comment on split-core CTs - same thing applies. Realize that in most split-cores, particularly those square/rectangular in shape may only have windings on 2 or 3 sides of the core, depending on where the cut is. Again, where I come from knowing the application at hand will produce a different CT.

Same thing applies to split-phase schemes with a single window-type CT.

There are many things a CT manufacturer can do to provide optimum flux balance in a core as long as they know they have to!!!

....just saying.

 

[scottf]

I agree with Randy.

There are many switchgear and other application engineers that don't understand that CBCT/zero-sequence CTs are manufactured differently than "normal" CTs.

Happened to us recently. Customer specified the ratings and didn't give us the application. If we had not asked, they would have gotten a CT that met their ratings, but would have failed miserably for their application.

 

[dpc]

So how are the flux summation CTs different from a "normal" toroidal CT?

 

[scottf]

dpc-

There are special winding and sometimes shielding techniques used.

I would say the most common that I've seen is using segmented windings. Meaning that for the a 50:5A, for example, instead of there being 1 secondary winding with 10 turns there are multiple sets of 10-turn windings connected together. This is sometimes combined with "over-sizing" the core (larger than normally need cross-sectional area) and using larger cross-sectional area wire than would normally be required.

 

[davidbeach]

Given what I've heard about the competency of the rest of this project it wouldn't surprise me one bit that the shields weren't brought back through the CT. Arc-flash issues have kept anybody that could report back out of the compartment where the CT is located.

Thanks for the various replies that actually helped try to figure this out.

 

[MeinKaun11]

I am just curious about the CT concept operation, if the SLG fault happened during the motor running, two primary phases will cancel out as davidbeach mentioned but 3rd phase on the primary of transformer will not cancel out and CT will dish out current equal to the primary load current in that phase. Which will be enough to trip the breaker.

Did I get this right ? If yes, what is lost in the concept.

 

[stevenal]

Still no un-detected return path for the third phase. With a delta connected un-faulted transformer, the return path for the third phase is via the two other phases. All should be measured via the CBCT.

 

[DRWeig]

MeinKaun11, yes -- as long as the two unfaulted phases carry balanced current during the fault. This is not guaranteed.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[ScottyUK]

Dave, the CBCT is on the HV side of a Dy transformer feeder. The Dy transformer should isolate any zero sequence current to the LV side only. I had to re-read the original post myself.

 

[davidbeach]

Finally, the gear was opened up today. Shields neatly terminated on the breaker side of the CBCT. Now corrected with insulated ground conductors brought back through the CBCT and terminated on the field side.