CT Placement 3

[davidbeach]

We know that when placing a window CT around a conductor that the conductor should be centered in the window and continue straight for some distance beyond the CT.

Is there some rule of thumb or standard that addresses what "straight for some distance" means?

 

[catserveng]

When working with Avk many years ago adding current transformers to the generators, we were told to use the depth of the CT window as the required straight run in and out of the CT, in those cases the window depth was 100mm, so we made sure the csble was centered and straight for 100mm into and out of the CT. The engineer from Avk was very specific and said the recommendation came from their vendors, who I think were ABB and ITL at that time. I did a quick look to see if I had anything other than some notes on the margin of an old manual, but that is all I have. From that point I always used that rule of thumb, mostly on CT installations for generator differential.

MikeL.

 

[KuanYau]

To davidbeach (Electrical)
(OP)
1. In my 50 years of working experience on LV systems, I have NOT noticed any practical significant error irrespective of whether the conductor is placed perfectly in the centre and continue straight for some distance before or beyond the CT.
This can be easily demonstrated by moving around the jaw of the portable clamp-on ammeter and flex the conductor before and after the CT. There would be no "practical significant error" when not under high precision laboratory measure.

2. There is no such requirement(i.e must be placed centrally and a certain straight length before and after)in IEC CT, switchgear and installations standards.

3. For good workmanship and appearance but not technical reasons, it is desirablebut not mandatory to place the conductor centrally and maintain straight before and after the CT.

 

[davidbeach]

This is a generator split phase differential application. Given space constraints, the easiest thing to do would be bring the bus bars out to the CT and back right beside the CT. It seems that would push that edge of the much closer to saturation than the other side. I'm not terribly concerned under full load current, but I don't want the CTs saturating at unexpectedly low levels of fault current.

I hope you aren't conducting those clamp-on ammeter tests with significant fault current flowing in the conductor under consideration. ;-)

 

[davidbeach]

Oh, and thanks Mike, that's an easy enough to remember rule of thumb. Maybe scottf will offer his opinion.

 

[scottf]

The impact of centering of the conductor AND return conductor placement is very much a function of the primary current level (magnetic field level). The higher the current, the more important conductor placement becomes.

In my experience, for currents up to about 5000A, the placement of the primary conductor in the window doesn't have a big impact on accuracy. This goes for round window profiles and for CTs with a decent core cross-section (I know that's not specific). For rectangular windows with flat bar conductors, it's a little more tricky and sensitive to placement.

Return conductor placement is a little more difficult to define. I'm not sure I understand exactly where the rule of thumb that catserveng mentions comes from. Since core cross section has a lot to due with localized saturation effects and phase cross-talk effects, in some sense the smaller the CT height (and width) the longer the conductor needs to be "straight" after it passes through the CT.

For generator CTs with high current ratios, they should employ some method of magnetic shielding to minimum the impact of non-centers conductor and return conductor effect. Normally, the vendor will specify the phase-to-phase spacing and conductor routing limitations (I know we do) and this varies from design-to-design. These limits come from heating and accuracy concerns.

 

[waross]

Are you following this Gunnar? If I had your shop I would be considering some experiments to determine the practical effect of a current carrying conductor near the outside of a CT window.
Respectfully
Bill

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[scottf]

Here is a decent paper done on the subject. Although it's geared a little more towards the design aspects, it illustrates the issue well.

 

[davidbeach]

Thanks.

 

[MatthewDB]

I have never seen issues with out of center.

I have been burned by right angle bends in the busbar beside the CT. The application was 3000:5 CT's on a busbar and we saw about a 5% error that went away when the CT was relocated about 200mm away from the bend.