ECN Forum Forums General Topics of the Trade General Discussion Area Switchgear problems

I have never seen this type of warning when applying CT's, but I have only been doing it for 35 years. CTs are regularly installed near 90° turns in bus bar and conductors.


CTs this oversized are very common for zero sequence ground fault measurements. If you look carefully you can see the back of the CT appears to extend past the rear phase bus.



For sure.

It is hard to see, but I could be convinced this is corner grounded conductor system. It is hard to follow the bussing to the right of the CT.

CTs work on the physics of induction -- of unsteady electro-magnetic currents -- in this case alternating current.

In every CT cabinet I've had the pleasure to know, the conductors are engineered so that they pass straight through the cores of the CTs -- with no sharp bends in sight.

When a CT is cutting the EM fields of two (2) conductor elements -- at right angles to each other it's suffering a phase shift directly from said geometry.

In the case at hand, the CT is deep into the field of the neutral -- visibly touching it. It's also shockingly close to its ninety-degree face.

These two (2) induced currents -- on the same wire -- HAVE to be out of phase due to the geometry.

Inducing currents that must be out of phase should toast like a feeble bolted-short: battling waves, if you will.

This phase shift can be eliminated by shielding the CT with a trivially cheap sheet of conductor: aluminum would be perfect.

In a corner grounded scheme, the current passing through the neutral bus would not be trivial. Even though it's earthed, it's still pumping as much energy as the other legs in what I assume is a three-phase system.

EUSERC hates such tricky schemes, so my experiences do not reflect those of the eastern states.

{It's EUSERC's position that clever schemes generate danger to firemen and Poco talent when the stress is on. It's better for all concerned if the Service options are highly constrained so that when an emergency strikes everyone is looking at the same familiar scheme.

California led the way on this front because of earthquakes and the rapid expansion during WWII. Bringing talent in from all over the country -- in a grand style and in a total rush -- caused wartime installations to become confusingly 'tricky.' The Poco's started suffering fatalities as their new hires didn't recognize oddities -- starting with corner-grounded schemes.}

I know how CTs work.I also know that in all my years of design and application, your concern of a 'right angle' has never been part of any 'mounting instructions' that I have seen.
http://www.gedigitalenergy.com/products/specs/it_basic.pdf

Do you have a reference, or is this something you have made up?

It is quite apparent that you are not familiar with a corner-grounded system. There is no neutral, however because one phase is intentionally grounded, the NEC requires it it to be white.

JBD:
I have not heard of any issues with CTs and right angles.

The split core CTs that were supplied with Emon and other check meters (submeters) had no references to right angles, only orientation of the coil.

I still feel that there was no melting, just a sloppy attempt at insulation.

The raw physics of current flow are not brought up to field electricians.

It is the domain of Physics professors, though.

I can't recommend Feynman's Lectures on Physics too highly.

He crams the better part of a semester into one lecture: 22 -- pages 22-1 through 22-18.

The right-angle impedence is a result of Maxwell's Equations. Lecture 20 -- pages 20-1 through 20-15.

The simplest way I can put it:

Imagine it's a DC flow of current... that's going in one straight direction. Around that flow a saturated magnetic field has been established.

Next, the DC current takes a right angle turn. The magnetic field HAS to turn to stay oriented around the DC flow. It's now in a different direction.

DC will only affect the CT for the very few first seconds -- as the field is being built up.

(DC does actually have impedance. All of the world's computers utterly depend on it to function. It's just that it tails off to pure resistance at all ordinary reaction times. And it's simpler to explain to the newbies that way.)

Now imagine that the DC is AC. A field that was vectored in direction alpha now is turned into direction beta. The CT measures CHANGES IN FIELD INTENSITY. It's as if you have just pushed a field compass around the circuit. You bet the needle moves. (For our example, think of AC as instantaneous DC.)

The current in the CT doesn't care WHAT the field intensity did -- up, down or flat, if it CHANGED then a wave form was induced into the CT conductor and core.

The twisting of the E & B fields obeys the right hand rule and a bunch of thick calculus equations. For field wiring, these issues are entirely neglected.

They only come up when you're pushing your luck: massive current flows, very close distances (inverse square law applies) and voltage amplification. (it's a CT)

CTs don't push a lot of current, but they do have the potential to really pump up the voltage. If you're unlucky enough to have a CT crossing two -- out of phase AC conductors -- then its own (higher voltage) secondary is going to be seeing currents it's not designed to carry.

QED.

I leave it to the readership to troll the Internet on this issue.

Suffice it to say that all surge protectors come with instructions -- that tell the installer to eliminate every possible bend in the conductors -- PARTICULARLY sharp right angle bends. This is because EACH and EVERY such turn is a locus of impedance. (per the laws of electromagnetism) If carried too far, during a lightning event, you can break down the equipotential voltage plane and cause shocks to all and every. This happens at the speed of light.

The sole reason for this specific instruction is because my assertion is correct. You don't need to coil a conductor to get an inductor effect. Even a quarter turn has an impedence effect. Yes, it's been proven in test labs -- going back forever.

Lastly, E-mon may have incorporated EM shielding in their gadget. It would cost them peanuts. All that's required is a thin metallic outer layer that's encapsulated in plastic along with the rest of the gadget. The only field that would influence their CT would be the one at the core.

The CT in question, plainly, does NOT have any shielding.

Wow Tesla. That was some pretty nerdy stuff but I understood most of it. Thanx for the explanation.

Then I am glad that I have an engineering degree which included a transformer design course.

To my knowledge shielding is not a normal feature in any standard 600V CT mounted in switchboard and switchgear.
The Emon CT is not encapsulated, effectively you can see the windings through the insulation just like in the OP.

I can not find a single CT application guide that tells about errors due to CT mounting, except for those concerning the winding distribution.

Please provide a reference from a CT manufacturer explaining this application consideration, otherwise it sounds like you are stretching your expertise in this area.

Here are links to CT design references.
http://www.mmgca.com/apps/MMG-ctdesign.pdf
http://library.abb.com/GLOBAL/SCOT/scot229.NSF/0cb8394a97bc4979c1256c6b004c4f2e/eb183a922d51bf17c12570ae004b5518/%24FILE/gan_pap.pdf

Your comment,
is almost incomprehensible. CTs are regularly applied where they surround out-of-phase conductors

JBD:

Did you mean the split core of the Emons is not encapsulated?? The windings that I see & remember slipped over the split care and they seemed to be encapsulated.

"The Emon CT is not encapsulated, effectively you can see the windings through the insulation just like in the OP."

The split corehad a plastic cover that snapped over it.

No argument from me, just for my own info.

I find it interesting that the OP has not made any comments, or answered any questions!!

I was thinking of a non-split core version, but it is not the first time I have remembered wrong. The instructions on their website definitely now shows 'molded' CTs.

Page 15 of this book shows one like I was thinking of.
http://www.submetering4less.com/downloads/levitonminimeterinstallationguide.pdf

JBD...

1) ALL instruction omits 2nd order and 3rd order effects, typically entirely, otherwise... until the end. This includes even advanced college courses. Launching off into such side-tracks maddens the instructor – and confuses the students.

Example: the Law of Gravity. At the start, Newton's law is presented as a two-body problem. The math is challenging even then. So much so that Newton invented calculus to solve the matter. (His notational conventions have never caught on. Leibniz's notations did. Centuries later the matter is still argued.)

The first scientist, Cavendish, to 'weigh the Earth' set up his apparatus to cancel out all other gravity influences – so that he could resolve the issue to at least the first decimals. To do so was brutal. Too early for a Nobel, he would've received one, hands down, if the prize yet existed.

Yet Newton's law is written to cover all matter at all times and to all distances. To do so with even three-bodies is considered upper division work, really brutal.

No-one has ever used classical math to solve even the four-body problem. NASA had to use computational mathematics... settling for an approximation that had enough accuracy to get the job done. In all calculations, the distant stars are entirely neglected, even though the theory says that they're influencing events across the light-years.

Gravity is also a 'central force.' For calculations it can be treated as if it is emitted from a single point – most of the time. This notion entirely breaks down when neutron stars approach black holes. Which brings us to some comprehension of the limits to 'theory projection.' When anyone is working from simplifying assumptions – idealized systems and interactions – that person has to be on the lookout for limiting conditions. Sometimes these situations can only be approached via 'thought experiments.' It was by this means that Einstein revealed that Newton's physics completely broke down at extreme velocities. Later it was proved that Newton's physics could not fit the extremely small, either.

In both cases, the scientific world was stood on its head. They'd found that everything that they assumed was universally true – wasn't!

2) In all my days I've NEVER seen an engineering course side-track off into 2nd and 3rd order effects – unless it HAD to. Generally, no-one can build anything based upon such deviations. Instead, in all of the practical arts, the attempt is to isolate upon one, dominant, effect (hopefully linear) and exclude/ dampen other factors.

Thus we have circuits designed as approximations, followed by peg-board mock-ups, which are then tuned by 'decade boxes' and 'pots', monitored by 'scopes until the results satisfy. I had a good buddy who earned his keep doing that. In his case, he was stomping out glitches in high frequency DC switching circuits. At every turn he discovered that he needed to provide additional grounding – ultimately for each signal line. You'll see that he was not alone: all of the modern pin-outs for CPUs feature endless grounding pins. The DC impedances can't be suppressed any other way. (He almost went insane trying to cure his state of the art motherboard.)

Under no circumstances should anyone think that the industry actually sits at a chalkboard and directly calculates the final production circuit.

3) No sales literature is EVER going to spout off product limitations. That's poor salesmanship. In the case of surge suppressors, the impedance issue is soft-peddled. I've never seen any cut sheet that explained the physics of E & B fields.

%%%

Going through the pdf linked:

“3.6 (5) Trim the [CT lead] wire to the approximate length to avoid coils of excessive wiring.”

Said coils are at risk of picking up signal contamination. The E&B fields never let up for a second.

Pg 16

“Failure to install CT's in the correct orientation and on the correct phase will lead to inaccurate meter readings...”

Being a signal analysis circuit that's been factory balanced, it's touchy.

Pg 20

“A: Meters are tested and approved for accuracy [trimmed at the bench] with CT's installed in the correct orientation. Installing CT's backwards [inverted] and inverting [flopping] the terminal connections has a slight affect on meter accuracy.”

This is further proof that the factory is concerned about 2nd and 3rd order effects. In this case, the only thing that could be at issue is the orientation of the leads as they entered and left the B fields. Because the entire nature of CT circuits is to amplify signals, the factory won't even stand behind even such a trivial shift in bias. This is another tip that the circuits are tuned. (Trimmed with variable resistors, 'pots')

The encapsulation evident in the CT's shows that whenever the factory can shield the CT's – they do so.

The CT's that lack shielding are plainly dimensioned for plain vanilla bus bars – rising vertically – in CT sections. By their very geometry, they suppress the contamination of the adjoining bus bars. They face the flats of the bus. This is not only cheaper to wind, but figures to provide a sweeter signal.

The influence of the adjacent bus bars is still there, it's just that the Inverse Square Law makes their impact of no importance to readings within the range of accuracy. So, they're never brought up as an issue. The very construction of the switchgear is highly engineered so as to suppress all side effects. This is but a part of getting it listed by UL.

For example, ferro-magnetic materials are ALWAYS influenced by B fields. Switchgear is constructed with just such materials! To stop the box, itself, from bleeding off too much energy, clearances are built into the design. Because of the Inverse Square Law, it does not take too much for this issue to be completely suppressed.

It rises its ugly head during massive circuit faults. Hence the use of bonding bushings along the axis of all GEC's. In such events, the inductance [choking] effect of the conduit has to be suppressed by making it a co-conductor. (The bonding bushings do that. They remove the B fields from the issue, the jolt will travel via the E field into the earth.)

In the extreme form, inductance into the eddy currents of ferro-magnetic materials is used to melt iron! Such gadgets use massive current flows tuned to a sweet frequency – and shielded, lest they do crazy things to every other object nearby.

The CT in the image looks like the installer came close to pulling off this trick.^^^

If he actually bonded the CT core to the neutral... bad things would happen. He, by so doing, had made the core part of the E field instead of the B field. Yikes. Such steels are not designed for low resistance! Pulsing juice would be enough to explain everything. This would be a '1st order effect, BTW.

I put this at the top of the list, because it doesn't look like the actual CT leads are fried, though I suspect they were kicking out very weird readings, too.