CT secondary cable shield grounding 4

[amps21]

Hi Gents,

What are the advantages and disadvantages of the shield of CT secondary cable being grounded at both the ends (Control house and breaker/transformer etc.) The above case is for a 345kV substation with shields grounded at both ends. What would be the effect of leaving an open end ? I know during LG faults, the this closed loop will have current.

Thanks

 

[davidbeach]

Ground it for safety. The circulating current on the shield will have minor impact. Having the shield float high enough to flash over could cause many nasty things to happen.

 

[scottf]

The advantage of grounding at both ends is better sheilding. As a manufacturer of CTs, this is waht we recommend, especially for metering circuits where the secondary current could be very low.

HOwever, some utilites have a practice to onlny ground at 1 end, as they are concerned that the ground sheild may see too much current during certain fault conditions.

My experience is about 50% of folks ground at both ends and about 50% of folks ground at 1 end only

Unless your company has experienced problems in the past with grounding at both ends, I would recommend you do it for metering circuits. For protection circuits, the error caused by magnetic interference should not normally be a concern.

 

[stevenal]

IEEE 525 regarding control cables:

"If only one end of the shield is grounded, large transient shield-to-ground and conductor-to-ground voltages may be
present at the ungrounded end.

A grounding conductor may be run parallel to the shielded cable to help protect the shield from being damaged when
fault currents are present"

 

[cuky2000]

Grounding the shield at both ends is a very effective
means for greatly reducing the surge voltage
level on control cables. It simultaneously eliminates
both the electric-field and magnetic field-induced
transients.

 

[scottf]

Cuky2000-

I agree with the magnetic field shielding, but I take issue with the electric field shielding. Electric field shielding comes more from twisting wire pairs than anything else on a CT circuit.

 

[waross]

If the shield is grounded at both ends, I would be more worried about externally induced shield currents. If the induction is enough to cause objectional voltages when grounded at only one end, I would investigate rerouting the shielded cable.

 

[davidbeach]

waross, CT cables have to go from point A to point B. Point A is where the CT is located, very little choice as to where that might be, and a brutal electromagnetic environment, particularly during fault events. Point B is somewhere else, a control room for example, where things are much quieter electromagnetically. The shield must be grounded at both ends, possibly elsewhere also, to maintain its properties as a shield rather than letting it float and produce very high induced voltages on the enclosed CT circuit. If you are concerned about shield currents, run a bare 4/0 copper ground conductor in parallel with the shielded cable and tie the shield to the 4/0 periodically Most of the circulating current will flow in the 4/0 and very little in the shield.

 

[waross]

davidbeach
I have read a lot of your posts and respect your knowlege.
Thank you for takeing the time to respond to my post.
You have raised another question in my mind;
With a 4/0 cable beside the shielded cable wouldn't the cable be influenced by the same field as the shield and have the same voltage induced in it?
Wouldn't the much greater current in the 4/0 be a result of the same voltage being induced in the much lower impedance of the 4/0 cable. Wouldn't virtually the same current flow in the shield as before regardless of the connections to the 4/0 cable?
I have had some hands on experience experience with induction and sheath currents. I would also expect the same common mode voltage to be induced in the ct conductors as in the shield. They are subject to the same flux that is inducing the voltage in the sheath. I would think that grounding the sheath at both ends and grounding the ct circuit at one end would result in more hazard. there would be a sheath current and little voltage at the end of the sheath. A common mode voltage would be induced in the ct circuit. (Both terminals of one ct would be at a different potential than the other ct)
After having worked on single conductor sheathed cables, and single conductor armoured cables, and medical isolation transformers with a grounded shield between concentric windings, I am firmlly convinced of the futility of trying to shield a magnetic flux with any non magnetic material.
However, I have read a lot about "Step and Touch Potentials". I would expect that during a traumatic elecrical event at a large substation or power plant, the ground potential gradients would be considerable. Is it possible that the different voltages and the resulting sheath currents may be be as a result of ground gradients? Induction currents, while probably present may be a small percentage component of the sheath current.
The comment that 50% do ground both ends and 50% don't ground both ends indicates that it may not be that important. Other comments lead me to suspect that the cause of currents in the sheaths of ct cables has been assumed rather than investigated.
The implications of complete understanding may be marginally more accurate measuring when current measurments are already acceptable. Given that a flashover from the sheath to the ct terminals may be a possibility, which would probably seriously impare the accuracy of the circuit; I think the most important thing is to avoid the possibly high voltage differences between the sheath end and the ct terminals. Those voting for induction, ground both ends. Those voting for ground gradient ground the ct circuit and the sheath at one end only.
Respectfully.

 

[amps21]

Thanks for your response guys.

I agree to the transients causing a spikey float potential / induced voltage at the open end. Even under normal operating conditions, it might have potential which is probably determined by cable shield to gnd capacitance and length.Under a fault condition it could get worse.

Rerouting the cables might work but at this stage in the job it wont be possible.

Thanks again

FYI The cables are 4C/#8 shielded.

 

[amps21]

Hi Waross,

I was wondering about the medical isolation transformers with grounded shielding between concentric windings, how are these magnetically shielding the flux? They cannot be completed in circuit cause it will be a shorted turn then.

The shielding that we have used in power transformer designs have been completely from a voltage gradient point of view and also to shield the low voltage (eg. tertiary) windings from surge transient induced voltages.
Magnetic shielding has also been used in various form using both magnetic and non magnetic materials to limit the fluc from enetring the transformer tank walls. Anyways i would like to learn more about the medical isolation transformers.

I would appreciate your reply in this context.

Thanks for your input on the CT cables shielding issue.

 

[waross]

hi amps21
As I understand the purpose of the shields, they form a grounded barrier between the primary and secondary windings.
It's been 15 years since I had to do some design for an operating room. Somewhere I have the applicable standards that I only used that once but I can't locate them this evening.
As my faulty memory serves, The safe current levels that we are used to, such as the 15 ma trip points on a gfci are only applicable to surface contact with the human body. When there is direct contact with the internal organs such as during an operation, a very much lower current level may be fatal if it contacts the internal organs. I was led to believe that it was possible for a surgeon who had a hand in contact with internal organs to terminate a patient by reaching up with the other hand and adjusting the big light with a slight current leakage. He could accidently kill his patient without feeling anything himself.
True or not, the standards were concerned that leakage current be much much lower than what we are used to normally.
Back to the transformer, I googled it earlier this evening with marginal results. I understand that the shield is to prevent leakage current from the primary to the secondary, and to prevent rf noise and transient voltage spikes from
transfering from the primary to the secondary. This is not definitive information, just what I gleaned from a few web sites. I don't know if the shield is rugged enough to prevent contact between the primary and secondary in the event of transformer burnout, but that may be a factor.
The point is, I didn't think you could shield magnetic flux with a nonmagnetic material. With copper or aluminum, if it's arranged to maximize eddy currents you may be able to dampen the field somewhat, and you can shunt it with a magnetic material, But shield it?
The transformer was quite efficient and the shield didn't seem to have any effect whatsoever on the magnetic flux.
My experience with shield grounding has been with 4-20 ma current loops that are grounded at the panel and insulated from ground thereafter. The conventional wisdom was that the shield would protect against electrostatic and rf fields, but that a shield current caused by either a ground gradient or other cause may induce a current in the circuit conductors. I wish I could be more help on the transformers, but you probably know more about them than I do.
yours

 

[davidbeach]

That 4/0 cable is tied to the station ground grid at both end, and probably other points along the way. The finer points of substation grounding get beyond me rather quickly, but the intent is that the ground and all non-current carrying metal in the substation is bonded together to get as close to one equipotential plane as possible. Yes, there will be gradients within the substation, but held to a minimum. Ground in the substation may well be at a very different potential than ground some distance away and that ground potential rise has to be considered to make it possible to safely enter the substation.

Sure, if that 4/0 was all by itself it might just cause more trouble than if it didn't exist at all, but it is part of a larger system.

 

[waross]

Agreed davidbeach
I worked on a high voltage capacitor station tears ago. The utility made available to us a lot of case histories of safety grounding incidents. The specific issue was induction voltages and currents in a grounded transmission line parallel to an energized line.
1> A workman was lifted in a basket supported by a steel cable
from a grounded crane to work on denergized grounded, transmission line. The crane was grounded by a driven ground rod. When he touched the transmission line the difference in ground potential between the crane and the cable knocked him unconcious.
2> I knew one of the men involved in this incident. The report was included with the information from the utility.
It involved work on a grounded 500KV transmission line parallel to an over 100 miles of energised 500 KV transmission line.
The line was grounded at both ends and the ground at the work site was inadvertently lost. In the resulting incident one man was killed and my aquaintance was injured.
3> The utility procedure for safety grounds was to drive one or more ground rods and to then flag off about a 50 foot square around the grouns rod to prevent contact with possible touch or step potentials.
Remember, these instances were under normal conditione and not fault conditions.
I think that there is a choice here between:
1> Grounding both ends of everything (CT CIRCUITS INCLUDED and yes, I know that this may impair accuracy) and living with the currents.

2> Grounding one end of everything at one end only and living with the voltages.

3> Grounding the sheath at both ends and grounding the ct circuit at one end and living with the currents in the sheath and the common mode voltages in the ct circuit.

Re the 4/0 again if it's grounded at both ends, the 4/0 and the sheath will be subject to the same voltage gradients. whether from induction or ground gradient.
My experience with a parallel 4/0 for grounding has been for grounding cable trays supporting feeders. The tray often does not have enough conductance to safely conduct a fault current for any distance. A ground conductor of adequate size is run in the tray and bonded to the tray at intervals.
When I was young I was taught that electricity flowed from Positive to Negative. A little later I was taught that electricity was a flow of electrons which flowwed from Negative to positive. We suffered the induced dyslexia of having to use our left hand to check a coil with the right hand rule. Then there was an attempt to teach me tha "Hole" theory; Elwctrons go from Negative to Positive but Holes go from Positive to Negative, and electricity is the movement of Holes.
Any one who has been around for very long has seen reversals in "Conventional wisdom".
Follow the codes and do what works, but we can still question. There may be room for improvement.
The changes in residential and comercial grounding methods are a good example of positive change of the conventional wisdom.

If we agree completely on everything then one of us may be redundant.
respectfully

 

[scottf]

waross-

A couple of points, as I think there may be some confusion...

1) There is absolutely no need to ground the CT circuit in more than one place. The ground on this circuit is to give the circuit a ground reference to guard against capacitively coupled voltage rises on the secondary. Grounding in more than one place has minimal effect on this and can have a major effect on on accuracy of the circuit, particularly for metering applications.

2) Grounding cable shielding is different, as the voltages impressed would be a function of current induced in the shield from magnetic fields of the overhead power lines.

3) There also can be a need to shield against the effects of the electric fields generated from the power lines. Shielding does very little in this case. The only good way for electric field shielding is the use of twisted-pair cabling. In a CT circuit the errors from electric fields come from voltages being impressed on the cable and the resulting current error caused from that. Not really a safety issue, as the impressed voltages are normally relatively low, but can be a significant issue from an accuracy perspective, particularly if the current levels are very low in the circuit. Imagine trying to maintain a 0.15% accuracy on a current level of 2.5 mA.

 

[davidbeach]

Never, ever, ever ground a CT circuit at more than one point. You may get away with it on a simple overcurrent relay, but you will never get a differential relay to work correctly if there are multiple grounds on the CT circuits. If you have multiple elements in the same CT circuit, multiple grounds may cause different elements to see different currents depending on where the elements and grounds are located. Ground once, and only once.

Shield grounded at one end only can/will have a voltage at the non-grounded end that is vastly different than surrounding "ground" at that location. Shield grounded at both ends, without other, much lower impedance, connections between "ground" at each end will carry excessive circulating current. Providing the parallel path moves most of the circulating current to the parallel conductor. Yes, with both ends of the shield grounded, the two ends will be at different potentials if there is any potential gradient between the two ends, but the difference should be much less with both ends grounded than if one end is left floating.

 

[stevenal]

Now that we're speaking of CT secondary grounding, another IEEE standard applies. See C57.13.3.

 

[cuky2000]

Shield grounding it is a controversial topic and often misunderstood phenomenon especially in high voltage environment such as substation or generating plant. The purpose of grounding the shielding of control cable in substation is different from those for power cable application since the primary concern in the first case is to minimize electrical noise that may cause tripping or other wanted effect on the control and protection system.

On the other hand, grounding the shield on power cable is more concerned with the rise voltage level for safety reasons and minimize the circulating current in the shield that reduce the cable ampacity. See the enclose information regarding power cable application

http://cuky2000.250free.com/Cable_Shielding.jpg

We should keep in mind that control cable in HV substation are subject surge produced by switching with the station capacitance of HV buses, apparatus and the grounding grid.
The wave produced in the surge are characterized by high frequency of oscillatory nature lightly damped accompanied by very strong interaction with rapidly changing electric and magnetic fields
_{[blue]NOTE: The information here were compiled from different sources as state below.[/blue]}

 

[waross]

Thank you scottf
I did not mean to suggest grounding the ct circuits at more than one point. It is an option from a mechanical connection point of view but a really bad choice from an electrical performance point of view. I accept your gentle rebuke and apologise for not being more clear in my writing. I suspect that if you consider the voltage gradient as a voltage applied to a parallel curcuit comprised of the impedance of one conductor in parallel with the combined series impedances of the ct and the various relay burdens plus the impedance of the other conductor, you could easily calculate the component of a common mode current that would pass through the relays as an error. I think it gets worse as the cables get longer. Again, I am not advocating grounding ct's at both ends but giving this as one example of why NOT to do it.
Thank you cuky2000
A very lucid explanation.
One point on the protection of the shield.
I had assumed that grounding the shield at both ends would mean connecting to the ground grid at the control room and connecting to the grid again at the ct location. I then assumed that the 4/0 would be connected to the same two points on the ground grid.
If the 4/0 is grounded at only one end, the sheath grounded to the the 4/0 only and not to the ground grid at the ct location then I agree completely with the explanation, but I'm not sure it would be more effective than the sheath grounded without the parallel 4/0. And with the 4/0 grounded to the grid at both ends will it not be subject to the same electromagnetic induction and the same ground voltage transients as the sheath. In this case we would be tying equipotential points together. If the parallel ground cable is large enough to reduce the voltage gradients caused by faults, then I would agree but I don't think that that is the case.
Thank you for your patience with me folks.