CCVT Ferroresonance

[txtim]

We have at least two sets of CCVT's that have gone into ferroresonant oscillations after a phase to ground fault. This occurs on breaker and one half configurations where (3) CCVT's are on a line as compared to radial where there is only (1) CCVT on a line.

The oscillations are peaking at about 138kV ph-gnd (1.7p.u.) on a 138kV system (normally 80kV to ground) which suggests maybe some type of ground potential rise?

Any experience with similar issues? Any suggestions on why the supression circuits do not seem to be effective (they are built into the CCVT's from the factory). This has hapened on 2 different manufacturers CCVT's.

 

[scottf]

You say oscillations are peaking at 1.7 pu. Is this the secondary voltage output of the CCVT that is peaking or is it really the line voltage that is oscillating.

If the line voltage is stable (which is probably the case), then the ferro-resonance is localized within the CCVT, which is what the ferro-resonance suppression device (FSD) is made to dampen down.

Normally this is caused by some failure in the FSD. However, since you mentioned it's happening on more than one unit and more than one vendor, then it might be something else.

If the FSD circuit failure can be ruled out, the most common issue is associated with the use of non-linear burdens with CCVTs. The most common non-linear burdens are isolation transformers or auxiliary VTs. If using these type devices, then they should be sized at twice the voltage rating of the circuit, for example 230:230 V instead of 115:115V. This will ensure that the magnetization characteristics of the burden does not interfere with the FSD's performance.

All of the above is applicable to CCVTs using passive FSD's, which are common in the IEEE markets and most of Europe. If you are unlucky enough to have CCVTs that use active FSD's, then you've got bigger problems

If you can share what types of CCVTs you have, I might be able to offer a bit more detailed guidance.

 

[txtim]

Thanks scottf for the reply. Similar questions were asked of me by the vendor when I inquired. We do not have any non-linear loads and I believe the ferro-resonance is actually from line voltage. Another utility is on the remote end and also saw oscillation potential after the breaker(s) opened. The remote possibility is that both ends have secondary oscillations.

I wasn't sure about mentioning specific vendors in this forum but they are so limited these days that I think it's safe to say we have both Trench and Ritz CCVT's , both having passive ferroresonance supression.

The CCVT secondary potential literally comes off the secondary winding to and inside PT junction box and on to one mirroprocessor relay/control line panel. No ties, splits, junction or interfacing with any other potentials involved except through relay sync elements (no transducers, fault monitors, aux CT's or PT's...).

One vendor suggested the use of shielded secondary cable but I have only seen that on 230kV and above.

FYI...the CCVT's are 6250pF standard.

thanks for the reply.

 

[scottf]

(You may have talked to me...I'm the AREVA/Ritz US Sales Manager).

Just to be clear, if the line potential is actually going to 138 kV line to ground, then you have system ferro-resonance, which is most likely not at all related to the CCVTs. From a system perspective, CCVTs can be modeled as a capacitance from line to ground. The type of cable used on the secondary, grounding on the secondary, etc....doesn't really change how the system "sees" the CCVT.

Shielded cable would be used in relation to secondary induced voltage transients...once again...nothing to do with the transmission line voltage. This is a technical bulletin the describes this issue:

http://www.ritzusa.com/Adobe/Induced Transient Potential.PDF

Given what you've listed here, you might be looking at a far more serious and potentially destructive system issue. You may have a system grounding issue and possible ferro-resonance between some capacitance (breaker TRV capacitor or line stray capacitance) and some inductance (VTs, transformers, reactors).

 

[txtim]

Thanks again scottf. At the risk of offending you, I inquired to Ritz and Trench and only heard back from Trench. Obviously, my inquiry didn't make it through the proper channels (until I joined this forum).

You raise an interesting but valuable point on sytstem oscillations and not one that I'm sure I could model. I will surely discuss with my peers.

On at least 2 different occasions, the trace waveformes recorded by our digital relays show 130-138kv ph-gnd on the two unfaulted oscillating phases and little to nothing on the faulted phase after the breaker opened.

 

[scottf]

Hmm...did you call the factory?

Let us know if you need anything further.

 

[txtim]

scottf,

I would have (and did) go throgh my regional distributor/rep. No harm, no foul.

Thanks for the replies and I will call if I need anything else.

regards.

 

[JensenDrive]

Without a description of the system feeding the CCVTs it is hard to envision what is going on. Are disconnect switch operations occuring at the time of ferroresonance. If you open one or two phases on your wye source and are feeding a delta load, your remaining phase(s) feeds through the delta load windings to back feed the CCVT capacitance on the open phases.
I believe the ferroresonance supppession cirucuits on CCVTs is to dampen ferroresonance in the secondary circuits. I believe it is too light duty to dampen ferroresonance in your power system componenets.

 

[scottf]

Jensendrive-

Can you define what you mean by "backfeeding capacitance"?

The CCVT FSD is only for dampening internal resonance between the CCVTs capacitor divider and the inductive components in the base tank that can occur cue to high-frequency switching surges.

If there is truly a rise in voltage on the line, it's unlikely that the CCVT is involved at all...surely not a voltage rise at the other end of the line.

 

[txtim]

scottf,

Since the FSD is between the capacitance divider and the base tank, is the CCVT incapable of "backfeeding" potential to the 138kV transmission line? In other words, is it likely that the CCVT capacitance be a significant factor in the equation in the transmission line resonance?

I'm curious to find out what would happen if we replaced one of the CCVT's with a PT and monitored for future faults. If what you are suggesting is true, that the transmission line is resonating, then the PT would likely detect that as well.

This resonance I'm speaking of is occuring at the fundamental frequency. Am I to assume that the FSD only activates in high frequency situations? It also seems like to voltage we are seeing should be dampened by some sort overvoltage protection (I don't have the data sheets in front of me right now).

GP

 

[scottf]

The FSD is connected across one of the secondary windings of the intermediate voltage transformer of the CCVT.

Yes...back-feeding appreciable voltage from the CCVT secondary to the H1 primary terminal is next to impossible with a CCVT.

Normally, the CCVT is not involved in system ferro-resonance. Normally, the resonance is between some inductance (power xmfr or VT) and a capacitance (like line stray capacitance or capacitors on breakers). Normally replacing a CCVT with a VT would make matters worse, not better.

You say he resonance is occurring at the fundamental frequency. That essentially indicates that it's probably not resonance at all, but rather maybe a capacitance voltage rise on an open line?

Ferroresonance is normally not a state that is continual. It's normally considered to be an unstable condition that typically ends in the failure of some device (normally the VT or power transformer bushing). If you had ferroresonance, the voltage waveform should look extremely distorted.

With all of the above said, I've reachd my limit on knowledge of system ferroresonance. You may want to consider getting a system study consultant involved if this is really a continuing problem. From my perspective, based on what you've shared, it sounds unlikely that it's actually a problem with the CCVT.

Feel free to call Areva/Ritz in Waynesboro, GA and ask for Scott if you wish to discuss it further on the phone. I may be able to pick the brains of some of our engineers as well.

 

[JensenDrive]

By Backfeeding, I mean this circuit. Hope it comes through.

Vag----open-----Cap(ag)------|delta load|
Vbg----open-----Cap(bg)------|delta load|
Vcg--------------Cap(Cg)------|delta load|

A very similar circuit would arise if only one phase was open. The capacitance if being energized in an abnormal way, via the magnetics of the delta load. In the case above, the phase C to ground voltage feeds through the magnetics of the delta load into the A to ground and B to ground capacitance. This is a series LC circuit where L is saturable and hence variable. Part of the backfed capacitance is the CCVT line to ground capacitance. As I understand ferroresonance, this is the most common circuit associated with ferroresonance.

From the description of the circuit, I cannot say whether this circuit applies. Do you have uneven pole opening on the wye source that could momentarily be creating such a circuit?

 

[scottf]

Jensen-

I've never heard of that scenario (delta connected load backfeeding line) at 138 kV. Isn't that more of something that could occur on distribution voltages?

 

[JensenDrive]

Yes, the circuit I described is definitely found moreso at distribution voltage, mainly because of all the single phase switching devices feeding delta loads. I was just resonding to the string by writing up the circuit where I had heard ferroresonance was most common, and asked if by some odd chance the circuit had arisen. Single phase switching at 138kV would not be very likely, but it is plausible, such as via a gang switch not working uniformly on all phases.