DC-offset faults on 38KV utility lines

[nkellerMP]

Hello all,
On occasion there has been known to be DC-offsets that occur on 38KV utility lines during some fault conditions. The fact that these faults rarely exist still prompts the switchgear controls to delay opening the breaker five cycles. Apparently the DC-offset fault is cleared in most cases inside of five cycles. I am wondering if anyone has any information on how to measure these DC-offsets on AC utility lines of this nature. Thanks for any help.

 

[cranky108]

I use the event recording feature of my relays to find my DC offsets on our 34 system.

I usually don't see much offset beyond 3 cycles.

 

[nkellerMP]

Your responce indicates to me that your able to detect and perhaps measure the presence of these DC-offest faults. I'm wondering what sensors/transducers and/or metering apparatus your using to detect the presence of the actual DC-offsets. Thanks

 

[cranky108]

I am using an SEL-351 relay.

 

[nkellerMP]

Thanks again for all your help. Just to be clear and verify we are both talking about the same thing I've attached a link to a document. Page four talks about Asymmetrical current faults. This is what I am referring to as a DC-Offset fault. Figure eight is a good example of what it looks like. I need to be able to read and measure this type of fault pretty much as fast as it occurs. All that being said do you still hold true that the SEL-351 relay is able to read measure this type of fault in real time?

 

[dpc]

You'll have to define what you mean by "real-time". The relay is going to capture an event waveform when triggered by a trip or other trigger parameter. It will record raw current data that will be the actual instantaneous current that will include both ac and dc components. This will be limited to a fixed number of cycles per event - something like 15 or 30 plus 5 or cycles of pre-fault data.

Asymmetrical fault current is the norm, not the exception. The decay of the dc component is a function of the system X/R ratio looking into the fault. It is also a function of the voltage phase angle at the time of the fault, so it will be different in each phase.

 

[nkellerMP]

All very good information. Thanks again for all your time. Correct me if I'm wrong. Assuming Asymmetrical fault always exists on the line during the fault the faster you open the breaker the larger the breaker has to be because the earlier you open it the more Asymmetrical fault current will still be on the utility line. If one was able to monitor the line in "real time" and act on the AC fault as well as the DC Asymmetry wouldn't one be able to "choose" the best place during the cycle to open the breaker inside of three cycles while the DC Asymmety is still mostly present thereby allowing the use of a smaller breaker not to mention prolonging the life of all related components. I'm told from parties involved in wind power generation that these Asymmetrical faults happen at random times and sometimes even only on one phase of there utility lines. My interest is to entertain the idea of monitoring the line and act on the fault based on not only the magnitude of the AC fault but based on the existence or non-existence of the DC Asymmetry as well. In order to do this I need to be able to track both components in fairly high speed(real time) which brings me back to my original question of what apparatus does the industry use to read these two componets. However, my interest is in the sensors and metering technique only, not the entire relay control. It looks like the SEL-351 would serve as a good model if it really does collect these components. How fast it collects them is one thing, it's the technique used in their collection is what I'm interrested in. Obviously an AC current transformer is used for the AC fault but since AC CT's quickly saturate when subjected to DC components how is the DC Asymmetry measured? This is the source of my confusion.

 

[dpc]

A typical ac breaker will not interrupt a fault until the current passes through zero after initial contact parting. To answer your question - the "industry" uses micro-processor relays these days, or for large transmission substations, perhaps digital fault recorders such as:

http://www.energy.siemens.com/us/en...ibution/power-quality/digital-fault-recorder/

As for saturation of current transformers due to dc currents, this is not instantaneous. Virtually any level of dc current will eventually saturate a CT, but it takes time. This is a well-researched topic so you should be able to find plenty of information on this.

I have no idea what you are trying to accomplish, but before you invest much time on this, you might want to do a good search of previous research. This is a mature field, so you may be re-inventing a new wheel.

Good luck.