Measurement of transmission line parameters in time domain 1

[Bronzeado]

Hi folks,

We are planning to start a R&D Project to measure the parameters of transmission lines, using time domain (not frequence domain) approach.
We intend to use instuments of good accuracy synchronized by GPS such those used to synchrophasors.
We expect that the results of this Project could validate the TL parameters we normally use in our system studies (protection, load flow, etc.).

I wonder if somebody could help us in finding specific literature on this matter, or if know people that has worked in this field.

Thank you in advance!

Best regards,

Herivelto Bronzeado

 

[davidbeach]

Synchrophasors are a great tool to nail down the positive sequence impedance of a line. Without a fault or a severe imbalance you can't come up with the zero sequence impedance. It is pretty easy to come up with a good (not quite as good as with synchrophasors) positive sequence impedance using conductor spacing. It is always the problem of getting a good zero-sequence impedance that is the problem. If you solve that, I'd certainly be interested in your methods.

 

[Bronzeado]

Hi Davidbeach,

Thank you for your reply.
We am not going to use synchrophasors (frequence domain)but the instruments used in that system due to their accuracy.

We intend to work with the waveforms in both normal and fault conditions. So we may get the zero sequence parameters (fault condition).

We will monitoring the both ends of the TL and will work with the voltage and current waveforms (time domain) to calculate the TL parameters.

We intend also to see how the TL parameters changes with the load and weather conditions. It seems to be a very interesting subject.

Best regards,

Herivelto Bronzeado

 

[DTR2011]

OMICRON has a device called the CPC-100 and when used in conjunction with a coupling amplifier (CU-1) can measure the positive (Z1) and zero (Z0) sequence impedance of a line in question. The lines must be de-energized for the test.

http://www.omicron.at/en/products/pro/primary-testing-diagnosis/cpc-100-series/cp-cu1/

It may be worthwhile to investigate.

 

[Bronzeado]

Hi smallgreek

Thank you for the information. I will investigate to see if this device can help me.

Remember that we are looking for measuring the TL parameters in the time domain without de-energize the line.

Best regards,

Herivelto Bronzeado

 

[davidbeach]

One problem with the OMICRON system, or even using synchrophasor data during faults, for calculation of zero-sequence impedance is that what you will get is the effective zero-sequence impedance. What you want get is the self zero-sequence and a separate value for any mutual zero-sequence impedance.

 

[marks1080]

Very interesting, but I find it hard to see why you would want to use time domain vs freq. domain.

Are you looking at using properties of wavelets?

 

[Bronzeado]

Davidbeach, smallgreek and marks1080,

What we really intend to do is to install two instruments with very good accuracy, one in each end of a TL, synchronizing them via GPS to get the same time window at every snapshot.
Then, from these "windows" we will try to get the "transfer function" of the transmission line, with the instruments remain on line and getting snapshot at:
a) every 10 minutes and,
b) during fault conditions.

In (a) we will try to get information on the TL parameters of positive and negative sequence and, also, how they vary with system conditions. Initially, the measurements will be done on a single TL (no others TL in parallel. Then, we will move to a TL in a corridor with more than 2 TL in parallel to verify the effect of mutual coupling.

In (b) we will try to get information on TL parameters of zero sequence and, also, see how it varies with system and weather conditions.

The time domain approach is used because we want to consider mutual couplings and voltage and current distortions.

I would apreciate your thoughts and literature on that. Thank you.

Herivelto Bronzeado

 

[marks1080]

The new protective relays being used in power systems might be able to do some of that grunt work for you. They have very good event records/data loggers and can also be time synced together (over very long distances) typically using an IRIG-b gps clock. This wouldn't help with your transfer function, but it is a very practicle way of getting the hard data. I am still wondering why you wouldn't consider a fourier transform. Maybe its just because it was drilled into my head back when I was in school but I don't think there is anything you can do in the time domain that you can't do in freq. Please let me know if there is something I am missing.

Mark