How does the positive-sequence voltage relay measure? 1

[stason]

Hello,

this is my first time posting in this great forum and I think I might have an interesting question.

I am performing some EMT simulations in order to study the effects of energizing a transformer. I have several transformers running in an islanded network and I get the sympathetic inrush phenomenon. Due to that the inrush magnetization current lasts longer and so does the voltage dip at a certain terminal.

At that terminal a positive-sequence voltage relay is installed and should issue a tripping signal 50 msec after the voltage goes below 0.84 p.u. It does go below 0.84 pu if I take the average. But due to the harmonics caused by inrush magnetization current this positive-sequence voltage is really jagged. It may jump from 0.7 pu to 1 pu several times during one cycle.

So my question is: does the relay filter the average value out of this jagged signal and will issue a trip or will it just follow the jagged signal and get resetted every time the voltage crosses 0.84 pu?

Thank you in advance.
Regards,
stason

 

[alehman]

You didn't say what relay you are using. You'll probably need to read the manual to determine how or if it averages the voltage and under what conditions it resets the trip timer.

Alan
“The engineer's first problem in any design situation is to discover what the problem really is.” Unk.

 

[stason]

the relay is ABB UKT 913. Unfortunately I don't have the manual nor could I find any manuals on the web.
Thanks!

 

[JJHorak]

In most digital relays that I am aware of protective elements pickup at only one sample below the setting; there is no inherent delay. If there is no time delay, you will have a trip with only one sample is out of range. Programmable delay settings will allow you to specify that the trip condition has to exist over a specified time duration and most such relays reset the timing instantaneously if one sample is seen outside of the trip level.

 

[jghrist]

Most digital relays use a filter, e.g. cosine, and only operate on fundamental quantities. It takes more than one sample to calculate an rms value. The rms value of a pure sine wave can be calculated with two samples, 90° apart, but you may have to contact ABB to determine what algorithm or filters are used in this particular relay.

 

[JJHorak]

I stand corrected again. One sample is not used in relays to trip. It takes a cycle's worth of data to accurately calculate RMS and fundamental frequency magnitude, though you can calculate either RMS or fundamental frequency data off of a quarter or half cycle of data if you assume that quarter or half cycle of data you use represents the rest of the cycle (no harmonics, no changing magnitude, no DC offset).

 

[stason]

Thanks you for your answers. I wrote an email to ABB and asked them about this specific relay. Let's see what happens.
Regards

 

[slavag]

Hi stason.
On this moment, please see attached
Good Luck
Slava

 

[stason]

Thanks a lot, Slava!

According to the manual the time delay can be set to 0 ms to issue a tripping signal after the measurement unit has picked up. Which means the relay does not average and will be resetted every time the measured value exceeds the reference value. Which in my case with jagged values would mean no tripping.

What could possibly inhibit the resetting of the relay is the reset time and the hysteresis (Email from ABB).

Thank you everybody once again.
Regards

 

[jghrist]

But due to the harmonics caused by inrush magnetization current this positive-sequence voltage is really jagged. It may jump from 0.7 pu to 1 pu several times during one cycle.

How can you operate an undervoltage relay based on sub-cycle values? A 1 pu perfect sinewave voltage will drop to zero twice each cycle.

 

[stason]

jghrist,

the positive-sequence voltage serves as a reference value, not the momentary values of phase voltages.
Regards