Power swing speed

[lz5pl]

Dear colleagues,
How do you test power swing function of distance protections? In some of them speed of power swing (dZ/dt)is settable, other manufacturers use fixed speed, but I never have found some defined values. It is clear that swing depends of system conditions, but my test set doesn't understand this and ask me to set some value for dZ/dt
Which values you use as a border between swing and fault?

------------------------
It may be like this in theory and practice, but in real life it is completely different.
The favourite sentence of my army sergeant

 

[slavag]

Hello lz5pl.
Test of two functions is always problem: Power swing and Out-of-step/Pole slip. I recommend in both of cases ask mnf. for test procedure. I'm not sure if possible build some common test procedure.
Good Luck.
Slava.

 

[ELEP]

Power swing test function can be performed effectively using automated relay testing kits like, Omicron, FREZA, Doble or equivalent. In Omicron testing kit, there is ramp function, you can produce swing waveform as desired. Also, you can prepare power swing with power swing followed by single/three phase faults.

I think AREVA relays have fixed power swing frequency i.e. 5ms, any swing more than this frequency will be considered as swing. In Siemens & ABB distance realys have user configurable settings.

During power swing there are symmetrical changes impendances of all phases but during fault conditions impedances are are different. This is the condition which differentiates swing & fault.

 

[ELEP]

Please make correction.

power swing frequency 5HZ instead of 5ms..

 

[lz5pl]

Thank you, ELEP.
I use either my Freja 300 , either Omicron, when the customer provides test set. My question was exactly about speed of power swing. Next time I will try with 5Hz, this is good starting point.

------------------------
It may be like this in theory and practice, but in real life it is completely different.
The favourite sentence of my army sergeant

 

[slavag]

Hi Lz5pl.
Other option for test of power swing.
Just idea.
Unfortunately, I haven't option on this moment test this function.
Assume, that power swing have detection limits from 0.2Hz (lower swing) up to 8Hz(fast swing).
BTW, 5Hz in this area.

Use 3ph current and voltage relay tester ( You have it).
Calculate according to Z ( trip level of distance protection) current and voltages.
Put current on 50Hz injection and voltages on ,for example,
55Hz ( in case of 5Hz is setting of relay) and check power swing function according to event list. Use 3ph symmetrical injection.
I think it's work ( or not).
Regards.
Slava

 

[slavag]

Hi Lz5pl.
I check this option on Out-of-step (Pole slip)protection function. With small modification of setting in relay it's work.
Slava

 

[lz5pl]

Thank you, Slavaq,
I have no relay available at the moment, so thank you for testing. This test with voltage and current I used before when I had only simple numerical test set with control options close to QuickCMC of Omicron, but control only via handheld device (no reliable software !). Imagine how inconvenient it was, but it works!
BTW, when calculating dZ/dt it is important to convert from primary to secondary Ohms. Seems obvious, but when energizing date is after three days it is easy to forget it

------------------------
It may be like this in theory and practice, but in real life it is completely different.
The favourite sentence of my army sergeant

 

[PhilGavin]

There are several methods.
Caveat: Some numerical relays are not fooled by the smallest step change in values. - make increment steps as small as possible.

Powerswing is characterised by slow three phase Z locus aproaching Z origin.

Powerswing (dz/dt) say to trip on all but Z1: My first try would be to ramp from L1,L2,L3 Vnom+ to L1,L2L3 Vnom- at contant 1*Inom.
Vary increments to vary dz/dt.

The angular relationship you set between V and I will determine the approach path of the Z locus. For example 0 deg should trip on Z1R characteristic boundary and 90 deg should trip on Z1X characteric boundary.

Out-of-step: This is a bit more complicated. The setting could be N Impedance loci passes with each pass having to be within X time of the last. For example four passes within 32 secs of each other.
This can be further complicated by an internal OoS zone that trips instantaneously on the first severe OoS.

Further to this the current has to be above a certain threshold value.

The trick here is to create a Zlocus that passes through the larger '4 pass' characteristic, whilst missing the first pass instantaneous zone.

Here I set value of current above the I pickup threshold and a corresponding value of voltage that allows me to absolutely JUST enter the outer '4 pass' characteristic.
Remember you cant even clip the interior characteristic's corner.
Then I set a frequency slip between the current and the voltage.
I use the frequency differential to create a slow rotating impedance locus around the origin. For instance first case where the impedance locus would cut the characteristic again within 32 secs would result in a trip on the fourth pass. Then I produce a slightly slower rotation allowing just over 32sec between intersects and after 4plus passes I expect 'No trip'.

Note: Slip frequency must be calculated not for 360 rotation, but to create arc between leaving characteristic and re-entering it. IOW you must take into account the angular displacement caused by width of the zone.(Simple trig.)
Also take into account the Z locus starting angle.

Do this and the result is surprisingly accurate.
See attachment to illustrate what you are trying to achieve.

The internal severe OoS characteristic can easily be found as described for power swing above.