Harmonics in CT secondary?!

[Bung]

Interesting little problem cropped up during the commissioning of some instantaneous overcurrent check relays. Because the relays were not performing as expected (they were OK on the bench with a clean 50Hz injection) we were recording the current waveform as supplied to the relay by the CT. It showed a high (>9%) 20th harmonic component in the current (also >6% 23rd harmonic). So we checked a few more protection circuits, and saw similar distortions of the current. Adjourned to a different substation with a different source, and saw a similar effect. All this was on the 132kV protection in 132/33kV substations. Checked the 33kV side, same effect evident.

The measuring instrument was pretty rough 'n ready - a numeric transformer differential relay with a harmonic recording capability, but I believe the results are indicative of the real situation (maybe 10% accuracy and not <1% from a harmonic analyser).

Where is it coming from? The harmonic voltages on the primary side are going to be very low, but I don't have the facilities to measure this order of harmonic at 132kV (all our 132kV VTs are CVTs, and will give me all kinds of false harmonic values). But the supply authority hasn't burned out any generators yet, so the 13% THD measured in the current isn't coming from voltage distortion on the 7000MVA fault level 132kV source. So we must have some very low harmonic impedances (= resonances) on the secondary side of the CTs. We saw this effect on both high impedance busbar protection circuits and on distance protection circuits (numeric type), so it probably isn't the relays.

Could it be CT cabling layout? Little green men with a bad attitude? It all happened while I was on holidays (doesn't the best stuff always happen like that?) so we are going out to try again with some more sophisticated instruments, but it would be nice if I could go armed with half an idea of what I'm looking at. All ideas welcome!


Bung
Life is non-linear...

 

[busbar]

This may not apply, but local {and possibly more distant} MV shunt capacitors might be checked for variation of harmonic measurements with capacitor switching. Unbalanced harmonics could conceivably be caused by open fuses in larger series/parallel banks.

Try benching your differential relay and seeing if injected currents from a relay test set give reproducible numbers.

There are ANSI standards for CT-secondary grounding {and symmetry} but they are basically common sense.

As a sanity check against whatever more sophisticated equipment is used for studies, a simple current shunt and differential scope may help as a basic reference against all the data that more sophisticated diagnostic instruments can spit out.

 

[alehman]

Unless I'm missing something, the only common factors relating to the various locations of measurement are the primary current itself and the measuring instrument. Large harmonic current at such high frequencies seems unlikely on the 132kV system. The fact that you are seeing similar harmonics on the 33kV side also seems unlikely (I doubt that the power transformer would pass much 20th harmonic current). I'm immediately suspicious of the relay you're using for harmonic measurement. I would recommend following up on busbar's last suggestion.

 

[jbartos]

Suggestion: What is the voltage THD on the CT secondary? The CT V-I characteristic is linear in certain V or I interval only. For a very small current I, and for a very high current I, the CT nonlinear characteristics will show.

 

[Bung]

Good point, jbartos, it seems we are right down near the ankle of the CT. To test that one, I will try a more heavily loaded primary circuit.

As for the accuracy of the waveform as recorded by the relay, we'll be able to confirm that when we stick a 'scope on it tomorrow.

I'm reasonably certain the phenomenon is an artifact of the secondary side - I just want to track it down. There are some around here who are convinced the problem is humungous levels of zillionth harmonics on the 132kV system, and I'm just a little sceptical of that...


Bung
Life is non-linear...

 

[TBDaniel]

We have a problem at our 925 MW gas turbine power plant with harmonics. We get spurious trips of the gas turbine vibration systems and unexpected CO2 discharges. The harmonics were traced back to the auxiliary transformers that provided the hotel load for the plant. When lightly loaded the 13.8/4.16 kV and 4160/480 V transformer onsite emit 5th and 7th harmonics that interfere with the operation of low voltage equipment. If we load the transformers the noise is minimized.