Power Waveform Issue

[drbond24]

All:

I've got some waveform data I've been scratching my head over. Can anybody out there make sense of this? Analyzing stuff like this is not my forte...

Is this enough information to draw any conclusions? I see that A phase had an overcurrent and it looks to me like B and C are moving toward each other before the breaker isolates. Does that indicate a phase-to-phase between B and C?

Thank you.

 

[brian john]

UMMMM is there a graphic to go along with this question?

 

[gar]

110308-2354 EST

There is a change in both I1 and I2 at the time current starts to flow in the neutral. The blip in the curves. There is excess current from I1 to In. Not so much from I2 to In.

Why is there current after the breaker opens?

.

 

[drbond24]

Sorry for the post and run yesterday. I'll be around today to take care of this more promptly if anyone has input.

UMMMM is there a graphic to go along with this question?

There is a picture inserted in the post. Here is a direct link to it: http://i254.photobucket.com/albums/hh84/bonddr24/screenshot.jpg

There is a change in both I1 and I2 at the time current starts to flow in the neutral. The blip in the curves. There is excess current from I1 to In. Not so much from I2 to In.

I noticed the blip, but didn't know what to make of it. Does excess current from I1 to In indicate a ground fault on that phase?

Why is there current after the breaker opens?

Probably just the way I have the graph scaled. I was more concerned with the stuff that happened before the breaker tripped. They all go flat just off the page to the right. There was a breaker misoperation, so the delay may be the time it took for the next breaker to operate.

My biggest questions are as follows:

1) Does the high current in phase 1 in conjunction with the neutral current mean a phase to ground fault?
2) Is there any significance to the change in position of phases 2 and 3? They are both shifted towards each other for the ~3.5 cycles before the breaker tripped.
3) Does neutral current simply indicate a phase imbalance, or does it automatically mean a ground fault?
4) Is it possible to have a phase imbalance without neutral current? I didn't think it was, but another unit that was connected to this same mess recorded the exact same changes in the phases (i.e. phase 1 increased and phases 2 and 3 shifted towards each other) but it recorded no neutral current. Is that possible, or do we have a busted CT somewhere that should have recorded neutral current on that unit as well?

I'm getting ready to break out the Power Systems Analysis textbook and see what I can remember, but its been 10 years since college and I haven't used this stuff since that time. I'm not sure how it will go.

 

[gar]

110209-0945 EST

drbond24:

Since you had no neutral current before the blip this means an abnormal load occurred starting at the blip. We can call it a line to neutral short for convenience, but it does not appear to be a very high overload. There seems to be an occurrence on both line 1 and 2, but less on line 2.

1) Does the high current in phase 1 in conjunction with the neutral current mean a phase to ground fault?

Likely.

2) Is there any significance to the change in position of phases 2 and 3? They are both shifted towards each other for the ~3.5 cycles before the breaker tripped.

I do not believe that the phase of line 3 changed. It is 1.8" peak to peak on my screen throughout the disturbance. But the phase angle of 1 and 2 both changed.

Suppose your normal delta load is resistive and balanced. Next connect a resistive load from line 1 to neutral this is also resistive. This added current load is phase shifted from the delta current of one of the three delta load resistances. But the line current from the full delta load in one line is in phase the line to neutral resistive load.

Next make the delta load inductive, then the line current phase angle will not be in phase with line current from a resistive line to neutral load. Under these conditions the composite line current phase angle will change when a resistive line to neutral load is added.

i do not know if I have clearly stated this. See if it makes sense.

3) Does neutral current simply indicate a phase imbalance, or does it automatically mean a ground fault?

If the normal load is a delta load on a Y supply, then there will never be any neutral current, balanced or unbalanced. An unbalanced Y load on a Y source will produce a neutral current.

4) Is it possible to have a phase imbalance without neutral current? I didn't think it was, but another unit that was connected to this same mess recorded the exact same changes in the phases (i.e. phase 1 increased and phases 2 and 3 shifted towards each other) but it recorded no neutral current. Is that possible, or do we have a busted CT somewhere that should have recorded neutral current on that unit as well?

I answered the first part above.

It is strange (unlikely) to have some other load with separate instrumentation show exactly the same results in time correlation with your first described event unless the loads are correlated in some way.

.

 

[gar]

110309-1023 EST

Looking back at the curves I see the I3 had a slight increase in current from before the blip to after the blip, but no blip in the I3. It would be useful to see several cycles before the blip.

.

 

[drbond24]

Here's another graph that I zoomed out a bit:

And a direct link to it: http://i254.photobucket.com/albums/hh84/bonddr24/screenshot-1.jpg

The graph is the phase current from an 800 MW Y connected generator. The y-axis is x1000 amps. An event occurred on the 765 kV transmission line near the plant which caused the data I've shown. A forest fire under the power lines filled the air with smoke which allowed the line(s) to arc because the smoke has different (i.e. worse) insulating capabilities than the air. Additionally, one of the breakers that was supposed to operate to isolate this transmission line didn't open properly, so the next breaker had to operate which killed the plant. I'm mostly interested in what happened out on the line to begin with.

The other unit I mentioned is a 1300 MW generator connected to the same transmission line. It recorded the same blip in phases 1 and 2 and the same overcurrent in phase 1, but it recorded no neutral current. This unit is isolated by different breakers, so once the other breakers finally isolated the transmission line with the problem, this unit kept right on chugging.

 

[Smart $]

110309-1023 EST

Looking back at the curves I see the I3 had a slight increase in current from before the blip to after the blip, but no blip in the I3. It would be useful to see several cycles before the blip.

.

Added time lines to graph...

I'm still waiting on load(s) info... (looks like he posted some while I was fiddle farting on my keyboard :roll

 

[jim dungar]

Are you actually measuring the neutral current (In) or is the waveform a calculated result?

 

[drbond24]

Added time lines to graph...I'm still waiting on load(s) info... (looks like he posted some while I was fiddle farting on my keyboard :roll

The load is the United States power grid.

If you require more information, just tell me. I'm not even sure what I need to provide here. The generator makes 26 kV and then each phase runs through a separate 317 MVA xfmr where it is stepped up to 765 kV for the transmission line. (In case that wasn't clear, there is not a big 3 phase transformer, there are three separate single phase transformers on this unit.) The neutral on the generator is grounded through a 75 kVA xfmr. The neutrals on each of the generator step up xfmrs are tied together and grounded.

Are you actually measuring the neutral current (In) or is the waveform a calculated result?

It is measured.

 

[LarryFine]

1) Does the high current in phase 1 in conjunction with the neutral current mean a phase to ground fault?

That's where I was going as I looked at the image. The L1 and N currents seem to coincide. I don't know enough to answer the other questions.

 

[gar]

110309-1207 EST

Looks like line 3 has no phase angle change from before the incident (blip) to after. I suspect that the current change of I3 resulted from the load's response to the other two line changes.

Looks like a substantial arc from I1 to I2 and additional arcs from I1 and I2 to neutral. Note that just after the blip that I1 and I2 are almost 180 deg out of phase. Possible if the current is flowing from I1 to I2. This statement might seem strange, but if the I1 voltage is relatively greater than I2's, then current would flow that way. This is also why the neutral current is about 180 out from I1.

The wiring and/or generator size difference may account for no neutral current on the smaller generator. Look at the circuit and see if there are impedance factors that would favor the large generator vs the smaller one on the neutral.

.

 

[Smart $]

The load is the United States power grid.

If you require more information, just tell me. I'm not even sure what I need to provide here. The generator makes 26 kV and then each phase runs through a separate 317 MVA xfmr where it is stepped up to 765 kV for the transmission line. (In case that wasn't clear, there is not a big 3 phase transformer, there are three separate single phase transformers on this unit.) The neutral on the generator is grounded through a 75 kVA xfmr. The neutrals on each of the generator step up xfmrs are tied together and grounded.

Are you saying you have a bank of 3 x 317MVA 1? xfmrs, plus a 3? 75kVA xfmr?

...and the only neutral connection to the genset is between it and the 75kVA xfmr primary?

If the "neutral" connections on the 3x bank are bonded together and grounded, is that grounding bonded in such a manner that ground-fault current will be measured as neutral current? ...or does it mostly bypass the neutral CT.

Can you provide a connection diagram, including the obvious ground-neutral paths? I'm having a little problem picturing the overall connection scheme, but I am a bit distracted right now.

 

[wirenut1980]

From the informtion provided, I would say there was a fault on I1 phase to ground on the 765 kV line. In increases as a result of the currrent unbalance created when the fault occurs. The voltage would have sagged pretty good during the fault, which explains the increase in I2 and I3 during the fault.

I2 and I3 become more in phase during the fault, and I bet that can be explained by the voltage phase angles shifting during the fault.

I'm not sure why the 1300 MW neutral current did not show up. Do you mean it is not measured, or it is there, but did not increase when the fault occurred?

What information are you specifically after?

 

[hurk27]

Not that I have any idea of what I'm seeing, but from what has been described and looking at the graph, I would say L1 flashed to L2 through the carbon in the smoke, then maybe because of inductive kickback or an insulator coated with carbon L1 flashed to the MGN for 3 cycles till extinguished by the final breaker opening?

Being that the generator was now out of balance to the Y connection could explain the rise of current in the other phases trying to balance the Y?

A voltage overlay would show this

 

[wirenut1980]

If I1 flashed to I2, wouldn't those currents be in phase and not 180 degrees apart?

 

[gar]

110310-0835 EST

wirenut1980:

No.

Consider a battery, and a two wire connection to a load.

At the battery end put a shunt in series with each wire. Now you have to assign phasing to the voltmeter connections to the shunts.

In a circuit diagram put the battery on the left with the + terminal at the top and the - terminal at the bottom. The direction of positive current flow in the circuit external to the battery is from the + terminal to the - terminal. Measure the voltage drop across the top shunt from the battery towards the load and the voltage is positive. Then measure the voltage drop across the bottom shunt from the battery towards the load. Now the voltage drop is negative, 180 deg phase shift in an AC circuit.

This is exactly the same condition with the phasing of current transformers in a normal 3 phase application. So when current flows from I1 to I2 and nowhere else the current transformer on I1 will show a positive polarity for the waveform when the I2 shows a negative polarity, and vice versa.

.

 

[drbond24]

Here is the voltage:

And a direct link: http://i254.photobucket.com/albums/hh84/bonddr24/screenshot-2.jpg

Are you saying you have a bank of 3 x 317MVA 1? xfmrs, plus a 3? 75kVA xfmr?

No, the 75 kVA is a single phase also and it just has the generator neutral. The neutral is tied to one side of the transformer, and the other side of the transformer is grounded. This is to limit fault current if there is a ground inside the generator. I don't think this has anything to do with the fault, I was just trying to cover everything. Didn't mean to confuse the matter.

If the "neutral" connections on the 3x bank are bonded together and grounded, is that grounding bonded in such a manner that ground-fault current will be measured as neutral current? ...or does it mostly bypass the neutral CT.

I'll get back to you on this one. I need to ask some questions here.

Can you provide a connection diagram, including the obvious ground-neutral paths? I'm having a little problem picturing the overall connection scheme, but I am a bit distracted right now.

I don't think I have anything that would be useful here. I've got a one-line of the system, but it doesn't show any detail on the ground-neutral paths.

I'm not sure why the 1300 MW neutral current did not show up. Do you mean it is not measured, or it is there, but did not increase when the fault occurred?

It is measured, but nothing was measured. The neutral current on the other unit indicates it did not change. Both units in questions are connected to the same two 765 kV transmission lines. The fault happened on one of these lines, and after the line was isolated the second unit continued to feed the second line with no trouble. It just seems to me that if both units are connected to the same stuff and one of them saw neutral current the other one should have too. However, I also obviously don't know anything about this stuff.

What information are you specifically after?

My interest is in what happened on the transmission line. Phase to phase fault? Single phase to ground? Double-line to ground?

 

[drbond24]

110310-0835 EST

wirenut1980:

No.

Consider a battery, and a two wire connection to a load.

At the battery end put a shunt in series with each wire. Now you have to assign phasing to the voltmeter connections to the shunts.

In a circuit diagram put the battery on the left with the + terminal at the top and the - terminal at the bottom. The direction of positive current flow in the circuit external to the battery is from the + terminal to the - terminal. Measure the voltage drop across the top shunt from the battery towards the load and the voltage is positive. Then measure the voltage drop across the bottom shunt from the battery towards the load. Now the voltage drop is negative, 180 deg phase shift in an AC circuit.

This is exactly the same condition with the phasing of current transformers in a normal 3 phase application. So when current flows from I1 to I2 and nowhere else the current transformer on I1 will show a positive polarity for the waveform when the I2 shows a negative polarity, and vice versa.

.

I think I understand that you're saying phase 1 was connected to both phase 2 and ground in my example...? With current flowing from phase 1 to phase 2.

 

[gar]

110310-1034 EST

drbond24:

Your voltage waveforms are useful.

V1 had the greatest drop, and V2 a substantial drop. V3 the least.

Note there is negligible phase shift of the voltage waveforms thru the incident.

I believe there was a lot of current from from V1 to neutral. If all the current was from V1 to V2, then I would expect both voltages to drop equally. So maybe more current flows from V1 to neutral than from V1 to V2. But I think I would expect a larger neutral current than what is shown.

Since the voltages are not shifted in phase very much I think the following plots would be useful:
Time base from 2 cycles before to 3 cycles after the incident (blip).
V1 and I1 on the same plot. Same for 2 and 3 as separate plots.

Another would be a single plot of V1, I1, I2, and In over the same -2 to +3 time range.

.