Fault Analysis Help 3

[rockman7892]

We recently had a fault on one of our 4.16kV motor starters that was pretty big in magnitude. It appeas the fault happend due to a loose bus connection where the starter section's vertical bus terminates on the contactor stabs. The attached picture shows a screw being missing most causing a loose connection on this bus, and as soon as motor was started fault occured. It is assumed that the loose connection was arcing, and somehow this arcing caused a L-L fault with one or both of the other phases.

I say a L-L fault because looking at my upstream relays it shows what appears to be a phase fault of aprox 22kA. Our upstream breaker actually tripped on a ground fault but the phase overcurrent element picked up when the 22kA or so was seen. I guess somehow this L-L fault produced some ground current as well.

Our system is a 4.16kV primary distribution system which is LRG to 400A at the utility substation.

I have the event reports/waveforms from both my main breaker (multilin 750) and the utility's 387E schweitzer relay. The 387E relay is monitoring the utility transformer differential with winding 1 being the primary side, and winding 2 being the secondary side. Winding 3 is looking at the transformer tank ground connection.

The utility says that they saw the same 22kA current magnitude on the secondary but claim they saw about 2000A on this tank ground CT. I question how they can see this magnitude of ground current if we are resistance grouned limiting current to 400A? Can this current get above 400A?

I'm by no means an expert at interpreting these event reports, and wanted to see if anyone was interested in looking at these event reports to see what appears to be the sequence of events and analysis of the fault. I am interested to learn from others input. I also do not have the correct software to view utilites relay files.

I will attached multilin 750, and schweitzer 387E relay in following posts.

Thanks for the help.

 

[Eleceng01]

dpc, if I am reading the drawings correctly, the SEL-501 is actually measuring the current through the 400A grounding resistor via the Channel Y B-Phase input of the relay.

 

[dpc]

OK - that is certainly possible.

Thanks,

Dave

David Castor

http://www.cvoes.com

 

[ScottyUK]

Dave,

I totally agree with the 'trust the relays' philosophy when discussing / arguing with Operations and/or Management who would rather just re-energise the equipment and risk damaging it. I've fought that battle on many occasions, with some wins and some losses.

In this instance though if a relay says 2kA flowed through a 400A NGR then my initial thoughts would be that the relay is not reporting correctly - usually because of human error - rather than trying to find an explanation for 2kA flowing in the NGR. The data the relay reports has to be credible.

eleceng01,

I like the "In God we trust..." line.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[jghrist]

Can you post a one-line that shows what the different relay elements are measuring?

 

[rockman7892]

I have found a way to view the Multilin 750 event only. I have attached the waveform from this event. I still need to figure out how to view certain particulars on the waveforms, however the attached shows the three phase currents, ground currents, and voltages during event. Unfortunately the plots do no show magnitude or durations. I have also attached magnitudes on a seperate for the first and 3rd cursor positions.

You can see during the event that all three phases had a high magnitude current (upwards of 20kA). Because of the scaling its hard to see the ground fault current which I'm not sure is either Ig or Isg. I have moved the cursor through all points on the ground current as well as looked at them in a raw data spreadsheet that was produced, and do not see any ground current over about 29A or so. This is contradictory with what the utility relays saw. This multilin 750 did trip on an Neutral OC which I believe is set up as the ground sensing element. I need to check settings.

The strange thing in this plot is that the Vbc voltage appears to be zero. I would think that this would be an error since on the relay during normal operation it shows this voltage. PT's are in an open delta arrangement. What do you make of this voltage?

I will try to get better polts and information from this relay.

Thanks again for your input on this one guys!

 

[rockman7892]

jghrist

I had posted the utiltiy relay scheme in addition to the one-lines I had posted on 25 Jan 10 9:51. This scheme shows both the 387E relay and 501-2 relay and the CT's which they are looking at. Please let me know if you want additional information and I will find it and provide it.

Eleceng01

Thanks for posting the plots of the two relays, its alot easier to understand with a visual. I think you may be misinterperting the two gronund CT's however and which relay they corrospond to.

Looking at the arrangement, it appears that the 387E is looking at the neutal CT on the neutral bushing of the transformer as shown as element z15 and z16 on the 387E on the drawings. The 501-2 relay is looking at a CT on element z110 and z109 that is looking at a connection between the trasformer tank casing and ground. There is a copper wire bonding the transformer tank to the ground grid, and there is a CT on this bonding wire.

I also notice that in your plots you are looking at the sequence currents on the high side of the transformer. Is looking at the sequence components on the high side of a transformer a usefull tool in analyzing faults, as opposed to maybe looking at the low side?

 

[Eleceng01]

rockman7892, you are correct - in coming back and forth to this over my day, I got turned-around. Sorry.

Here are the SEL-387E LV side wiggley-squiggleys (as my old boss used to say). I have also plotted the neutral current and it does reach around 2000A after all.

 

[jghrist]

I had posted the utiltiy relay scheme in addition to the one-lines I had posted on 25 Jan 10 9:51. This scheme shows both the 387E relay and 501-2 relay and the CT's which they are looking at. Please let me know if you want additional information and I will find it and provide it.

Sorry, I missed that one.

The 3-line shows the 387E measuring the transformer neutral current, but the 1-line shows this measured by a 501.

The 1-line shows two 501's, one on the high side and one on the low side, but the 3-line shows only one.

 

[davidbeach]

I haven't had the time to do any event analysis, but seems that if there is any current (beyond noise) in the transformer tank-to-ground connection then all bets are off about how much neutral current you might see (what exactly is the circuit at the time the current is recorded?) or what it means. In the plot posted by Eleceng01, it appears that the currents are quite low, INEUMag doesn't get much above 50A, so something seems to be missing, not sure where the 2000A comes from.

 

[Eleceng01]

david, remember that the SEL-387E reports in secondary quantities so the 50A secondary is actually 2000A (CTR = 40 in this case).

 

[dpc]

OK, I'm jumping back into Scotty's camp - you really need to physically verify what these CT secondaries are connected to.

We can stare at the event reports until our eyes cross, but it's all for naught if the basics of the systems are not identified correctly.

Something is fundamentally wrong if you can see 2000 A through a 200 A resistor.





David Castor

http://www.cvoes.com

 

[rockman7892]

jghrist

The utility relaying shown on the one-line was a proposed scheme by the plant distriution system designers, however the utility changed the scheme slightly to what is shown in the 3-line drawing. The 3-line is what is currently in the field.

Thanks for posting the 387E secondary current Eleceng01. I am now puzzled as to why we would see 2000A on the neutral if this system is resistance grounded to 400A? Does the resistor heating during a fault change the resisance? I would think if anything the resistor would heat up, and thus be a greater resistance. Is it possible that somehow this resistor is shunted?

I have always been confused about the transformer tank to ground connection. Why would you want to monitor this current if all current would flow from ground to the neutral? Is this CT usually only to look at current in the event that there was an internal transformer fault between the tank and the neutral?

If the utility relay saw this 2000A fault current, then why do you think that this does not appear on the Multilin 750 relay on my main breaker? As I mentioned, scrolling through the 750 event report, I only see about 29A max ground current.

Does anyone find the fact that the Vbc voltage wavform on the 750 shows 0V? would it appear that there is something wrong with this measurement?

I know that we have a 400A LRG system to limit ground to keep ground current to apropriate levels. However if something is indeed wrong with this grounding system, then can the fact that the ground fault magnitude was high play a factor in making this L-L fault worse? Can you tell from the event repors what happened first, wheather a L-L fault turned into a ground fault or vise-versa with a ground fault turning into a L-L fault?

I am going to place a call to the utility to to come out and verify all relay connections.