Troubleshooting ground fault tripping breaker. 2

[rockman7892]

I am trouble shooting the attached circuit that keeps tripping both of the breakers in the circuit (coordinatin issue) from what appears to be a ground fault. Both circuit breakers keep tripping intermitently. The circuit shown is a 480V L-L single phase circuit. The load will sometimes run for hours before tripping only to be rest and then trip out right away. After waiting some time the breakers can be reset and the circuit continues to operate.

I put a power meter on the circuit to see what is going on when the breakers trip. During normal operation I see current on both legs of the circuit at about 170A. However right as breaker trips I see the one leg as indicated on the sketch jump up to about 700A or so. I thought this was weird that only one leg would go high and not the other so I suspected there may be a ground fault. I set up another zero sequence CT around both of the legs as shown and find that when the breaker trips the meter sees the 700A on the one leg as shown as well as the zero sequence CT. Zero sequence CT reads 0 during normal operatioin. Seeing this I am suspecting there is a ground fault on the primary of the circuit somewhere on the one phase as indicated? Do you agree?

We went through and meggered both the cables and the primary of the transformer as shown in the circuit and everything is clear. We cannot seem to find eveidence of a short anywhere.

My thought is that it has to be a ground fault on the primary of the circuit. If there is a ground fault or any other type of fault on the secondary of the circuit then this would only appear as an overload on the primary of the transformer correct? In other words I there was a ground fault in the secondary of the transformer wouldn't this appear as an overload on the priamry with current seen in both primary legs as opposed to just 1?

Do you agree that what I am seeing must be a ground fault on the primary of the transformer circuit?

The only thing I'm not sure about is the SCR. Could something be going on with this to see what I am seeing?

 

[Skogsgurra]

There is a single-phase thyristor control in the circuit. If the two thyristors do not conduct exactly the same angle, you will have a DC component that may fool the CTs.

Have you looked at the current with a DC transducer? If there is any DC when the circuit trips, then you have to find and correct the error in the gating pulses for the thyristors.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[gattie]

Rockman
Suggest you disconnect the Transformer & energise the circuit.
It is possible that you have an intermittent earth fault between the t/f windings & laminations.
Good Luck

 

[rockman7892]

Skogsgurra

Your comment is interesting. I dont know a whole lot about thyristors so can you explain how this causes the DC component that you mention. Could this DC component lead to the tripping of circuit breakers?

The thing that I find most strange is the fact that both circuit breakers (feeder breaker and cabinet main breaker)are tripping everytime even though they are coordinated as such that they are not overlapping. Is there something going on that could possibly be causing BOTH circuit breakers to trip everytime?

Is it possible that something is causing the reactor in the circuit to saturate thus causing an increased current on the one leg?

gattie

Thanks for your comments regarding the transformer windings. We tested insulation of cables and transformer primary winding at 1000V DC for about 30sec-1min and get readings of over 1GOhm. Do you have any other suggestions on how to possibly located a source of the intermitent fault?

 

[Skogsgurra]

re: DC component in thyristor controlled AC circuits.

A transformer does not tolerate much DC. DC will saturate the core and that makes the impedance of the primary winding low, which in turn increases primary current.

DC in a CT also saturates the core. That can make the CTs show different readings even if measuring the same circuit. That would explain the different readings in left and right CT. What it doesn't explain is the 700 A measured by the zero sequence transformer.

How can bad gating pulses produce DC?
If you never turn on one of the thyristors, you will get single-pulse rectified DC in the circuit. Now, if you turn one of the thyristors on a little later than the other, you will also get DC - although to a lesser extent. The problem could be either an asymmetrical gating circuit (analogue ones are more prone to this than digital ones) or a bad gating transformer, a bad connection between gating board and thyristors or also ripple in the supply voltage to the gating board.

Check either the timing of the gating pulses or the DC component in the primary circuit.

Also, if your control makes fast changes in the output, you will get 'semi-DC' during the ramping up or down of the AC. That can also saturate the core and cause tripping. Especially if you are close to rated primary voltage.


Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[rockman7892]

skogsgurra

Thanks for the explanation. Unfortunately I dont have a CT or meter capable of measuring a DC component.

What I did today was move the meter and the two CT's to after the reactor in the circuit and located then right on the primary bushings of the transformer. I also put a third zero sequence CT around both of the legs again at the transformer primary.

Again the breakers tripped and when I looked at the meter recording it showed about 40A or so of current on each leg and 0A on the zero sequence CT. Both breakers again tripped. What would the fact that the current at the transformer primary is low and balanced indicate is happening? Would this point to reactor saturation?

Also I plotted the two breaker curves and have attached them. I circled the area where the breakers two instantaneous regions overlap. Considering both of these breakers are tripping at once at times, would this indicate that the breakers are seeing current at or above the instantaneous pickups of these breakers. In this case this would indicate current of aproximately 1000A or greater?

 

[rockman7892]

I have changed out the SCR and the controler supplying the gating pulses to the SCR's and the problem still persists.

I have now checked everything on the primary circuit, and when test insulation at 1000V for 1min we get upwards of 1Gohm on cables and transformer primary. I dont know what else could possibly be happening on the primary of the circuit.

I want to now possibly shift focus to the secondary of the transformer and the secondary components, however an issue on the secondary would not make sense for the unbalanced current and zero sequence CT readings on the primary. Do you agree? Could there possibly be something on the secondary?

Any other ideas would be greatly appreciated.

 

[waross]

Put a scope on the CT output and look for a DC offset?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[rockman7892]

Waross

Thanks for the suggestion. Should I look for this DC offset during normal operation or would it only be present during the fault?

Would the DC offset simply appear in the way that is would shift the AC current waveform upwards?

What would the presence of this DC offset indicate?

 

[waross]

An offset will shift the AC waveform up or down.
If aqn offset is present, give Gunnar a star. It will mean that he has suggested the correct reason for the issues.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Gutre133]

Hi Rockman,

Maybe the thyristors are making glitch. In this case you should have a short pulse strong enought to trip in the instantenous of the breaker.

You should see these glitch on the scope.

 

[rockman7892]

We finally found a smoking gun with this one. After our most recent breaker trips we found that we were not able to even get the load restarted and it was tripping almost instantly. We put a megger on the circuit again and sure enough found that one of the feeder cables was shorted to ground. It was the same cable that saw the current rise on the meter.

It appears then that since we checked cable with megger previous times with results over 1Gohm we must of had an intermitent fault that may have only been happeneing when something was heating up etc... and it got progressively worse until the point of total failure.

Does anyone know of a technique for finding if there is damge to a cable when an intermittent fault is suspected but the megger does not initally show anything? Can you leave the megger on for longer periods of time as opposed to maybe a minute or so to help try to find the fualt. Should you use a voltage than 1000V DC for testing a 600V rated cable?

Thanks to everyone who offered their advice in this thread. I learned something new in regards to asymmetrical thyristor firing resulting in DC offset as a result of this thread.

 

[Zogzog]

1000VDC is the minimim recommended test voltage for a 600V cable and 60 Seconds is the minimum time the test voltage should be applied.