Ground trip when 480VAC motor starts

[dithomas]

I did find an old thread on this topic but there was never a conclusion.

I have a problem with a 480VAC chiller with a Y-Delta starter. The motor is rated at 390FLA.

If the building is in operation with lights and conveyors running at around 350 to 400 amps and the chiller attempts to start the 2000A main breaker (this is the only device with GFI) trips. The trip unit is a Digitrip-520 with the ground setting at 800A 0.50 second I2T for a 3P 4W system with a neutral CT and X0 Ground bond on the transformer side of the Neutral CT.

If the building is not in operation on a Saturday with only lights the chiller starts with no issues. We performed three starts.

Today I took a good look at the switchboard to make sure CT polarities were correct and account for all neutral and ground connections of the out going conductors, and tested the GF function on the trip unit which was OK.

The switch board main section is a new semi custom 2000A 3P 3W section which was installed to replace an existing GE Bolted Pressure switch. The load bus has been configured to match and line to the existing GE feeder fused switches. There had to be some custom bus detail for the neutral bus.

What I discovered today is that the Neutral CT is very close, within 3/4", of the phase A bus running vertical along one side of the N CT. I am now wondering if, with a certain amount of 3P 4W building circuits in operation, that the flux produced by the motor starting inrush (630A) in the Phase A bus is coupling over into the 2000:1 neutral CT causing an out of phase or distorted secondary current in the N CT to be sensed by the trip unit and tripping the breaker.

Have I gone over the cliff looking for a cause or is there some merit in my theory?

Dan

 

[collies99]

Seems plausible but could there be another cause ?

With a loaded building, it may be that there is already appreciable unbalance or neutral current. The motor starting transient would add to that unbalanced current (but not always) and operates the ground protection.

I2t protection is faster at higher current, which may result in faster tripping during starting surges. Any chance of changing the setting to the definite time setting for the ground protection just beyond the transient time during start-up to prevent nuisance tripping ?

Have you measured the neutral current on the main with and without the building loaded just to confirm if neutral current exist?

 

[jraef]

No, it's not flux from the adjacent bus.

It's more likely the Y-Delta starter, because it's likely an open Transition. There can be a HUGE current spike when the starter changes over from Y to Delta, it's called a 'transition spike" and is a very common issue. It's because in the changeover, there is also a phase shift involved, but on an open transition, the motor must be disconnected briefly to avoid a bolted fault. So when reconnected, if there is any remaining residual magnetism in the motor, it is still regenerating when the transition occurs, but the new magnetic field is out of phase with the existing one. So it ends up like connecting two out of phase generators. There is a huge voltage spike and current surge, well known for tripping breakers.

The best solution is to junk the Y delta starter for a soft starter. 2nd best is to change to a Closed transition Y-Delta, but that actually costs more. You can also try 'rolling' the conductors on the 'S' contactor that go to the motor, meaning shift them over by one but maintain the same relative rotation. This changes the phase shift and might mitigate it. That one is free, except for your time.

Read this patent application, it was for a device to indicate if the phase shift was necessary or not, but it describes the problem.

http://www.patentstorm.us/patents/5142213/description.html

"Will work for (the memory of) salami"

 

[VTer]

Make sure the neutral current sensor matches the cb phase current sensor and ensure that all the current sensors match the trip plug. Different sensor ratios will cause you problems.

"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic — and this we know it is, for certain — then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". – Nikola Tesla

 

[dithomas]

Thanks to all for the quick reply.

The starter is a closed transition starter.

The N Ct matches the rating plug. All polarities are correct.

At one time the Ground trip was set at 1000 amps and 0.5 definite time. The breaker tripped with the building in operation.

I cannot imagine that much ground or neutral current. I had the Fluke 434 on the stater when we had the successful starters and saw a perfect start in rush with a 7 second delay to transition. The motor ran a full speed for 1.5 second before the transition. There was only 7 amps of ground current when the motor started. The inrush at transition was minimal.

I will explore the CT and stray flux from phase A theory a little further. Found an ABB white paper on stray flux and CT's.

May relocate the N CT and see what happens.

Thanks.

Dan

 

[dpc]

I had the same thought as Jeff - the open transition wye-delta starters can be a likely suspect. The amount of running load could be a red herring. The transients that occur on transition will depend on the voltage phase angles at the transition, and this is basically random. It might start 5 times in a row and then trip on the 6th.

The ground fault delay is already at max I believe. But you should be able to increase the pickup to 1200 A. It might help.

 

[collies99]

When the breaker trips, does it trips immediately after the start button is depressed or can you discern a noticeable delay time befire tripping ?

Which LED actually illuminated ? Ground LED ?

 

[waross]

Does the breaker trip before, during or after the transition?

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

 

[LionelHutz]

I don't follow your system explanation. You have a neutral CT on the neutral bus you're using to detect ground faults? Shouldn't that CT be on the neutral ground bond wire? Otherwise, you are just monitoring neutral current, not ground current.

I also don't see how a motor starting could cause neutral current when there is no connection between the motor starter or motor itself and the neutral bus.

 

[VTer]

Hi Lionel, this is a residual connection. There are also current sensors on the phases on the main breaker. Under normal operation Ia+Ib+Ic+In=0.

I also do not see how the transition of the starter will cause the GFP to trip. The delay is at .5seconds. The transition does not last that long and the magnitude may not be that large if it is a closed transition starter. I would spend my time investigating the current sensor wiring and make sure that the sensors are operational. Have the sensor been tested? I would also set the GF pickup at 1200A and see what happens.

"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic — and this we know it is, for certain — then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". – Nikola Tesla

 

[dithomas]

Kind of hard to explain.

When the breaker did trip it tripped as soon as the motor started to roll.

The Ground LED lit on the Digitrip520.

Eaton uses a Neutral CT on the Neutral Bus on a 3P 4W system and three phase CT's in the breaker. The Digitrip mounted in the Magnum DS breaker, with the proper jumper installed, looks like an internal zero sequence system. This configuration allows all of the 277 circuits to function and a ground current can be detected because it is outside of the zero sequence circuit.

I am hanging on to my "phase bus too close to the N CT" because in a standard switchboard or switchgear the N CT is far way from any phase bus. If you think about how a CT works my theory is possible. I will know the answer soon.

Thanks for responding. Discuss is always good!

 

[collies99]

You are correct on the Neutral CT. EATON has it for ground source protection on the main. Uses could be back up for ground faults on the feeder or protection for ground faults on the switchboard.

For the GROUND LED to lit, the digitrip must see ground current from the NEUTRAL LED unless the jumper enabling the NEUTRAL CT is not installed and GROUND protection is then actually residual protection even though the NEUTRAL CT is connected, another item to check first.

 

[dithomas]

Vter, Did you see my comment that this is a new custom Magnum DS main breaker section that was install to replace an existing bolted pressure switch and that the neutral bus detail in the main section is in an unusual configuration in order for the normal Eatom neutral bus location below the phase bus to be at the top of the section above the phase bus to match the existing GE feeder sections. i noticed yesterday that the Neutral CT is located within 3/4 of the Phase Bus which is not the usual location for a Neutral CT. I think that the phase A bus flux, when the motor inrush is present, influences the neutral CT secondary current and the trip unit operates.

The existing BPS had a GE ground relay and a zero sequence CT. The GF relay was set at 900A Max time. The chiller was not a problem until the new main section was installed.

I am a semi retired Hot Strip Mill guy that became a Westinghouse/ABB/C-H/Eaton field service engineer and I have seen many strange things.

Facts:
On a Saturday we have started and run the chiller with no problem. A couple of starts where attempted during the week with trip out.
A Fluke 434 set on Inrush indicates a Y-D starter and motor in good condition with typical starting current and transition currents. I have never had the Fluke on when the main tripped.
The motor was baker surge tested and found to be OK
All phase, neutral, and ground conductors are as they should be.
The Digitrip tested Ok with the secondary injection test.
CT polarities are OK
The starter is a closed transition starter. All connections and resistors were check and found to be OK. The chiller is the only large motor on the system.

The only thing that is out of place is the Neutral CT and it's non standard very close (3/4") proximity to the Phase A bus.

Inrush current in Phase A influences the secondary current on the Neutral CT! I will check out my theory this week.

Dan

 

[VTer]

Couple of times already I had a case with Eaton where the wrong ratio neutral sensor for ground fault was packaged with the circuit breaker. Secondary injection test will only give you the status of the trip unit. It does not tell you anything about the sensor circuits or actual sensors. Did you personally inspect the neutral sensor ratio to see if it matched the breaker sensor ratios and the trip plug. Please share those ratio numbers if you have them.

When the new switchboard section was added, did you perform acceptance testing to ensure all components are operational?

"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic — and this we know it is, for certain — then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". – Nikola Tesla

 

[dithomas]

Well guys I did a Mr. Wizard experiment this morning and recreated the field situation.

I simulated the building load conditions and motor starting.

I found that
1. Current thru a bus mounted 3/4" away from a CT will influence the secondary current of the CT.
2. The stray flux will add to the secondary current of the CT when the phase current and primary current of the CT are in phase.
3. The stray current will subtract from the secondary current of the when out of phase.
4. When enough out of phase current is passing thru the bus the secondary current will switch polarity.

The questions now is: How does a Digitrip 520 trip unit react, configured to look like a zero sequence system, to a sudden change to the neutral signal polarity with no change in the phase current polarities.

I was told by a friend that my aluminum foil hat was causing a shift in the global magnetic field!!!!

Dan

 

[rcw retired EE]

I have also seen that problem before.
You already answered your last question: "Ia + Ib + Ic + In= 0".
If the In output switches polarity or is not accurate the sum is no longer zero and the unit trips.

 

[unclebob]

I would do a primary injection in this breaker, making sure the trip unit sees all 3 phase currents.
My guess is that one phase CT is defect/mis-wired internally. The breaker will trip at the GF pick up if only 2 phases are measured by the DT520.
If the load is lower than 800 A, it won't trip. You could verify this easily, if possible, by setting the GF pick up lower than the load current.

In your case, verify that terminals B8-B9 are jumped together (for residual connection and to disable zone interlocking). B5 and B4 are the CT secondary terminals, with B5 being the polarity (X1). B6 and B7 are not used.

Bob