Chiller Ground Fault Issues

[brian john]

A bit long winded, but I could use some help/insight.

We have a site with several Chillers’ they have reduced voltage starting.
One of the chillers keeps tripping the feeder circuit breaker on ground fault at start up.
Chiller is a 3 phase 480 VAC
Once started the chiller runs fine.

Here is what we have.
800 amp GE circuit breaker with a MicroVersa Trip PM Trip Unit, with a 300 amp rating plug
There is a neutral CT that is not utilized, though it is connected to the circuit breaker.

Long Time 6 delay 24
Instantaneous 7
Short Time at 9 delay 1
Ground Fault .2 and delay 1

The circuit breaker was primary injection tested last year
Secondary injection tested last week and tested satisfactorily
The GFP was primary injection tested last week and tested satisfactorily

The GFP pickup is based on the circuit breaker sensors so ground fault pickup is 160 amps.

We utilized a high speed recorder to monitor voltage and current, we monitored A, B, and C phase currents plus a single CT was placed around all three phase conductors and a separate CT was placed on the EGC.

In our test we did not get a trip of the feeder circuit breaker, this is what we did see.

AT START UP
A, B, C inrush currents in excessive of 1,100 amps, (due to limitations of the CT’s we could not capture the exact inrush currents)
The zero sequence CT-150 amps
The EGC CT-135 amps (it is not feasible to capture true ground current as there are multiple ground paths).

Prior to us arriving on site the building electricians swapped circuit breakers between chillers and had the same issue with this chiller with a different circuit breaker.

This chiller has been on line for about 10 years and the issue occurred a few times, but in July the tripping on GFP became a regular occurrence.

The chiller was meggered at 1000 VDC and the readings were above 2000, Megohms (by others)
The feeder was meggered with similar results.

We could not monitor the other chillers at this site as they were in use, I’d like to see if there is a similar issue with other units.

With transformers we have had ground fault tripping issues when the GFP relay was set low, the design engineers made changes in the coordination study to overcome this issue.
Is it reasonable to assume that there would be some zero sequence current on a 3 phase high inrush load on start up?
Or is my assumption incorrect?
Is it possible to have induced current on the EGC during high inrush start ups? (Is this what I am seeing?)
This chiller operated with minimal issues for 10 years, could this issue have been existing all along and only gotten worse due to aging of the motor, bearings or something with the chiller that added additional load during startup?

 

[hurk27]

To me it sounds like you have a intermediate ground fault on start up, it could conductor insulation worn off over time, or a poorly made up connection that is making a ground connection as the inrush current is jumping the conductors around, but it does sound like something is grounding as it starts, I doubt that at 160 amps it's capacitive coupling, and even at start up you should not have start up current on the grounding, I would be looking at all the connections to the compressors, back to the supply but it does sound like something is going to ground, have you tried to isolate it down to one phase? can you put a recorder on each phase to see if you have a higher inrush on one, because the one you have is maxed out your not going to see it with that one, you will need to obtain CT's that will increase your range, or loosely couple them so you can see which phase is higher on the inrush, but this is not a very accurate way to do this, but might get you in the ball park enough to see which phase might have the ground fault in it. it could even be in the windings if you have a loose winding and it only faults at start up.

 

[zbang]

Did you look at phase-ground voltages? Following the intermittent ground theory, if one lead is grounding, that voltage should be significantly lower.

 

[brian john]

We do have that information I will recheck....

 

[ATSman]

GF Jumper

GF Jumper

800 amp GE circuit breaker with a MicroVersa Trip PM Trip Unit, with a 300 amp rating plug
There is a neutral CT that is not utilized, though it is connected to the circuit breaker.

If I seek the resources of my feeble mind, I recall that on a three wire circuit such as a chiller motor, if the supply breaker is four wire with a neutral CT then there is a terminal block on the side of this breaker that allows you to change a jumper connection to short out the GF CT circuit to prevent tripping.
The trip unit internal logic is looking for neutral CT current to balance the load current that does not exist since a motor is 3 wire. It may be a matter of moving this jumper to correct problem. I think you would still maintain GF protection on the motor but am not sure. I do not think you have an actual ground fault. My first thoughts but I will look further into this. Anyone recall?
Where is Zog or Gar?

 

[brian john]

800 amp GE circuit breaker with a MicroVersa Trip PM Trip Unit, with a 300 amp rating plug
There is a neutral CT that is not utilized, though it is connected to the circuit breaker.

If I seek the resources of my feeble mind, I recall that on a three wire circuit such as a chiller motor, if the supply breaker is four wire with a neutral CT then there is a terminal block on the side of this breaker that allows you to change a jumper connection to short out the GF CT circuit to prevent tripping.
The trip unit internal logic is looking for neutral CT current to balance the load current that does not exist since a motor is 3 wire. It may be a matter of moving this jumper to correct problem. I think you would still maintain GF protection on the motor but am not sure. I do not think you have an actual ground fault. My first thoughts but I will look further into this. Anyone recall?
Where is Zog or Gar?

I thought about that and need to dig the GE instruction book out. But if the 3 phase load is balanced as it is in a motor (except at start up?) the neutral CT is seeing what it normally would see no current.

 

[ATSman]

I thought about that and need to dig the GE instruction book out. But if the 3 phase load is balanced as it is in a motor (except at start up?) the neutral CT is seeing what it normally would see no current.

You hit the nail on the head.
On startup the sine wave becomes distorted and is no longer cancels by the logic circuit and appears as a GF.
Don't forget a motor winding electrically looks like a transformer with a shorted secondary till the rotor gets up to speed. Then back EMF takes over and the current drops to normal.

 

[Jraef]

You don't mention what kind of reduced voltage starter you are using. If it's Wye Delta, that can cause all kinds of problems and if it is solid state, a severe imbalance on startup could be an indication of a malfunctioning ramp control board or a failing SCR.

 

[zog]

What is the CAT# on the MVT+?