Motor causing ground fault; now gone, why?

[tankfarms]

For the past 2 days, I've been constantly getting a ground fault alarm (we have an ungrounded system here) on one of our 480V MCC bus. The alarm toggles simutaneously with the run status of a pump we have. (we have a DCS system here that we can monitor). So, it's simple, we isolate the pump, and check service/load.

We got electricians check it, each phase-to-gnd megged good (2 Gohms), he disconnected the leads from the load side of the starter, and pulled the starter close, and no ground alarm; then he re-connected the leads back to the load side of the starter, and ran the motor, again, no ground alarm either, and from this point on, we no longer get this ground fault as we run the motor.

Just some background info
The pump is on a level control, so it's not continous run
Motor FLA=86amp, 3-PH, 75HP, SF=1.15, 480V
it was measured that the motor was pulling approx. 100amps when it run

could it be possible that something (thermally) is going on inside the motor? as I know the motor is overloaded at the moment, which is a seperate issue we've been working on, but could it be this causing some intermittent insulation problem which then causes a ground fault alarm?

At the time writing this, the motor kicks on/off with a level control, and no ground alarm is observed at the moment. kinda mysterious here.

Any input is welcome. Thanks.

 

[weressl]

For the past 2 days, I've been constantly getting a ground fault alarm (we have an ungrounded system here) on one of our 480V MCC bus. The alarm toggles simutaneously with the run status of a pump we have. (we have a DCS system here that we can monitor). So, it's simple, we isolate the pump, and check service/load.

We got electricians check it, each phase-to-gnd megged good (2 Gohms), he disconnected the leads from the load side of the starter, and pulled the starter close, and no ground alarm; then he re-connected the leads back to the load side of the starter, and ran the motor, again, no ground alarm either, and from this point on, we no longer get this ground fault as we run the motor.

Just some background info
The pump is on a level control, so it's not continous run
Motor FLA=86amp, 3-PH, 75HP, SF=1.15, 480V
it was measured that the motor was pulling approx. 100amps when it run

could it be possible that something (thermally) is going on inside the motor? as I know the motor is overloaded at the moment, which is a seperate issue we've been working on, but could it be this causing some intermittent insulation problem which then causes a ground fault alarm?

At the time writing this, the motor kicks on/off with a level control, and no ground alarm is observed at the moment. kinda mysterious here.

Any input is welcome. Thanks.

A relatively high resistance ground fault can be intermittent based on variuos factors what is IN the ground fault path and how high is the current, etc. If it develops into a high resistance phase-to-phase fault taht could burn the path out without tripping OL or SC protection but now you will have a weak spot somewhere that eventually results in a high level fault. There are various 'noise' generator type fault finders or TDR devices that can locate you fault.

 

[iwire]

could it be possible that something (thermally) is going on inside the motor? as I know the motor is overloaded at the moment, which is a seperate issue we've been working on, but could it be this causing some intermittent insulation problem which then causes a ground fault alarm?

I am certainly not a motor expert but it would not surprise me that you could have a fault that for now is only making when the motor is hot and expands and with this being an ungrounded system it is not 'blowing a hole' in the surrounding insulation when it does fault.

If it develops into a high resistance phase-to-phase fault taht could burn the path out without tripping OL or SC protection

Seems unlikely with this being an ungrounded system.

 

[mull982]

For the past 2 days, I've been constantly getting a ground fault alarm (we have an ungrounded system here) on one of our 480V MCC bus. The alarm toggles simutaneously with the run status of a pump we have. (we have a DCS system here that we can monitor). So, it's simple, we isolate the pump, and check service/load.

We got electricians check it, each phase-to-gnd megged good (2 Gohms), he disconnected the leads from the load side of the starter, and pulled the starter close, and no ground alarm; then he re-connected the leads back to the load side of the starter, and ran the motor, again, no ground alarm either, and from this point on, we no longer get this ground fault as we run the motor.

Just some background info
The pump is on a level control, so it's not continous run
Motor FLA=86amp, 3-PH, 75HP, SF=1.15, 480V
it was measured that the motor was pulling approx. 100amps when it run

could it be possible that something (thermally) is going on inside the motor? as I know the motor is overloaded at the moment, which is a seperate issue we've been working on, but could it be this causing some intermittent insulation problem which then causes a ground fault alarm?

At the time writing this, the motor kicks on/off with a level control, and no ground alarm is observed at the moment. kinda mysterious here.

Any input is welcome. Thanks.

Of course your first step is always to check the ground fault monitoring control wiring including that to your DCS to make sure there are no loose connections causing a false signal.

Secondly your comment about something getting hot and causing intermittent ground connections though the insulation is very possible. I had a case once that drove me nut where a cable had a weak spot that would go to ground everytime the cable was loaded heavily and got hot and would trip the breaker. We must have megged the cable almost 10 different times and always had extremely high readings that indicated cable was fine. Finally one day electricians started pulling off every cover in boxes along the run until they discovered the bas spot in the wire. We replaced the cable and the problem went away for good.

It is very possible that you could have the same thing going on when the motor windings or motor feeder cable heat up.

One thing you could do to verify there is a real ground fault condition happening and not some false alarm due to control wiring would be to monitor the L-G voltages on all 3 phases. If and when one of the phases goes to ground even if its intermitent you will see the L-G voltage on that phase go from its normal L-G reading to a reading of 0V. If you can trend or monitor this with a power meter you can then compare event times when you get the alarm on the DCS to the power monitor to hopefully see what is happening is real. May also be a good idea to monitor motor amps if you are not already to see what the motor is pulling when the alarm comes in again.

 

[tankfarms]

Of course your first step is always to check the ground fault monitoring control wiring including that to your DCS to make sure there are no loose connections causing a false signal.

Secondly your comment about something getting hot and causing intermittent ground connections though the insulation is very possible. I had a case once that drove me nut where a cable had a weak spot that would go to ground everytime the cable was loaded heavily and got hot and would trip the breaker. We must have megged the cable almost 10 different times and always had extremely high readings that indicated cable was fine. Finally one day electricians started pulling off every cover in boxes along the run until they discovered the bas spot in the wire. We replaced the cable and the problem went away for good.

It is very possible that you could have the same thing going on when the motor windings or motor feeder cable heat up.

One thing you could do to verify there is a real ground fault condition happening and not some false alarm due to control wiring would be to monitor the L-G voltages on all 3 phases. If and when one of the phases goes to ground even if its intermitent you will see the L-G voltage on that phase go from its normal L-G reading to a reading of 0V. If you can trend or monitor this with a power meter you can then compare event times when you get the alarm on the DCS to the power monitor to hopefully see what is happening is real. May also be a good idea to monitor motor amps if you are not already to see what the motor is pulling when the alarm comes in again.

That's exactly what I did. It's being monitored at the moment. Will keep you posted here.

 

[Jraef]

What kind of GF monitoring do you have? Zero Sequence (with a core balanced CT) or Residual Current? What is the trip or alarm threshold? These are important factors in determining where to start chasing this problem.

For example, if it's a ZS GF, a very small amount of ground current could be triggering the response, such as one might find when a warm motor is turned off in a moist environment, and who's insulation is not quite failing, but is certainly leaking when wet for a period of time. So the first time you energize it after a long enough period, there is a small leakage to ground until it warms up and evaporates the moisture. Very common problem. If it's a RC GF, the current necessary to cause a trip is significantly higher, but starting current imbalance may be the trigger. Every motor has slightly different winding resistances between the phases and although it usually doesn't make much of a difference, under the right conditions that imbalance can get magnified and could be enough to be causing a trip.

So do a little more investigation and think about those kinds of situations.

 

[hurk27]

What would a single phasing issue cause with a GF detector on the system?

Could there been a loose connection on the load side of the starter but when the electrician removed the conductors and replaced them the problem went away?

 

[tankfarms]

What kind of GF monitoring do you have? Zero Sequence (with a core balanced CT) or Residual Current? What is the trip or alarm threshold? These are important factors in determining where to start chasing this problem.

For example, if it's a ZS GF, a very small amount of ground current could be triggering the response, such as one might find when a warm motor is turned off in a moist environment, and who's insulation is not quite failing, but is certainly leaking when wet for a period of time. So the first time you energize it after a long enough period, there is a small leakage to ground until it warms up and evaporates the moisture. Very common problem. If it's a RC GF, the current necessary to cause a trip is significantly higher, but starting current imbalance may be the trigger. Every motor has slightly different winding resistances between the phases and although it usually doesn't make much of a difference, under the right conditions that imbalance can get magnified and could be enough to be causing a trip.

So do a little more investigation and think about those kinds of situations.

Thanks, unfortunately, no such drawings are available to tell what kind of relay is being used, but I assume it works on observing the L-G voltages. So, sometimes, it may not even be a "ground", rather, could be a voltage dip due to some power issues on the line.

 

[tankfarms]

for the past 24 hours, it seems this ground has gone. Same motor has been operating under the same condition with automatic on/off pump controller. We didn't really do any "repair" yesterday when we checked, the only thing that we did was disconnecting the starter load-side leads and reconnecting them back, now I suspect it was a connection issue that may have caused this "ground". How could a connection issue cause a "ground"? I wonder.

 

[Jraef]

for the past 24 hours, it seems this ground has gone. Same motor has been operating under the same condition with automatic on/off pump controller. We didn't really do any "repair" yesterday when we checked, the only thing that we did was disconnecting the starter load-side leads and reconnecting them back, now I suspect it was a connection issue that may have caused this "ground". How could a connection issue cause a "ground"? I wonder.

Something connected to ground!

You can speculate all you want, but unless you know how your ground fault alarm system operates, you are just guessing.

 

[G._S._Ohm]

It could also be capacitance to ground.

 

[stew]

I can tell you one thing for certain and that is with this motor having a full load nameplate amperage of 84 and it drawing 100 amps full load it wont be long before your problem will be very apparent. This is nearly 30% overload on this unit and I would guess that the thermal degradtion caused by this overheating is the root cuase of the problem.It wont be long before the load lets the smoke out on this motor!!

 

[Jraef]

I can tell you one thing for certain and that is with this motor having a full load nameplate amperage of 84 and it drawing 100 amps full load it wont be long before your problem will be very apparent. This is nearly 30% overload on this unit and I would guess that the thermal degradtion caused by this overheating is the root cuase of the problem.It wont be long before the load lets the smoke out on this motor!!

Your math is a little off, and it's only 1A over the Service Factor, but your conclusion is not that far fetched at all. Running a motor into the SF continuously will definitely increase the motor temperature, which does degrade the insulation. That may be what's behind a moisture problem when it has been sitting for a while. If that's what the problem is...

 

[stew]

yes my math is a little off but so is yours. . 84 amps at 1.15 service factor is 96.6 amps. 1.2 service factor is 100.8 amps. running any motor continuosly at service factor or above will result in rapid and premature insulation breakdown fer sure!. dont know how i came up with 30% but 20% is definatley a problem!!

 

[LHarrington]

Tankfarm,
Let me start by says you have received answers for some of the best that this forum has to offer. I have encounter and very similar situation to yours came to work one Monday morning and two identical compressors with 200hp motors that were failing on GF alarms. I was able to get one running temporary and the other had to be sent to the rewind shop immediately to be rewound because the insulation in the windings had failed. Once it was finished and on line we pulled the second and had the same procedure done. The one thing we had in common the day these failed was it was raining cats and dogs there was a lot of moisture in the air. It was the moisture that provided the last missing piece of the puzzle to start the GF troubles.
Good luck and listen closely to these individual advising you, they are the best asset to have and better than any tools you can buy.

LHarrington

 

[weressl]

yes my math is a little off but so is yours. . 84 amps at 1.15 service factor is 96.6 amps. 1.2 service factor is 100.8 amps. running any motor continuosly at service factor or above will result in rapid and premature insulation breakdown fer sure!. dont know how i came up with 30% but 20% is definatley a problem!!

Thermal design of insulation is such that a motor running at service factor will lasts it's lifetime design, if no other parameters exceed the design limits. Running it above will shorten the insulation life.

Not all motors designed to run with the same temperature rise under normal, full operating load. As an example IEEE841 requires the use of Class F and limits the average winding temperature rise to Class B or 80*C and limits the maximum exposed temperature to 200*C. Some run hotter, like a range of seven 10HP motors tested the temperature rise ranged between 47*C to 86*C, actually exceeding the design limits set by the Standard. (This gives you an example where the Standards are set, but no third party independent agency verifies it, the actual compliance is always suspect when manufacturer self-proclaims compliance. In other conuntries the Manufacturers are required to self-certify, so they are liable if the equipment fails to perform to certified values.) If the same insulating materials used, obviously the insulation of the hotter running motor will fail sooner.

 

[stew]

and in this case the motor is running above the service factor so we can expect that this insulation is rather crispy critters already depending on how long this condition has existed. Our motor shop rewound lots of these that came in well cooked. We wound with all class H materials so we were able to give our customers a better winding than the manufacturer gave on the original. Whenever we saw this type of failure we always tried to alert the customer to the fact that thier was either an overloaded condition or voltage problem that needed to be checked after the unit was reinstalled.We were a factory warranty service center for manufactures such as Lincoln and Baldor and I was fortunate to have been able to attend motor failure anaysis training session for both at the factory level, Became quite proficient in analysis of failure by visual inspection of failed units.Interesting work.

 

[weressl]

and in this case the motor is running above the service factor so we can expect that this insulation is rather crispy critters already depending on how long this condition has existed. Our motor shop rewound lots of these that came in well cooked. We wound with all class H materials so we were able to give our customers a better winding than the manufacturer gave on the original. Whenever we saw this type of failure we always tried to alert the customer to the fact that thier was either an overloaded condition or voltage problem that needed to be checked after the unit was reinstalled.We were a factory warranty service center for manufactures such as Lincoln and Baldor and I was fortunate to have been able to attend motor failure anaysis training session for both at the factory level, Became quite proficient in analysis of failure by visual inspection of failed units.Interesting work.

Let's get back to the original data and problem. The OP stated that the current was ~100A @ 86FLA nameplate, which is approximately 1.27% over service factor. Hardly a critical condition, especially when the OP posted that this was a cyclical operation giving off time to the motor to cool off. Without knowing what ambient operating temperatures existed and how long this condition existed, the crispy-critters expectation is not supported by the information. The OP further stated that the GF condition existed for the past two days, then disappeared, so I would have difficulty of making the above logical extention form that information.

 

[bobsherwood]

We had an AHU doing that. Turned off on GF every once in a while. We did not find the problem until we broke down the motor. We found one little burn spot on the rotor and a small burn on the stator. We figured that if the motor tried to start when these two burn spots where lined up, it would trip off???? Not a motor expert but, I could not find anything else wrong with it. The burn spots where small enough that there was no direct short when we megged it.????