motor stopping rapidly during a fault 16

[electricpete]

We had a 500hp hermetically-sealed centrifugal chiller motor trip on instantaneous current 15 seconds after starting.

Normally I would think instantaneous trip is due to electrical fault, not mechanical origin, however...

Personnel present reported that the machine stopped very abruptly <1 seconds after the trip, whereas it normally takes 10-15 seconds to coast down during normal shutdown.

It occurred to me maybe there is a higher different load torque imposed by the machine at the time of trip during shutdown. When I talked to the chiller mechanical engineer, he said the torque should be roughly the same.

Is it possible that certain fault currents can act similar to dynamic braking to stop motor quickly?

(I kind of doubt it... suspect other simpler explanation... just wanted to ask about this possibility)

 

[DanDel]

Hi 'pete.
An external fault that didn't open a device would make the motor decelerate very quickly, because it would be supplying power into the fault.
If the motor starter or short-circuit device opened under a fault, there would then be no current flow to create any kind of braking. Unless there was a fault in front of the opened device, or some other path for current, like capacitors or resistors?

 

[dpc]

I can't come up with an electrical cause for the deceleration, assuming it is running across-the-line. The motor would contribute fault current to an external fault for only a few cycles, after that it is just a blob of rotating inertia.

Could it be due to some high back-pressure condition on the compressor??

 

[steamboat444]

Does this chiller have an unloader valve that may not have functioned properly?

 

[electricpete]

Thanks for the ideas.

Well, there is more information. Let me start again from the beginning.

While starting the motor following chiller maintenance, the motor ran for 15 seconds and then tripped (all three phases instantaneous). At the moment of the trip there was also a lound bang heard, and the normal decelleration whine was not heard.

Motor was meggered. The term box smelled, one pair of leads was open, all three leads had ground. Motor is definitely bad. But not the end of the story....

Because of the abnormal de-acceleration, we disassembled to the point where we could gain access to the shaft. It is a screw-in connection between motor and compressor shafts. When motor transmits torque to the compressor it acts to tighten that screw connection. It was found totally unscrewed.

Compressor was verified to be capable of rotating freely. We did not get a chance to check free by-hand rotation of the motor yet based on our limited access.

Motor has not been determinated in years, so there is no chance it was connected backwards.

So we have two firm symptoms: motor bad, coupling unscrewed. In addition a band at the time of trip, and lack of decelleration noise.

Under what scenario do those symptoms fit togehter? I have some ideas and we are continuing to troubleshoot but I'm interested to hear your thoughts.

 

[electricpete]

type alert

&quot;in addition a bang at the time of trip...&quot;

 

[OperaHouse]

I had a customer call me from Florida and told me thay sometimes get momentary phase reversals and was loking for something that could stop a motor very quickly. I quizzed him more about this because it didn't seem likely. Sounded like the same setup as yours where it screwed into the motor. Came to the conclusion that what really happened was a phase was lost, this stopped the motor under load, pressure caused the motor to spin in reverse, That caused the motor to single phase spin in reverse, reverse unscrewed the impeller and ground it into the case.

 

[OperaHouse]

Wanted to say that you really need a phase loss relay that ia almost instantanious to remoe all power. Some have delays that can result in a lot of damage if the impeller screws out under power.

 

[electricpete]

We had a field service/maintenance guy from chiller manufacturer Y*rk helping us a little bit.

That gentleman was pretty adamant that a very common cause of shafts unscrewing is momentary interruption and restoration of power (bus transfer/reclose). Worst case in the neighborhood 3-10 cycles or 0.5-1.5* open circuit time constant. We all know that can result in large transient torques capable of breaking shafts. It makes sense they might momentarily be reverse.

IMHO, that reclosure scenario is irrelevant for us. It applies most directly to people fed from the distribution circuits where there are reclosures. We live in a power plant connected directly to 345kv breaker-and-a-half switchyard via transformers with no auto reclose or transfer on any of those breakers.

Here is my theory: Something got caught in the air gap between rotor and stator and caused both (1) – rotor to stop suddently and (2) – stator winding damage. Maybe that something was a rotor bar coming out of it's slot and catching on the stator.

When the rotor stopped abruptly (1), the tendency of the compressor inertia to keep the compressor spinning caused the unscrewing action. When the stator winding got damaged (2) there was instantaneous trip all three phases and severe damage indicated by open circuit measured later.

There has been no further inspection done beyond what I described above. But we have made the decision to rewind the motor and rebuild the compressor (possible compressor damage in many of the scenario's…. tough to rule out without disassembling). Next step is to build the rigging for removal which actually will take over a week due to resources. I don't expect to have any more info until then…. including no ability to check motor shaft for binding.

So plan is set, motor will be rewound, no further problem to solve until we see actual physical evidence. In the meantime it is still interesting to speculate. Does my scenario make sense? Other scenario's?

 

[electricpete]

To Operahouse - I have a comment on your unscrewing scenario. It's an interesting aspect of unscrewing you may already be familiar with but a little non-intuitive. Please don't take offense if I spend awhile explaining how I think it should work.

All standard equipment is screwed in a manner that is tightened by normal forward motor torque. In order for torque associated with the load to unscrew the connection, the load torque must push in the normal rotation direction, not the reverse rotation direction.

An experiment to prove it's true? Hold a comb in front of you in both hands.
Comb represents shaft.
Right hand represents motor.
Left hand represents load/pump.
&quot;forward&quot; direction is comb rotating away from you on top.
&quot;reverse&quot; directon is comb rotating toward you on top.

Now try applying forward torque with the right hand (motor) and restraint with the left.
Compare to applying reverse torque with the left hand (pump) and restraint with the right.

The comb deforms the same either way. If there were threaded connection in the middle it would tighten either way.

Now apply forward torque with the left/pump and restraint with the right. The comb deforms opposite direction (opposite from normal operation torque applied with motor/right). Likewise when you apply reverse torque with the right/motor, the comb deforms in a direction opposite normal.

I think the last scenario is what we had. Motor bound up possibly due to rotor bar catching on stator and applied reverse torque causing unscrewing. If it was just the bearings binding, there would be no reason for a simultaneous electrical fault on the motor.

Sorry to go at length into the unscrewing thing. We talked about it a lot today and it's very fresh in my mind. I think most people would jump to the wrong conclusion based upon intuition initially.... i.e. assume that pump needs to push reverse to cause unscrewing. That is not the case. The pump in danger of unscrewing is the one which has higher external pressure applied to it's suction than it's discharge.

 

[busbar]

There are larger than 500hp hermetic-refrigeration units, but they can make an expensive mess with a stator electrical failure in refrigerant. Seems like putting a lot of eggs in one basket.

I can’t add much advice to the situation, but if there are any sort of warrantee issues with the equipment, I would want to see one of your guys present when the rebuilder first breaks it open. Something mechanically seized associated with short-order winding-insulation failure is a strong possibility.

 

[jbartos]

Suggestion: If the fault, e.g. phase to phase to phase, was somewhere between the motor terminal box and the motor protector, then the motor protector would trip instantaneously, if the instantaneous trip was present, and the motor would stop abruptly, e.g. within less than one second, since all the motor dynamic energy would be discharged into the fault. This type of occurrence may be reconstructed to confirm it.

 

[electricpete]

busbar/jbartos - thanks for the comments.

I struggle with the question of whether a fault at the motor leads or any point inside the motor would cause rapid deceleration.

I am skeptical of that scenario. But I would be interested to hear any more thoughts on it or whether anyone has seen that effect in action.

 

[jbartos]

Suggestion: It would not be difficult to set up an experiment, e.g. with a small three or single phase motor, which would include a bolted short in the feeder, i.e. motor leads, to see the motor behaviour.

 

[GusD]

Hello Electricpete

Folowing your last lead regarding the abrupt stop being caused from possible rotor bar problems.The rewind shop should be able to have a very good idea of what the problem was once they take it apart.
Are you thinking of overhauling compressor because of the
motor problem or is it time for an overhaul?
I am curious to know what the motor shop comes up with.

Regards
GusD

GusD

 

[electricpete]

You are right, the rotor bar scenario would be abundantly clear when we get a look. Still that is over a week away and a lot of people are asking questions.

So I go though the thought excercize which I think is interesting. Entering a guess is free, and if the motor tear-down proves you right, then you have bragging rights to say you knew it all along ;-)

Compressor was overhauled 10 years ago but that is the most recent of any of our 8 compressors (20 year cycle)... so it is not exactly due for overhaul.

Compressor teardown is based upon strong recommendation from Y*rk and a group decision-making process from the people in the room at the time. One aspect is that if my scenario holds true, the axial movement of motor or compressor must have been at least 1/4&quot; to unscrew. Normally motor is free to move axially but not clear if it would be during the jammed rotor bar or other scenario's. If the compressor moved more than 20 or 30 thousandth of an inch axially, it's thrust bearing is probably damaged.

We have no easy way of checking the thrust bearing. Don't have motor available to run to check the compressor. Waiting for motor rewind to find out that the compressor is not good would set the schedule too far back. With the extent of disassembly necessary to inspect the thrust bearings you might as well do an overhaul. Also there is a degree of uncertainty do we really understand the problem and will we understand it much better in the near future.

 

[edison123]

pete,

Did the motor bearing(s) seize causing abrupt stop ?

 

[electricpete]

edison - it is a possibility. But in this case would there be any logical reason that the motor would trip on instantaneous overcurrent all three phases and upon investigation windings found grounded/opened?

 

[DanDel]

Couldn't it just be a winding failure causing Inst trip of the external device and rapid motor decel because it was feeding the fault for a few cycles? The compressor might have had more inertia as the motor decelerated and unscrewed itself at the coupling.

 

[edison123]

pete,

If the bearings had seized at the time of motor start, then definitely overcurrent trip will occur due to almost locked rotor condition. If the windings were old and if the stator current were high, then stator winding could have been damaged even leading to earth fault. Was the overcurrent relay calibrated in the recent time to check its operation ?