MV Motor Runs Uncoupled, Locked Rotor Fault When Coupled 3

[Plum_Point1]

We have a Siemens 1750 HP 3600 RMP 4000V motor that is used on an air compressor. We recently had the windings replaced. All motor shop test results looked fine. We got the motor back, done an uncoupled run to identify magnetic center, aligned and coupled the motor. We get a "locked rotor fault" when attempting a coupled run. We uncoupled the motor and it will find magnetic center and run fine, no issues. It is only on the coupled run that we have an issue. We have realigned and coupled multiple times with oversight from our local air compressor rep even and are still dealing with the issue. We have done extensive motor testing and have found nothing wrong. Any ideas on what we could be dealing with here?

 

[waross]

Misplaced decimal place on the current limit parameter of the drive?
Compressor seized?
Verify that the compressor is free to turn.
Re-check drive parameters.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Plum_Point1]

This motor is run off of an autotransformer (korndorfer) starter. The fault is coming from our motor protection relay which has all previous settings that it had when the compressor ran fine before the motor repair. The compressor is free to turn by hand so doesn't seem seized. The start current is going to around 800 amps on start, which is normal, and is staying high (above 700 amps) until the motor trips on locked rotor fault from the motor protection relay.

 

[LittleInch]

Going the right way round?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[waross]

Is the motor turning?
Is the motor accelerating at all?
Have you checked for an open phase?
I have seen that on an auto-transformer start.
An open auto-transformer winding, a blown fuse or a badly burned contact may reduce the acceleration so that the protection trips before the motor comes up to speed.
Is the unloader working properly on the compressor?
Seen that also.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Plum_Point1]

Rotation is correct.
The motor does turn and seems to be accelerating. We have checked for open phase. Everything at the breaker (autotransformer, contactors, etc.) all tested good. Unloader valve has shown open.

 

[RRaghunath]

It seems the motor torque has reduced following the rewind and hence the motor is taking longer time to come to full speed.
Since the motor is turning (couple run), you may try increasing the locked rotor trip delay setting in the motor protection relay to allow a little more time.

 

[waross]

If delta windings were reconnected in star, that may explain the slow acceleration.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[electricpete]

Most of our large motors have time overcurrent (overload) trips, but not "locked rotor fault". In case on some very large motors (8000hp with large flywheel) where we have relay schemes that are termed as "locked rotor", it is supervised by a distance relay that judges the acceleration by the apparent impedance / angle. If those particular relays tripped when we observed that the motor was actually accelerating, I would investigate a relay malfunction.

But you may have a different type of relay / logic for this "locked rotor". Maybe you can tell us more about it and whether you judge it to be inconsistent with observed acceleration.

Let's set aside the locked rotor terminology and assume this was a time overcurrent trip. Then the logical factors might be:

1. Power supply. Low or unbalanced voltage provided at the motor terminals during acceleraiton. This also could include malfunctions of your korndorfer reduced voltage starting equipment
2. Reduced torque capability of motor.
3. Increased torque demand of mechanical load.

#1. and #3. are less likely since you didn't change anything (but still not ruled out). 2. is the first item to focus on and I'd say within category 2 some of the possible items in order from most likely to least likely could include:

[ul]
[li]Incorrect field hookup. Similar to what BILL mentioned.[/li]
[li]Incorrect winding performed by the rewinder.[/li]
[li]High resistance termination causes unbalanced current and reduced torque.[/li]
[li]Rotor bar degradation (not obviously related to previous work but can reduce the torque[/li]
[li]Coupling up sleeve bearing machine in the wrong axial location could in theory misalign rotor and stator iron to reduce effective length of rotor/stator overlap and torque capability although I've never heard of that causing a problem (it would have to be a really marginal starting application to begin with and maybe this could push it over the edge).[/li]
[/ul]

Among the troubleshooting, I would be inclined to try to capture starting waveforms to see if there is indication of sustained current unbalance or anything else unusual. Perhaps rotate manually if you have a historical basis for comparison.

 

[Gr8blu]

Plum Point The way you describe the trip message (locked rotor fault that occurs under coupled conditions only) leads me to believe that the old current-vs-time data does not match with what you have for the rewind. This means one of only a couple of things: 1) is there a back-pressure condition in the compressor (for example, a valve closed that should be open, or an accumulation of moisture within the compressor itself)? 2) something else happened to the compressor (e.g. a bearing issue) that got covered up by the need to rewind the motor in the first place and is only now coming to light as an increased torque requirement at some point on the speed curve, or 3) a change in the winding itself as aa result of the rewind.

You indicate that you've checked the unloader and it is working correctly - did you also look for condensation?
You also think the compressor turns "freely" - but is there undue noise coming from it during the acceleration that might point to a different mechanical issue?
Was the repair done by an accredited shop that is also familiar with the original machine design? Or was this their first time with this kind of equipment? Did they use their regular burnout process - and possibly damage the lamination steel coating, creating additional losses that might be just enough to tip the scales in terms of accelerating torque? Did they connect the winding incorrectly - say a WYE (star) connection instead of a DELTA? Or an incorrect turn-per-pole and/or circuit connection?

Is this unit part of a larger process - i.e., are there other units that are pressurizing the downstream circuit? Was this unit supposed to be started against an "empty" process or at some other point? I did run across that once: the unit in question was supposed to start first because it did not have enough "oomph" to overcome the residual back pressure created by the other units when one or more of them was in operation. Once the user shut the whole system down, the "suspect" unit started just fine against an "empty" line.

Converting energy to motion for more than half a century

 

[edison123]

Is the decoupled no-load current before and after rewind the same? If yes, then rewind is fine.

Muthu

http://www.edison.co.in

 

[zlatkodo]

Edison is right.
We need to know the NLA both before and after the rewind.
ACW

 

[Plum_Point1]

I haven't examined the relay files yet to fully understand what the "locked rotor fault" is exactly showing, but it is obvious that the current never drops down after startup. We have another compressor that is redundant to this one that jumps to 800 amps on startup and is below FLA (215) within 4 seconds or so for comparison. At that time this compressor is still well above 700 amps as previously mentioned.
No noise or vibrations during the attempted runs. The repair was done by an accredited shop that is familiar with this design.
As far as the winding connections go there is no evidence to show whether it was made up correctly or not. All of our testing says it is, but no pictures from the shop and we haven't opened it back up yet.
At the moment the only doubt we have with anything physical is with the coupling and whether the grease is compacted enough to now allow the float it needs to get to magnetic center. By hand we weren't able to move the coupling axially at all. Yesterday we pulled the grease plugs and moved the coupling axially to get rid of some grease and was able to then move the coupling axially with ease. We plan to attempt another test start this morning after doing that and moving to magnetic center prior to the start. All process related variables have been confirmed to be lined up for a normal start. Drains have been open to ensure no moisture buildup.
I don't believe we have any data from a no-load run from before the rewind that I am aware of to be able to compare. I will request this from the motor shop from the previous rebuild.

 

[waross]

I doubt that magnetic center is a problem, but , to that end;
Does the compressor have end play?
Does the coupling allow for expansion?
Is there anywhere in the design that will allow for thermal expansion?

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[electricpete]

I don't really know what data you have in front of you.

If it were my plant, here is some data I'd think about gathering during the next attempted start:

[ol 1]
[li]* See if you can capture the waveforms for the 3 phase currents (on the same trace) during start. Often that can be done with minimum fuss and tells a lot even without the voltages[/li]
[li]* Also try to capture the voltages if do-able (pt secondary of course). It helps put the current in context and also may reveal power supply problems.[/li]
[li]* Capture a vibration time waveform during start (ideally 1k samples / sec). From that you might be able to estimate speed vs time. That's just data, easy to gather if you have an on-staff competent vibration guy. Whether it will give any useful insight remains to be seen.[/li]
[li]* Timing - Watch time from start to trip (of course) and then also watch the coastdown time. Check direction of rotation during coastdown (it's easy to see then) as someone mentioned although I'll bet you've already done that. Coastdown time falls in the category of data... may be difficult to interpret since you don't have a direct comparison that trips during start, but it's easy data to gather which might provide insight about unusual friction or loading during coastdown.[/li]
[li]* Timing of the korndorfer. I'm not really familiar with it but how long does it take to get to full voltage and is that before or after time of trip? If the expected timing can be verified during the run, that may help.[/li]
[li]* observe air system parameters for clues. you know more about that than i do.[/li]
[/ol]
Contact repair shop to discuss and get the full written report on what was done if available:
[ul]
[li]How did the stator-only surge test look (any indication of differences between phases can reveal winding differences).[/li]
[li]How does the winding resistance test look (well balanced).[/li]
[li]Was the rewind like for like, or did they make some changes?[/li]
[li]Was the no-load current unbalance reasonable considering the voltage unbalance?[/li]
[li]What other thoughts do they have about your situation?[/li]
[/ul]

 

[oldfieldguy]

Had a client send a 7000HP 4160V motor in for rewind. It came back with the same symptoms you describe.

Got with the rewind shop. Instead of 2300-volt windings to connect in a wye for 4160, they did 4160 windings. When hooked in a wye, those expect 7200 volts.

Motor ran well when uncoupled. That's no mystery, we used to commonly run 2300-volt motors on 480 to check bearings in the shop.

old field guy

 

[zlatkodo]

Hi, Plum_Point1
It would be good to know how the motor behaved prior to the last failure.
It's possible that the same cause of the previous motor malfunction, has not yet been resolved.
Or if you have pictures of the disassembled motor illustrating the condition of the winding, stator, rotor, etc. before the last repair.
If it's not confidential, please provide the card from the rewind shop showing winding details and main measurements of the stator core in order to check whether the connection should be Wye or Delta. This information is likely available if the rewind shop is accredited.
Recently, we had a case where a similar motor was incorrectly internally connected as 1Delta instead of 2Wye (3 leads out).
Winding Design and Motor Repair

 

[jraef]

Just so you are aware, there is a possibility that your Korndorfer starter is the problem. Siemens in particular has investigated this issue and did a white paper on it for IEEE. Look for the following:

“Medium-Voltage Reduced-Voltage Autotransformer Starter Failures—Explaining the Unexplained“ Paper PID-04-26, presented at the 2004 IEEE Pulp and Paper Industry Conference, Vancouver, BC, Canada, June 27–July 31.

The basic conclusion is that the Korndorfer stating method, when applied at MV, puts a lot of voltage stress on the transformer windings, causing them too short. It might be that your need to rewind that motor was a symptom of the auto transformer in the starter having failed first, so that original cause still exists and is potentially damaging your new motor windings!

This has come up for me on a couple of occasions where users have asked to replace MV RVAT starters with Solid State Soft Starters because of this issue. I had never heard of that before.


" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden

 

[LittleInch]

Jesus, you've got a good memory!

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Plum_Point1]

We were able to successfully start the air compressor motor after removing a plug in the coupling to relieve grease and allow for axial movement. We also moved the shaft out to magnetic center prior to the start attempt. It seems the culprit was us being hydraulically locked from the grease in the coupling. Thanks for all the input everyone.