Generator rotor fault

[Addy71]

Case study - 34 MW generator, was not generating sufficient reactive power. Checked at FSNL, PMG current is 12 A (typically 4A). Checked all points, found main rotor IR = 0 M?.

Rotor threaded out & DC voltage drop test conducted on rotor (DC to winding, checked drop between winding & ground). Rotor coils = 12. Large drop occurred between 2nd & 3rd coils, +ve to –ve changeover occurred at the 6th coil. Boroscope inspection showed melted copper between 2nd & 3rd coils under the turbine end retaining ring.

Next, passed 250 A DC current through the rotor body & checked drop between winding & body. Null point was detected between the 2nd & 3rd coils about 700 mm for the turbine end retaining ring. We have now started the repair activity.

Other observations: Both bearings have significant pitting marks on them, forming an axial band across the bearing lower halves. Rotor winding resistance has gone down from 0.235 ? to 0.116 ?. Bearing pedestal insulation is healthy.

Questions:

1. Do we have a single or double rotor earth fault? Can we expect a fault at the 6th coil also?

2. Have the coils been shorted too? Or is the lower resistance due to the earth fault?

3. Why did the bearings get damaged by shaft current? As these are white metal bearings with no prox. probes, how does one detect this problem in advance? I have seen high amplitude, high frequency peaks in case of anti-friction bearings. Would these also show up for plain bearings?

4. A shaft grounding brush was provided & was found healthy. Why did it not take care of the shaft current? Could the earth fault be responsible?

5. Could electrical signature analysis have some early indication of this problem? Has anyone experience in monitoring generator rotors with it? We have an EMPATH unit that could be tried I guess. It did not pick up anything in the stator current & voltage when we checked two months back.

Any ideas on these issues will be highly appreciated.

Regards,

Aditya

 

[waross]

1 To avoid further damage, I suggest you megger the remaining coils to ground with the known bad coils out of the circuit.
2 Melted copper is a pretty good indication of shorts and grounds.
3,4. I'll defer to others on these.
5 A suggestion for discussion. Monitoring the voltage of individual stator coils (The coils on one stator pole.)would possibly show a lower voltage when the failing pole passed.
You may be able to display the voltage from the coils on one stator pole on a scope. If the time base of the scope is set to one revolution you may be able to see the bad field pole passing.
respectfully

 

[ScottyUK]

1) Probably a double earth fault. Is this a rotating rectifier machine or a slipring machine? If the former, the second GF is on the rotor or the exciter rotor; if the latter, it just might be somewhere off the rotor. This could / would contribute to 3).

2) Difficult to tell. An RSO test might give you a more accurate location of the fault(s). If you have molten copper on the winding it is going to be a rewind job anyway, so svae your money. Do your stripdown carefully and methodically to determine the number and nature of failures.

3) See 1) for a hypothesis. The shaft grounding brush is a drain earth primarily designed to prevent the accumulation of static. It is not usually a power earth designed for heavy current.

4) How do you know the shaft grounding was effective when the unit is running?

5) Google 'air gap search coil' for a means of detecting rotor winding problems on a running machine. A coil is mounted in the clearance between stator and rotor approx midway along the machine axis. The coil signal is processed to give a flux measurement of the rotor poles as they pass the coil. A 'signature' from a healthy rotor is used as a baseline for comparison.

----------------------------------

I don't suffer from insanity. I enjoy it...

 

[Addy71]

Waross,

I can't megger the individual coils as that would require de-brazing all joints. Big job, client won't accept. I'm not sure how to check individual coil voltage in running, but I will try monitoring stator voltage in the future.

ScottyUK,

The machine is of rotating rectifier type. The exciter rotor has been separated & is healthy. What I wonder is that double earth faults are supposed to be highly destructive. EPRI describes severe retaining ring & shaft damage. This was not seen.

We have just placed an order for an RSO. A consultant we hired earlier stated that it could detect shorts but not their location. Please let me know if you have a procedure for this.

I don't know if the shaft grounding was effective in running, but I did check the bearing pedestal insulation after removing the rotor. I got an IR value of > 99 megohms. So, how did current pass through the bearings?

The search coil sounds interesting. We did approach GE but they refused to sell. Do you know any other suppliers?

Regards,

Aditya

 

[ScottyUK]

Hi -

Worst case for double ground fault is very severe damage, but depends on the location of the faults. Two faults on one turn will have little driving voltage and the current will be low; two faults at opposite ends of field winding has large driving voltage and more scope for damage.

Are both bearings insulated? There is a potential return path through the non-insulated bearing to ground or through the prime mover bearings via the coupling. Insulated bearing is there to prevent shaft currents, not to isolate shaft from ground.

The RSO is pretty specialised - you might speak to Doble Powertest in the UK for advice and potential assistance.

Search coils themselves are pretty simple. Rocoil in England can make coils which are suitable. You might want to discuss this with John Exon of Exon Engineering

http://www.exon1.demon.co.uk/

and make use of his consultancy service. He has forgotten more about generators than I know, and his day rates are very reasonable.


----------------------------------

I don't suffer from insanity. I enjoy it...

 

[aolalde]

I agree with ScottyUK, it takes at least two grounds in order to define a current path, unless one of the field terminals was already intentionally grounded. I think the starting of your rotor field problem was turn to turn failure, which finally developed the grounds. Your problem is that several spots on the field coils could be close to develop a similar failure pattern, for that reason a total new winding will be the only way to ensure a satisfactory long term repair. That makes the exact ground location futile. How old is the generator?

A Company that has developed (Flux-Track) rotor field monitoring based on a probe picking the slot leakage flux is Iris Power in Canada.

http://www.irispower.com

Hopefully you could connect to:

http://irispower.com/newsite/techpa...ED FLUX MONITORING FOR TURBINE GENERATOR….pdf

 

[Addy71]

Aolalde,

I think the generator problem started with loose overhang spacers under the retaining rings. We had overhauled this generator twice, in 2002 & 2005.

Both times, inspection showed these to be loose, but client did not allow ring removal. Instead, epoxy was used to glue them back in place. What bothers me is that a RSO test in June 2005 did not show any inter-turn shorts. The generator is about 10 years old.

The retainiing rings have been removed now. The winding has shifted axially, also plenty of melted copper & damaged spacers. A full rewind is under way. I will try to post some pictures soon.

Regards,

Aditya

 

[motorspert]

Aditya
Was the RSO test performed stationary or with the machine running? The advantage of the flux probe is that it shows the situation in the running condition.

Another company who do flux probes are

http://www.generatortech.com/start.html

these guys are always very helpful

The issue with all monitoring and inspections is if the client ignores the results then it makes it all pointless. But looks like in thios case its has been expensive for your client!
Good luck with the rewind