13.8 kV GENERATOR-INSULATION EVALUATION. 3

[aolalde]

A GE Turbo-Generator 25 MW, 13,800 Volts, 60 HZ, 3600 rpm, was running satisfactory and under normal condition. We stopped it for scheduled two year routine insulation test.

Phases A and B passed satisfactory Insulation Resistance (IR) at 10 kVDC for one minute, with 5800 and 6000 Megohms. The PI was 3.51 and 3.55. A controlled step voltage up to 28 kV, with 2kV/min steps followed the same results as compared with the test taken two years ago. (Around 0.5 microamperes with 10 KV and 12.5 microamperes at 28 kV).

On April 15, 2004, for phase C, the minimum Insulation Resistance (IR) tested was 5200 Megohms after 1 minute (phase C) and the Polarization Index (PI) 3.13

Controlled Over voltage Test (step Voltage Test) was done to the stator winding and phase C showed; IR = 1300 Megohms at 1 minute and PI= 2.08. The leakage current showed deviation as compared to the curve of previous test results, the test was stopped at 22 kV and 79 microamperes leakage. On April 19 the stepped voltage test on phase C was repeated and stopped at 16 kV with 31.2 microamperes leakage, to prevent imminent insulation breakdown.

Questions:

Will this generator stay operational until a scheduled outage for rewinding?

Should we stop the generator now and proceed with rewinding?

The difference of the costs involved with the decision is quite big, rewinding now being the most expensive.

 

[electricpete]

There is some risk that if you put the machine in service it will fail. I have been through the same decision as you on 13.2kv 3500hp slow-speed motor. We had nonlinear dc step results. We also had slight increase in pd although still below alarm level. For various reasons we elected to run and schedule repair later. The motor failed within a month. Rewind was required. Rewind probably would have been required even if we took the motor out at the earlier opportunity but there would have been less disruption of plant maintenance and operation.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[aolalde]

Thanks Electricpete.

Certainlly we have the pressure since we do not have spare bars and they will take several weeks for manufacture.

How severe could be the expected damage to the Laminations or the rotor, considering that we have diferential current protection to clear the ground fault of the windings?

 

[electricpete]

I think if it is a gradual slow bow upward like I~V^2, the tradional wisdom says the problem is end turn contamination.

If it is more of a sharp threshhold type increase the traditional wisdowm says the problem is in the slot.

Ours was a slow bow (indicating end turn). The failure occurred in the slot. We came to find out there had been water getting into the motor over a period of time. It didn't show up in the meggers (apparently not wet at the time) but it apparently created tracking paths which affected the step voltage and pd tests and later led to failure.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[edison123]

An AC Hipot at 1.2 to 1.5 x rated voltage will confirm the dielectric stength of the winding if your client is willing to take the chance of winding failure now than during operation(which may result in core damage).

 

[electricpete]

Another test available to you is Doble power factor test. Some attractive features of this test:
1 – It uses normal voltages, not hi-pot.
2 – There is a large database of results available at Doble by machine voltage, cooling, and manufacturer.

There may be some limited visual inspections you can do. Terminal box area for sure (you probably have already done that). Look for oil leaking external and underneath the stator. Possibly in-place visual inspection on horizontal machine sometimes very easy to pull of covers on the end and then pull off bracket for end-turn inspection. I am sure you are very familiar with these mtoor inspections but I’m just trying to think what you can look at.

Some more meandering thoughts:

If all phases acted similarly, that would be lower concern, suggesting contamination, moisture etc.
The fact that one phase is different than the others is probably more concern of localized problem.
The fact that only one phase is acting differerent than the others also suggests the problem is not in the end-turn area… in that area for tracking between phases you would expect it to show up on 2 tests.

Thinking out loud: you could in theory try the test with all 3 phases connected together. (we normally start that way and only break it if we see a problem). I’m not sure how much that would tell you.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[jbartos]

Suggestion to aolalde (Electrical) Apr 20, 2004 marked ///\\How severe could be the expected damage to the Laminations or the rotor, considering that we have differential current protection to clear the ground fault of the windings?
///Normally, the current differential protection of the Generator, ANSI Device No. 87G will protect the generator stator for phase-phase shorts or lower impedances. A little bit of clarification is needed.

If there are larger current leakages to the ground above the system grounding protection set points, then the system grounding monitoring and protection should provide an alarm and a trip, if available.\\

 

[aolalde]

This differential protection senses the three motor winding phases in and out current. For instance the current in (T1) and out (T4) goes thru a CT and the CT output to a current relay. Normally zero current is out, in whatever winding failure event which modifies that zero current output, the relay trips out.

 

[jbartos]

Question to the previous posting: Please, are you interested in a generator protection or motor protection?

This differential protection senses the three motor winding phases
///The original posting refers to the generator.\\ in and out current. For instance the current in (T1) and out (T4) goes thru a CT and the CT output to a current relay. Normally zero current is out, in whatever winding failure event which modifies that zero current output, the relay trips out.

 

[electricpete]

On the general question of core damage during stator fault. I have seen 6 or 7 stator faults on 4kv and 13.2kv induction motors at my plant over the last few years. None has caused any significant stator core damage. I agree the protection probably makes a difference and differential can be among the most sensitive protection available.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[jbartos]

Comment on aolalde (Electrical) Apr 20, 2004 marked ///\\Questions:
Will this generator stay operational until a scheduled outage for rewinding?
///If the signature evaluation pattern and its extrapolation is used, it may be possible to predict an approximate generator service time.\\Should we stop the generator now and proceed with rewinding?
///Please, would you elaborate on:
"""The difference of the costs involved with the decision is quite big, rewinding now being the most expensive."""\\\

 

[aolalde]

Jbartos.

Sorry about the lack of concentration. This is a Generator not a motor. The protection is as described.

I agree, a catastrophic fault will be more expensive. We are betting on the protection.

 

[jbartos]

Comment: The problem with the generator insulation good age prediction is nowadays somewhat become obstruse, since thare are many nonlinear loads and resulting nonlinear currents including harmonics, spikes etc.

 

[jplinn]

I realize this is too late to be of use, but I have been in the field since May working on our generators. If the situation happens again maybe this will be useful.

The same situation has occurred several times in the last few years on the generators I manage. On all of the occasions I reviewed the winding configuration and determined the generator de-rating for operation with a cut out coil based on nameplate. (Note: This conservative as my experience is that many older generators have significant margin.)

After calculating the decrease in rating and evaluation is made as to acceptablity of the lower rating. If exceeds the turbine rating no problem, if it is too low, consider the following - liminting power factor and/or running at temperatures over rated temperature until coils can be manufactured and installed. Use insuation aging calculations and operating time to a rewind to determine the acceptability of the aging effects on the insulation.

Test the unit to a voltage level that will give you confidence that the winding will last until a planned rewind. (I uses 2.2 to 2.3 * Phase-to-phase voltage).

If it passes, but has significantly different I vs V profile, plan on a rewind next outage. If profiles match it is your decision.

If a coil is cut, check operating temperatures on all generator RTDs and check negative sequence currents on initial startup.

During this process potential rewinders should be brought in to measure up the machine.

 

[electricpete]

In aolalde's situation, how would one determine which coil to cut out?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.