Electric motor insulation drop after cleaning

[Leendabests]

Dear Team,

I have experienced several problem regarding to the compressor motor. Our client have sent the compressor motors range between 250 HP to 450 HP. We are following standard service step by step. After dissemble the motor, we have done IR test at each phase of coil to check the healthy of the coil. Then we have cleaned the coil with high pressure water to eliminate the contaminants. After that we oven baked the stator. After the stator been taken out from oven, we do again IR Test at 1000 V. We found out that the IR dropped after it cooled down. My question is what is cause of insulation drop after clean the coil? Even after we re-varnish the coil, the IR still remain very low. Please advise how to prevent the IR from drop drastically.

Thank you

 

[flexoprinting]

Don't know what environment yours is exposed to, our compressors are in a closed ventilated room with a significant oil mist exposure which eventually softens and the insulation. Is possible the pressure washer blasted off some of your softened insulation producing a short within the coil?

 

[theode]

Sometimes when using pressure washer,the dirt is moved rather than cleaned.
We used a steam cleaner for such work as it breaks down grease easier.
Make sure the stator has thoroughly dried out including insides of slots.
Let the stator cool right down before testing,as a slightly warm stator gives a false reading.
Sometimes a second wash is needed.
Reasons for low readings,dirty slots,old insulation breaking down,damp still tracking.
Test,wash,stove to dry,varnish while hot,stove to cure,cool down,retest.

 

[Compositepro]

What is your oven bake procedure? Large massive objects can take a very long time to come to temperature and often people just do not realize this. In order to remove water from cavities the temperature of that water has to rise above 100C, and there has to be enough time for the required heat to transfer into the water, as water has a very high latent heat. 95C will not remove water from a nearly complete closed cavity. A powerful blower in the oven is also needed for a reasonable heating rate.

 

[Parchie]

What motors are these? MV? In my other life, we maintain oven temp at 75% of the motor insulation class for 6 hours and then let the motor cool before taking insulation test.

 

[electricpete]

I agree with previous comments - make sure it has been baked long enough. I think steam clean is more common than water clean and will not require a long of a bake.

Also
1 - what values are you reading?
2 - what is the P.I. before / after?
3 - are you temperature-correcting?

=====================================
(2B)+(2B)' ?

 

[electricpete]

3 - are you temperature-correcting?

If not, uncorrected readings will be lower for hot motor (recently out of oven) then for cool motor with similar insulation condition.

=====================================
(2B)+(2B)' ?

 

[krisys]

From the motor data sheet (if available any) you may check what is the heating time constant and cooling time constant of the motor. Then it becomes easy to judge how long you have to keep in the oven to heat the mass and then cool the mass uniformly.

 

[Leendabests]

Hi guys,

I would like to share one of the insulation drop cases. For this case we have done hipot test using Baker Digital Winding Tester. The result as attached. The insulation has increased but there is EAR(error area ratio)which is impulse test has reached more than 100%.

http://files.engineering.com/getfile.aspx?folder=3f6864f7-5da4-4daa-8efd-78b098997c42&file=IMG.pdf

http://files.engineering.com/getfil...4899-4f12-a21f-4c88c9f409ac&file=IMG_0001.pdf

 

[electricpete]

Before clean/bake @ 1 minute: 20 mega-ohms
After clean/bake @ 1 minute: 8,381 mega-ohms
So, this is the opposite of your original problem statement.

Your new question is related to the surge test results (completely different topic)???

Assuming that's the case, (no relation to original questioh), the "after" definitely shows a significant deviation among phase-lead-pairs worthy of investigation. Was rotor installed at the time? Check the test connections. Consider repeating test and examining behavior as slowly increase voltage on a given phase (zero crossings should not change)

=====================================
(2B)+(2B)' ?

 

[biggeorgetechnical]

In my experience, these surge results would tend to indicate leakage between the turn-turn or phase-phase insulation is now present.
(A quality DLRO would allow you to detect if it is a dead short)

Based on the info discussed in this thread, I'd suspect that contamination might have washed into some cracks in already compromised insulation (evidenced by the preliminary low meg ohm readings) and was subsequently detected by the surge test.

This underscores the reason that the surge test is accepted as a quality control test. Meg ohm will not detect turn to turn flaws or phase to phase faults (unless the phases are separated and tested piecemeal) As mentioned by Electricpete, this problem may be voltage dependent, and "disappear" at lower levels. Further cleaning may make it "disappear" as well. Flexing the winding by hand may allow a motor repair tech to determine exactly where the flaw is located...

"When you go looking for problems, you will find them... But they may not be the problems you went looking for!"