Polarisation Index and isolation on a generator 1

[mottt]

this morning I test a generator 6 kV, 3,3 Mw, this generator have been cleaned and varnished, but the values is no so good on my opinion....
Phase to earth---- with 1000 V 10 Gigaohm
Phases to earth ---with 2500 V 9 Gigaohms
Phases to earth--- with 5000 V 3 Gigaohms
Phase to phase---- with 1000 V 30 Gigaohms
Phase to phase---- with 2500 V 25 Gigaohms
Phase to phase---- with 5000 V 10 Gigaohms

The measures on the diferents phases are very close between them...
Then I test polarisation index, measuring isolation on 1 minute and isolation after 10 minutes, and them divide the ten minute value between the one minute value....
phase to earth was with 5000 V, 3 gigaohms at 1 minute and 4 gigaohms at ten, and the polarisation index was:1,33
this is a value too low, i had read that a good isolation on a winding is IP 2... or higher....
the same for the phase to phase isolation....
I hope someone of you, will know more about this concern than me....
thank in advance

 

[edison123]

How old is this generator winding ? If it is over 20 years, then the insulation would have reached its end-of-life condition. The decreasing IR values with higher megger voltages seem to indicate this. You could try heating the winding (externally by some heaters / hot-air blowers or internally by passing dc current) to about 90 deg C and check if PI (polarization index) improves. You can also try short-circuit run of the generator to dry out the both stator & rotor windings.

I am assuming your meggers are calibrated and read true values.

 

[rotman]

With insulation resistance on several GigaOhm level, PI value is not so essential as in case when IR is eg.hundreds or dozens of MOhms. With your IR values and PI=1,33 I wouldn't concern much about the insulation condition (esp. if this is not a new machine).

 

[mottt]

thank you, the machine is 10 years old, we cleaned it cause the oil from the bearing lubrication system go inside the windings, but now it´s clean....
I´d read that when a winding shows a redution in the value of the isolation with trips of voltage, means that isolation quality is going down....
Now the question is, what´s going in the winding when you got a good insulation but a bad PI?? Contamination on the winding will show a less than 1 PI, isn´t it???
ok, thank all of you for your kindy attention

 

[TheBlacksmith]

As stated, your initial insulation readings are quite high, many people would not bother to do a PI with those initial readings. (We don't unless readings are less than 500 kohms.) I suspect that the high initial readings leave little room for improvement in the values over ten minutes, resulting in the low PI readings.

 

[electricpete]

I agree - at these high levels, the P.I. doesn't mean much and I would consider the winding good.

IEEE43-2000 section 12.2.2 states "When the insulation resistance obtained after the voltage has been applied for 1 minute is higher than 5000 Megaohms, based on the magnetiud of applied direct voltage, the total measured current can be in the submiroampere range. At this level of required test instrument sensitivity, small changes in the supply voltage, ambient humidity, test connections, and other non-related components can greatly affect the total current measured during the 1-10 min interval required for a P.I. Because of these phnomena, when the one-minue insulation reading is higher than 5,000 megaohms, the P.I. may or may not be an indication of the insulation condition and is therefore not recommended as an insulation resistance tool"


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

 

[mottt]

thanks electricpete... your reply it´s a real technic answer...
it´s clear for me

 

[edison123]

IEEE std 43-200 states

"The value of insulation resistance may decrease somewhat with an increase in applied voltage; however, for insulation in good condition and in a thoroughly dry state, substantially the same insulation will be obtained for any test voltage up to the peak of the rated voltage.

A significant decrease in insulation resistance with an increase in applied voltage may an indication of insulation problems. These problems may be due to imperfection or fractions of insulation aggravated by the presence of dirt or moisture; or the problems may be due to the effects of dirt or moisture alone or result from other deterioration phenomena"

I do not agree that you cannot get a good PI with high IR values. I routinely test large generators and motors get PI over 3 even with one minute IR over 5 gigaohms.

I would still recommend drying out to get a consistent IR regardless of the applied voltage and a good PI.

 

[Zogzog]

I agree about the PI being a poor indicator at high IR levels. But nobody has asked the important question. Are these IR values temp corrected to 40 degrees C?

 

[mottt]

ok, that´s true, it´s not translated to 40 degrees, i forget it... and the values are measured at 25º C...
I knowed the point of the increase of the voltage have to give more or less the same insulation.... but i really don´t know what to think.

In other way... what is the minimum insulation you will start up a generator of 6000 V, in example 1,6 Mw??

 

[bud26]

according to some gen. maufacture requirment:the minimum insulation resistance is kv/1000+1 meggerohm. so the minimum insualtion resistance of generator of 6000v should above 7 meggerohm.

 

[electricpete]

KV+1 applied to old insulation systems, but for a modern winding, IEEE 43-2000 indicates a minimum insulation resistance of 100 megaohms (temperature corrected).

That's a pretty low value. I usually look for more than 500 megaohms.

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

 

[Zogzog]

I agree th pete on both the IEEE spec and his recommendation

 

[rmw]

I want to go back and know how you cleaned the windings and what you did to dry it afterwards before you applied the varnish. I am curious as to what you may have applied the varnish on top of.

rmw

 

[mottt]

ok, this machine was full of oil cause of the excesive flow of oil into the bearings, and cause this is a horizontal machine, the oil go inside of the machine and into the winding cause of the movement of the refrigeration air... this is an IC 21 machine.
Cause of this i dismounted the generator, and take it to the office, where i dismounted and clean the windings with water + dielectric liquid, then when all the oil was removed ( or seems to be removed) I give it heat, in a controlled way, checking all the time the temperature that was stable and the isolation....
this is how i do

 

[rmw]

One word in your post strikes fear in my heart - "water".

rmw

 

[electricpete]

Also note the word "heat", which is supposed to calm your fears about the word "water".

I have seen low PI's over and over again when the IR is in the gigaohm range. The ability to get a good PI measurement depends on your instruments and technique. Many instrumenets flatten out and just can't measure accurately that high.

Personally, I would not worry. If I was worried, I would check how long the heat was applied and possibly apply it longer, or else check power factor. But again, I personally would consider those readings you gave perfectly acceptable.

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

 

[edison123]

You need another dry-out. Try short circuit run.

 

[electricpete]

There is some subjectivity in making the decision.

But certainly doesn't hurt discuss all sides and to be conservative if uncertain.

Also could be there is some residue remaining of the oil contaminant or the cleaning fluid. Often people use a volatile fluid for cleaning which will evaporate... you said you used dielectric fluid? I'm not familiar with that.

One thing I didn't notice before is the consistent decrease in resistance as voltage increases, even at comparatively low test voltages from 2500 to 5000. That is not normal.

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

 

[mottt]

tomorrow i´ll see the reference of the dielectric fluid this is specially for cleaning electric windings....

and yes, the decrease in resistance didn´t like me. In any way at this moment i´m mounting the machine on the power plant...but still i´ll like to know more about this sucess...

and what´s the technique for get high IP´s?? normally the device do everything?? I used a megger Chauvin Arnoix