Why Don't Molded Splices Suffer From Partial Discharge?

Status
Not open for further replies.
B

big john

Senior Member
Location
Portland, ME
With hand-taped splices and terms, air voids and partial discharge is a huge concern: A huge amount of effort goes into excluding even very small voids or imperfections so there will be no PD.

Yet, with molded splices and terms, it would appear that you can have significant gaps (1/16-1/8" or more) between the parts of the conductor and the inside of the molded body. Yet there doesn't seem to be any concern about this at all, and everything seems to work.

Anybody shed some light on this?
 
big john said:
With hand-taped splices and terms, air voids and partial discharge is a huge concern: A huge amount of effort goes into excluding even very small voids or imperfections so there will be no PD.

Yet, with molded splices and terms, it would appear that you can have significant gaps (1/16-1/8" or more) between the parts of the conductor and the inside of the molded body. Yet there doesn't seem to be any concern about this at all, and everything seems to work.

Anybody shed some light on this?
Click to expand...

There can be PD in gaps in molded splices. PD will occur if the peak voltage is above about 300 V (have seen as low as 265Vpeak in tests) across the air bubble. If there is enough mold material distance between air bubble gaps the voltage can be kept below 300 Vpeak. e.g LOTS of molding material, thick, etc.

An FEA field map (Ansys, Maxwell, quickfield, etc) of the molding material and the air bubbles will tell the story.

My own design for HV small transformers are all vacuum potted to eliminate air bubbles.

PS: for those totally unfamiliar with PD (aka corona) one can treat an air bubble as a low capacitance in series with the capacitance of the tape or molding. Since the dielectric constant of air is 1 and other materials as high as 4 or 5, by constructing a series capacitor model the low capacitance air bubble has a higher voltage drop across it than adjacent series high capacitance sections of the insulating system.
Every time the cycle reverses the capacitance of the air gap discharges with a tiny spark, each spark is like an electrical discharge machine (EDM) cut and eventually eat thru the insulation. Air bubbles can also be the start point for treeing, which is another subject.
 
Last edited:
junkhound said:
PS: for those totally unfamiliar with PD (aka corona) one can treat an air bubble as a low capacitance in series with the capacitance of the tape or molding. Since the dielectric constant of air is 1 and other materials as high as 4 or 5, by constructing a series capacitor model the low capacitance air bubble has a higher voltage drop across it than adjacent series high capacitance sections of the insulating system.
Every time the cycle reverses the capacitance of the air gap discharges with a tiny spark, each spark is like an electrical discharge machine (EDM) cut and eventually eat thru the insulation. Air bubbles can also be the start point for treeing, which is another subject.
Click to expand...

This forum needs a like button- well said :)
 
Sorry for the confusion, yes, I'm talking about medium-voltage splices and terminations 5-35kV.
junkhound said:
There can be PD in gaps in molded splices...
Click to expand...
I guess my question is why isn't there a lot more?

There are clearly decent sized gaps inside molded components. Why doesn't that immediately create a point of significant partial discharge?

It's hard to find a good example, but in this photo all the space around the conductor lug is voids of different sizes:
hqdefault.jpg

In a hand-taped termination that would be totally unacceptable. Are hand-taped jobs just completely overkill? What gives?
 
big john said:
With hand-taped splices and terms, air voids and partial discharge is a huge concern: A huge amount of effort goes into excluding even very small voids or imperfections so there will be no PD.

Yet, with molded splices and terms, it would appear that you can have significant gaps (1/16-1/8" or more) between the parts of the conductor and the inside of the molded body. Yet there doesn't seem to be any concern about this at all, and everything seems to work.

Anybody shed some light on this?
Click to expand...

If there are air gaps between the premolded splice or termination and the cable insulation I would think it is too big. I recall pushing them onto the cable using enough force to stretch the the molded piece. Any small air gaps would be taken up by the dialectic grease that also served as a lubricant to slide on the body. I would expect air gaps of the size you mention to be a problem
 
big john said:
Sorry for the confusion, yes, I'm talking about medium-voltage splices and terminations 5-35kV. I guess my question is why isn't there a lot more?

There are clearly decent sized gaps inside molded components. Why doesn't that immediately create a point of significant partial discharge?

It's hard to find a good example, but in this photo all the space around the conductor lug is voids of different sizes:
View attachment 20375

In a hand-taped termination that would be totally unacceptable. Are hand-taped jobs just completely overkill? What gives?
Click to expand...

Yes those are large gaps but not in the insulating materials. I believe once the insulating materials are applied it is important to not have gaps within the insulation. I don’t think a gap between the conductor and the insulation is a problem.
 
The change between the high voltage and ground surface matters. The old terminations had stress cones to help taper the field to reduce the sharp corners of field where the field bent back around from the center conductor to ground.

The new slip-on splices spread that over a long sleeve and thus reduce stress. No need for stress cones and such.

Gap or no gap, if there is no stress then there is no problem.
 
big john said:
Sorry for the confusion, yes, I'm talking about medium-voltage splices and terminations 5-35kV. I guess my question is why isn't there a lot more?

There are clearly decent sized gaps inside molded components. Why doesn't that immediately create a point of significant partial discharge?

It's hard to find a good example, but in this photo all the space around the conductor lug is voids of different sizes:
Click to expand...
Looking at your photo. It may not be visible but the area around the lug is a semiconductor. The outer surface where this semicon meets the insulation is smooth/rounded to reduce stress from the electric field.

That removes corners and kinks that would normally cause a concentration of field lines at the insulation interface.
 
mivey said:
Looking at your photo. It may not be visible but the area around the lug is a semiconductor. The outer surface where this semicon meets the insulation is smooth/rounded to reduce stress from the electric field.

That removes corners and kinks that would normally cause a concentration of field lines at the insulation interface.
Click to expand...
IOW, the area around the lug/void is not just black insulation but a combination of semicon and insulating layers.
 
Status
Not open for further replies.
Top