Corona effect

The corona discharge you see is because the dielectric (the air) between the cables is high enough to partially cause breakdown. The corona discharge is the start of this insulation breakdown and it causes a loss of power which is used to ionise the air between the cables which results in a noxious gas called ozone to be given off, this ozone further reduces the insulation's effectiveness and so you are witnessing the beginning of the end.

Unless you stop it by moving the cables further apart, or by introducing another insulating layer between the cables.

How to avoid it is to reduce the voltage gradient between the cables to a safe level for the air to not become ionised, as above.

John.

OH, REALLY THANKS

But, one another Q is: what do you mean by "voltage gradient"? do you mean making cables parallel?

Voltage gradient is the difference in voltage between the cables, divided by distance. So many volts per metre. If the cables are at the same voltage there is no gradient..or if they are far enough apart the gradient is low...

How does the corona effect work?

Is it the case that where volt potential is brought in close proximaty divided by a dielectric, that with the increase rate of the dielectric divider, that the potential for corona effect increases?

1: i.e 12V separated by a dielectric of 6 with a gap of 100 microns compared with the same gap but a higher dielectric of say 40 (some materials have this rating and higher up to 80)?

2: does the nature of the material used as the dielectric divider contribute to conteracting the potential of corona effects?

The other name of Voltage gradient is electric field strength...

Full marks for waffle - get to the question? What you doing with a 22kV short circuit, can you be out alone at night without your dad's permission?

Hi there:

Indeed you have a corona effect discharge, this happen at about 20KV, having 1-inch equal 10KV at 22KV you should see a blueish color of about 2.2-inches. The short may be happening when you have several arc that short youre system out to ground. Please email me you application, I work in high KV systems mainly on electrostatic precipitators, we tend to use non-conductive PCV tubes on cables sytems that have high KV adding a grounding wire on the outer casing of the tube and grounded on only one side to earth. The tube needs to be about 8-inches in diameter to self contain the corona effect.

Email me youre application. fhromero at yahoo.com

youre pal frank

If you are seeing the visible blue glow of corona, then the spacing between the cables may be too close, the dielectric constant of the cable insulation system may be too high, and/or the insulation on the cables may already be damaged. Corona occurs with the electric field in air at normal pressures approaches 30 kV/cm. Once corona begins, the combination of ultraviolet light, reactive gaseous byproducts, and the discharges themselves combine to rapidly degrade most common insulating materials. This degradation can ultimately cause complete dielectric breakdown of the insulating material accompanied by a flashover (short circuit).

If the separation distance between the cables can't be changed, you might try converting to cables that use an insulation system with a low dielectric constant (k), such as cross-linked polyethylene (XPE), with k~2.3. This should cause more of the voltage stress to appear within the cable dielectric, and less voltage stress to appear within the air gap between the cables. In order to have long term system reliability, you must prevent corona from being formed in the first place.

Hey bert, the effect of a greater dielectic constant works against stopping the corona effect doesn't it? I don't know what the gap per constant ratio is, but a greater dielectric contant in comparison to a smaller dielectric constant over an equal width will allow the passing of a greater current before breakdown occurs. surely?!

Suppose you have two conductors separated by a layer of insulating material and a layer of air, each of equal thickness. The electric field across each layer will be shared unequally based up the inverse of their respective dielectric constants. Increasing the dielectric constant of the insulating material has the effect of reducing the voltage stress across the insulator, thereby increasing the the electrical field across the air gap.

It can be shown that the peak electrical field within the air gap for this system can be estimated as:

Eair = (Vp*Xair*€/(Xd+€*Xair))/Xair (electric field in air gap)

Where:

Vp = peak voltage between conductors

Eair = Peak electric field across air gap (volts/cm)

Xair = air gap thickness (cm)

Xd = Total Cable Dielectric thickness (cm) = 2X insulation thickness on each wire

€ = Dielectric constant of Cable Dielectric

Let's take an example of two conductors with 0.375 cm of insulation on each conductor (0.75 cm total) and a 0.75 cm air gap between them. Let's also assume an applied voltage 22 kV RMS (or 31.1 kV peak).

Case 1:
If the insulation is crosslinked polyethylene (€ ~ 2.3), then the electrical field seen in the air gap (Eair) will be 27.6 kV/cm. Since this is less than 30 kV/cm, you should not get corona between the cables. No problem.

Case 2:
However, suppose we change the cable dielectric to PCV which has a higher dielectric constant (€ ~ 4.9). In this case, the electrical field seen in the air gap now climbs to 34.4 kV/cm. A problem! Since this is greater than 30 kV/cm, you will now likely get corona discharge between the cables. Now, suppose we decide to increase the insulation thickness to 0.5 cm on each wire to try to eliminate the problem. Now the electrical stress jumps to 44.2 kV/cm in the air gap - we've actually worsened the problem!!

The bottom line: once you have a strong enough E-field to develop corona between the cables, merely increasing the thickness of the cable insulation will NOT eliminate the corona problem! You either need to increase the distance of the air gap, reduce the dielectric constant of the cable insulation, or remove the air gap entirely (using vacuum impregnation BETWEEN the individual cables, or substituting a multiconductor HV cable) to inhibit corona.

Hi,

what happens at very small distances?

I saw the Paschen curve that gives from distance times pressure the breakdown limit.

Below very small distance there will not be any breakdown.

This seems to be 6µm in air.

Does anyone know if only the tunnel-current is flowing if the gap is below this distance or are there other currents too?

RHABE

Answer 12, combined with Question 13 and answer 16 puts the whole thing into perspective. thanks for the great answers... and there's nothing counter intuitive about it. It makes perfect sense. corona is due to the ability of a building current across a gap, where as tunneling would be greater enhanced at bottle neck points or kinks that narrow the gap. ...My question was more in relation to capacitance and potential promblems that may result, but still taken and asked within the context of this thread...

Individual conductors in multi-conductor HV cables are now often screened to reduce the likelihood of insulation breakdown resulting from the stresses produced by the strong electric field. This solution to the breakdown problem that began to plague underground distribution systems in the 1930's was probably overlooked or resisted at first, because it seems counter-intuitive. Electricity at very high voltages can behave very strangely. It's worth the trouble to find a way to suppress the corona you are experiencing, because the ozone it produces will seriously degrade any organic insulation on the conductors, and it isn't good for your health, either. Ozone, by the way, is an unstable and very reactive oxygen molecule that contains three oxygen atoms instead of the normal two.

What effect would an exterior magnetic field have on the corona ? would it be possible to install rare earth magnets, to magnetically shield one wire from another?

GF

It is such a pleasure to read forum threads from someone who really knows what they are talking about!!! Thanks Bert Hickman !!!!!!

I'm not sure I understand the proposed multi conductor cable.

And in general Is this a 60 Hz power distribution apllication? Also would it be possible to achieve transmission without corona by using a transformer with several voltage taps to more evenly distribute the voltage gradient?, or is that infact the multi conductor cable proposition? I was also thinking several layers of floating shield? to distribute the voltage gradient?

And in addition, not that this solves anything, but out of curiousity; Are there some materials with dielectric constants well over 100? I seem to recall that Titanium and or Aluminum oxide had extremely high dielectric strength and I was just curious to have some commentary on that as far as capacitance goes.

Geoffrey Reed

Thanks for the kind words. The transformer solution sounds complex. However, your suggestion of using shielded HV cable is an excellent one, since all of the voltage stress would be confined within the insulation system of each individual cable. Shielded HV cable is made to resist the formation of internal corona through special insulation grading techniques, and the grounded shield eliminates E-fields outside of the cable. This approach is also significantly safer to personnel and equipment, particularly if this is a high power application.

I'm not familiar with the original poster's exact application. I had assumed that the application was AC since it involved slip rings. The poster said he was seeing visible corona between insulated HV wires at these (relatively) low voltages, and this would normally occur only with an AC (or high frequency) voltage between nearby wires.

Re: your question about dielectrics with high dielectric constants:
There are a number of ceramic dielectric materials (ferroelectrics) that have very high dielectric constants. Typically these are specially prepared and sintered mixtures of barium, strontium, or lead titanate. They can have static dielectric constants that approach ten thousand! For example, the Y5U dielectric material used in Vishay's HV ceramic capacitors has a dielectric constant of about 8500. A ceramic dielectric can also withstand very high voltages, and single dielectric layer ceramic "doorknob" caps are commonly available with working voltages of up to 50 kV.

In comparison, aluminum or tantalum oxide have comparatively low dielectric constants (about 7 and 11 respectively). But because these use a very thin oxide dielectric layer, electrolytic capacitors made with these materials have very high capacitance per unit volume. However, their working voltage is limited to about 50 volts (tantalum) or about 500 volts (aluminum).

8500! man, thats wacked!

is there a Farad limit on that when building HV capacitors?

Bert, I think you better check your sources for the dieclectric properties of Rutile (Titanium oxide). My old engineering pocket manual gives a dielectric rating of 40 to 80 depending on the composition of the material. Or mayhaps, I've miss read my book?

Excellent answers Bert.... its a pleasure reading them !!

John.

BERT.

Has any research been undertaken regarding the use of Fe oxide in the preparation of high voltage insulation for cabling? I could see a possible advantage regarding field effect close to the conductor.

Does the use of composite conductors such as aluminum and steel effect high voltage transmission corona? and is there any difference in the performance of these conductors, compared with differing frequencies, such as 50/60 cycles?

I have been experimenting with the development of corona for th purpose of generating ozone used in fluid sensitization. Handled properly it is far safer than chlorine, with less impact on the environment.

GF

GF:
"Has any research been undertaken regarding the use of Fe oxide in the preparation of high voltage insulation for cabling? I could see a possible advantage regarding field effect close to the conductor."

Not to my knowledge. HV cables use regions of semiconducting dielectric (doped with graphite or carbon black) to "smooth out" small defects/projections on the inner conductor and outer shield. This effectively eliminates points of E-field concentration that would otherwise initiate electrical trees and lead to eventual cable failure. Some forms of iron oxide may have sufficient electrical conductivity to perform a similar function. However, because the electrical breakdown mechanisms are considerably different within solid dielectrics, magnetic insulation effects will likely not apply.

GF:
"Does the use of composite conductors such as aluminum and steel effect high voltage transmission corona?"

Not to my knowledge. Corona formation and losses are virtually independent of conductor materials or internal construction.

GF:

"...and is there any difference in the performance of these conductors, compared with differing frequencies, such as 50/60 cycles?"

There are measurable differences in losses due to corona, core eddy current and hysteresis, and skin effect losses between 50/60 Hz. A given conductor will be lossier when used at 60 Hz versus 50 Hz.

GF:
"I have been experimenting with the development of corona for the purpose of generating ozone used in fluid sensitization. Handled properly it is far safer than chlorine, with less impact on the environment."

I'm not familiar with fluid sensitization. However, I definitely agree with your comments about the relative safety of ozone versus chlorine. Ozone is sometimes used (instead of chlorine) to disinfect water, since it can be electrically generated at the treatment plant, does not require storage of dangerous quantities, forms no halogenated hydrocarbons with organic impurities in the water, and it self-decomposes into pure oxygen.

Dear BertHickman,

thanks for the answers and the explanation of Paschen curve.

I am not understanding why at field emission the breakdown is immediate as there should be no (or may be very few) secondary electrons generated.

So the resistance in the circuit should limit the current and only at low resistance and high amps a plasma channel should develop? Am I right? As the field strength goes up at lowering the distance the current should go up? What is the characteristic of this type of electron emission - anything like an impedance or highly nonlinear and if so how to model?

I have a foto of a very peculiar corona - better to say the damage it did.

This is on the surface of a diamond that was used for cutting metals in ultraprecision machining. I assume that the chips that came off the surface did some charge separation by rubbing on the surface. The charge accumulated until breakdown and breakdown occured on the rake face of this cutting tool. There is a totally eroded part - the foot of the tree (2µm)- and a shallow(1...2µm) tree-structure 15x15µm in size.

So from this a next question arises: to which length is the light emission extended, do we see any reasonable part of it, and is there a limit for the branching .

And is this tree-like figure an averaged image that is generated by the limited time resolving capability of our eyes or is (if pulsed) this "tree" existing as one flash with all branches existing simultaneously?

Considering the suggested iron-oxide I would not think that this is really good as the magnetic hysteresis is big in these oxides and this will cause heating and promote breakdown.

Thank you for your information - your collection of books is really great - I will compile a list what I have for exchange or sale in 1 or 2 weeks and send it to you.

Thanks

RHABE

More questions: will the erratically coming radioactive decay particles start the breakdown at much smaller voltage - I assume so, that should happen in Geiger-Mueller counting tubes. What about naturally occuring or artificially generated ions in the air? - same situation?`

Hi Rhabe,

For gaps less than 4 microns, the breakdown mechanism is explosive field emission, not electron avalanche (Townsend breakdown). Explosive field emission is a rather complex runaway process that initially involves lower current field emission accompanied by the formation of gas ions within the gap which results in the buildup of a positive space charge near the cathode. The positive space charge intensifies the local E-field seen by microscopic cathode projections. Once the current density in a microprojection reaches a critical level (10E12 - 10E13 A/square meter), it suddenly explodes, liberating a shower of electrons and dense metal plasma which quickly bridges the gap, causing it to arc over almost instantaneously. Once initiated, the discharge assumes the negative resistance characteristics of a normal electrical arc.

The damage to the diamond tool sounds interesting - I'd like to see the image. It sounds like a small electrical tree.

The formation of electrical trees, whether slowly via partial discharge, or suddenly as in electron beam induced discharges, is accompanied by emission of light in the active growth zone(s). In the case of electrical trees created via successive partial discharges, these emissions are usually very low level, and not readily visible without image intensification. In the case of e-beam charge injected Lichtenberg Figures, most of the figure and light is created as a propagating discharge that removes most of the excess charge in one large, very bright discharge. There are often numerous secondary discharges that remove remaining stranded pockets of charge. Visible branching ends once the E-field (at the tips) drops below the threshold for further dielectric breakdown in the material.

Bert

Bert.

Thank you for your comprehensive reply to my questions. I miss spelled sanitize, it came out as sensitize which is rather confusing, but you never the less, confirmed the process.

I have a further question to ask, regarding High energy: Some years ago I patented a high speed wire pay out system, useful in coil winding. I live within the Lightning capital of the world : Florida.

After watching a documentary on the study of lightning, where rockets carrying wire conductors up into charged clouds, for the purpose of guiding these shafts of high energy to a place, close enough for study, (mostly photographic), It occurred to me that if such energy could be collected in this manner, then I should be able to set up my cottage industry.

They used precautions, such as firing the rockets with pneumatics rather than wires leading back to their "BUNKER".

What manner of task could be accomplished with this fleeting but colossal energy?

I thought like, John L Baird who tried to use the city's power supply to make diamonds. If I place a barrel of graphite at the launching pad, perhaps I could melt it? into a decent sized carat.

Do you think the idea has merit?

GF

GF,

Not for making carat-sized diamonds. Although there are numerous ways to briefly reach the high temperatures and pressures necessary to create discrete diamonds, diamonds created through high energy rate processes (explosive, electrical Z-pinch, or high energy impact) are microscopic in size since the appropriate conditions (2000 - 4000 K, 10 - 30 GPa) only exist for very short times. Making larger, defect-free crystals simply takes more time (hours or days) and precisely controlled and very rigorous conditions. This is true for crystal growth in general, and diamond is no exception - but the conditions are much more extreme for diamond than for most other crystalline materials.

If you have access to technical literature, here an interesting article that reviews various methods for using high current pulses to create micron-sized diamonds:

"Diamonds Produced by Electric Wire Explosion", V. E. Fortov, V. N. Korobenko, A. D. Ralhel, A. I. Savvatimski, 11th IEEE Pulsed Power Conference, 1997. Digest of Technical Papers, Volume, page(s) 226-230, ISBN 0780342135

Another interesting paper that provides a good view of the required conditions:

Nature Article (1993)

On a more "practical" note, artist Allan McColluman, using the Camp Blanding, Florida lightning research facility, did use triggered lightning to create numerous fulgurites using various mixtures of minerals:

http://home.att.net/~allanmcnyc/rakov.html

http://home.att.net/~amcnet/eventmain.html

Bert

Wasn't there a student that was trying to make an odd form of carbon found in desert sand? He had theorised that it could be caused by lightning strikes, but after several tests from this very bunker was unable to produce the same carbon structure.

I say these guy are full of crap!!!

To figure where the cross-over is occurring, just put your tongue across the + and -.

Hi, I have been involved with high some high voltage case studies of insulation failures. The cables were 500MCM, 25kV, xlpe, 133% insulation. Despite this high degree of insulation we were experiencing failures of the cables, at termination end, after six months of service. The terminations were installed new, using a high degree of cleanliness and care. A great deal of corona (high voltage) tracking was noticed, after failure, on the interior and some on the exterior. The one case I am thinking about caused about $750,000.00 damage to some indoor switch gear. Not much you say? Consider the arc only damaged two of the electrical insulators, all other damage was structural! The investigation showed that the installer failed to remove a piece of vinyl tape, used as a marker, on the semi con. This marker insulated the normal bleed of the corona back to ground through the the semi-con.

On the other hand, you are dealing with slip rings which have a high degree of carbon in them. The carbon dust will adhere to your stress cones and cause corona (tracking) to ground. A simple way to deal with this, is to, shut the machine down and clean the stress cones. Another is to install 35kV stress cones, if they will fit in the space provided. Good Luck

Could be worse, it could be me.

Mayt2U