Stream inside paraffin oil when exposed to high volage

[coquelicot]

Here is the experience: I put two electrodes inside paraffin oil (separated by at least 3cm), and submitted the electrodes to high voltage. At 20000 V - 30000 V, strong streams can be observed inside the paraffin oil. Since paraffin oil is a good dielectric, I just would like to understand the mechanism that causes these streams.

 

[sophiecentaur]

Do you mean there is visible movement of the oil? Which direction?
What material are the electrodes made of and were they (and the paraffin) scrupulously clean? It sounds as if there is something on the surface of one of the electrodes that is becoming ionised and it's then pulled across the gap, causing a 'stream'. If that were the case, I suggest that the motion of the stream would be from Anode to Cathode and that 'large' positive ions are moving (?).
What do you think and have you searched for the effect elsewhere?

 

[coquelicot]

The electrodes are made of pure copper. I was searching nothing, but simply tried to isolate an electrical high voltage filter with paraffin oil, then I observed the phenomenon. Since, I have repeated the experience with pure copper electrodes and without circuit. I don't think the movement of the paraffin oil is due to impurities inside the oil (I think it is very clean, and the streams are too strong to be due to that). The question of the direction of flow is interesting (I simply observed movements inside the oil) and I will try to make a new experiment to answer this question (not sure I can do). Anyway, I am almost certain that the streams are due to electrostatic/electrodynamic forces but I ignore the underlying mechanism. Persons used with electrodynamics of fluids may already have the solution.
EDIT: I've repeated the experiment and there is no particular direction of flow. The paraffin oil is simply agitated, much like vortexes.

 

[Asymptotic]

Where did this paraffin oil (which I take to mean a highly refined mineral oil) come from?
Was it developed specifically for use as a high voltage insulator?

Do small bubbles form on the electrodes? Does it produce any sort of sound?

I'm wondering about oil moisture content. Solubility is affected by temperature (even a water content as low as ~20 ppm condenses out at 0°C) but 30 ppm is a reasonable value for transformer oil in service. Perform a crackle test by heating a pan to 205°C, and slowly adding an oil sample drop-wise. It'll crackle if the oil contains too much moisture. The crackle test isn't a particularly good test (detection level varies from 100 ppm to 1000 ppm on up) so although it won't tell you if oil moisture level is OK, if it crackles, there's way too much water. Accurate, and lower ppm measurements require lab testing usually via the Karl Fisher titration method.

 

[coquelicot]

It is a high quality paraffin oil for food and medical purposes. It is new and clean. If it may not have been developed specifically for electrical purpose, it fulfils its aim as a very good insulator (breakdown voltage of 10000V per mm at least), as I could check. No bubbles nor any sort of sound are produced. I'm also certain that the moisture content is very low (if any).
Can someone try to repeat the experience (paraffin oil = mineral oil in US) ?

 

[Asymptotic]

In US parlance 'mineral oil' is more often used. Don't know it makes much difference in general usage, but tends to avoid confusion when describing oils that are primarily paraffinic versus naphthenic (oils that have a lower faction of heavier, waxy chains). Mineral oils are a science unto themselves as there are thousands of ways to slice and dice crude oils into their constituent parts.

For example, CAS #8012-95-1 is a common food grade mineral oil which is heavier in paraffinic components than a typical light naphthenic transformer oil such as Conoco #76 (CAS #C125-30, 64742-53-6).

I'm also certain that the moisture content is very low (if any).

Unless you have run a test on a fresh sample there is no way of knowing, and it's not amenable to gut feelings. Mineral oils absorb atmospheric moisture - not very much, but it doesn't take much to mess with dielectric fluids. For instance, unused oil in sealed, full drums has been known to fail dielectric testing if stored in a location with significant daily temperature variations for a prolonged time due to moisture getting past bung seals as the drums expand and contract.

This graph is from "Understanding Water in Transformer Systems" by Lance Lewand of Doble Engineering, and uses the ASTM D1816 test protocol with a 0.04" (1mm) gap. As you can see, it doesn't take much dissolved water to appreciably depress dielectric breakdown voltage.

Even though the focus is on their test equipment line it's worth reading "The Megger(TM) Guide to Insulating Oil Dielectric Breakdown Testing" to gain a sense of how meticulous dielectric oil testing is.

As to your original question, it isn't something I've ever encountered, but the effect you've described may be related to phenomena described in abstracts on the IEEE website. These are behind a paywall, and I haven't been a member in years so didn't look any further, but you could dig around for other cites regarding electrohydrodynamic vortices.

"Ring vortices in point-plane gaps at sub-breakdown fields"
"The role of electrostatic and hydrodynamic forces in the negative-point breakdown of liquid dielectrics"
"Electric-Field Distortions at Solid-Liquid Dielectric Interfaces"