AC/DC equivalent power delivery question

If your rectification is full wave, the power will be essentially the same. Of course, there will be some loss in the diode forward voltage, but in the context of "HV", this should be minute.

You don't say how high your HV is. The devil is in the details on this one. I can assure you there is virtually no current flowing *through* the glass. At high voltages, current can flow along the surface, through contaminants on the glass. Unless you have extremely high voltages, I would not expect any current to cross 2 feet of air.

Look for surface contamination. Even a very small amount can lead to measurable current flow. That's why power insulators are ribbed, to increase the surface distance and reduce the current.

What is the principle of your oil and water separation? Don't they separate naturally? Or do you have an emulsion that does not separate? What does the current do that separates them?

A couple of tests were done at 9000V, the minimum was 850V.

The devil is in the details on this one. I can assure you there is virtually no current flowing *through* the glass. At high voltages, current can flow along the surface, through contaminants on the glass. Unless you have extremely high voltages, I would not expect any current to cross 2 feet of air.

Well, that is what I need to know. The process is supposed to be just an electric field, so I expected very low current. I put a resistor in series with the aluminum vessel and measured the voltage drop, I expect the aluminum vessel is emitting current to somewhere. Because I had a voltage drop across the resistor when the other probe was not close. But, it was a pretty rough setup, I think it is being improved today.

I'll suggest cleaning and not touching.

I'm pretty clueless as to why, but I think it causes the oil to aggregate and the bigger droplets are more likely to float to the top, or as my son has corrected me, the water is pulled to the earth by gravity. Yes it is an emulsion.

I'm surprised you are getting a result. An emulsion has droplets so tiny that they will not coalesce. But perhaps they can still be encouraged to coalesce. I don't think it is as simple as getting them to float to the top. They are so small, Brownian motion keeps them distributed. I'm not sure what keeps them from coalescing, perhaps surface tension?

It may be that the AC field moves them back and forth, causing collisions that overcome the reluctance to coalesce. I really don't remember much about emulsions. Maybe Bill has retained more. If the agitation of the AC field is causing them to coalesce, I would not expect much difference from the DC field, other than that the peak to peak field is half. If you can, try a smoothed DC field. If the effect goes away entirely, that would be a strong clue. But I realize 9000V caps are not so easy to come by.

Tell your son he is being pedantic. The water is pulled down harder than the oil is, so the oil floats to the top!

You do find interesting things to occupy yourself with.

It is not a electric field separation is not a recent, there are lots of papers on it.

The first test was with a electric field of 500V/cm, with the surprising result when compared to the control sample. The only paper I have seen stopped at 1000V/cm, but the result was still improving. We did get inconsistent results with increasing voltage, my need to revisit. One paper shows an improvement with squarewave, sinewave was not quite as good (it didn''t say 0.707% as good, but I wonder? :-) and triangle was worse. The squarewave could have been DC, but I question it. DC does work, but the paper says because of electrolytic corrosion it is out of favor. If there is no current through the glass, I didn't know how you get electrolytic corrosion, however there are systems that sink horizontal or vertical plates or mesh into the solution, so I can see corrosion in that method. Maybe they were just not complete with that idea. I have about six 1uf 2000V caps I could string in series.

6 of one 1/2 a dozen of the other :-) It happens,, no matter how it is described.

Do you mean from bean sprouts to oil/water separation, before that we troubleshot the air compressor motor and found a bad contact in the centrifugal switch, and then it was the four pole double throw switch I should never have disassembled, but I did get it back together and working. Thanks, Mikek

Only if the current is the same. Which is HIGHLY unlikely.

AT that sort of voltage you can get corona discharge off any sharp edges of metal and leakage currents in ordinary water can be large. We had to add the right combination of corrosion inhibitors to our systems when operating water cooling systems floating up at 8kV.

Backing pumps also required some clever protection too since although air at STP is a pretty good insulator soft vacuum isn't at all. Hard vacuum is an insulator again but the in between state is dodgy.

I can't see this being generally applicable as a method of oil separation since the currents flowing in normal water will be high!

Dissolved salts make water not a particularly good insulator.

I'm now going on the idea that there is no current flowing through solution* and it is just an electric field. So, it is just the electric field setup between the center electrode that is in the solution, and the electrode that is on the outside of the glass. I'm assuming the RMS voltage will be about equal with AC or DC.

• I'm pretty sure the current flow we saw was that spewing off the 8" x 1-3/4" aluminum tube to (everything else?) We are spending the weekend tightening up the wiring to eliminate as much as we can. Mikek

I'm now going on the idea that there is no current flowing through solution and it is just an electric field. So, it is just the electric field setup between the center electrode that is in the solution, and the electrode that is on the outside of the glass. But, yes plenty of corona discharge.

The quote from a paper, looks like I have it backwards, its the water that is that under goes coalescence.

"Electrostatic coalescence was invented for the petroleum-related industries, and this method uses electric fields to promote the coalescence of small water droplets in water-in-oil emulsions. In electric fields, small water droplets move towards each other or collide with each other due to the electrostatic forces, such as dipole-dipole interaction and dielectrophoretic force, acting on them. As a result, the merging and coalescence of those droplets will increase the droplet size and eventually lead to settling and separation of water from the oil phase."

To reiterate, I think there is no current through the solution, but, there are system that have both electrode in solution. Mikek

Just to make sure, your electrodes are on the *outside* of the container, not in contact with the emulsion, right? Since there is no current through the glass, there's no current through the emulsion either.

Oh! There is one electrode in the solution. Still, no current through the glass, other than by capacitive coupling. Is this 60 Hz? You might want to do a rough calculation of the capacitance of your set up. Is the outside electrode just a wire, or do you use foil around the container with a significant area? This could make a very significant capacitor!!!

The RMS will be identical, as the polarity at any given instant doesn't matter after squaring. Root-mean-squared. The order is square the curve instant by instant, take the mean over a cycle, and take the square root. You are free to pick either polarity from the square root, but the positive number is usually more meaningful.

So the glass container inside the aluminum tube? Is the spacing fairly tight?

Not sure what you mean by "tightening".

Sorry, I don't understand your sentence.

What does that mean, a DC square wave? Do you mean only one polarity, V+ and 0V?

I'm not following you. Do you mean they used TWO electrodes in the emulsion?

I have the HV set up with a fullwave bridge. I pick off the AC at the input to the bridge and the DC out of the bridge. Without any load, (other that air), The DC voltage is 91% of the AC voltage, pretty close from 500V to 9000V. I don't think it is the quality of the meter, I'm doubtful it is a crest factor problem at 60 Hz, unfiltered from a fullwave bridge. It seems to be more than diode loss, and also the 91% is from 500V to 9000V, straight line, diode loss would show more at the bottom end. Any Ideas? Mikek

------------------------------------------

** A standard DMM reads the *average* value of any wave on DC ranges. On AC ranges, it reads the equivalent rms value of *sine waves* - by scaling their average rectified value up by a factor of 1.11. So the average DC output from an ideal bridge rectifier will read 90% of AC rms voltage being supplied. ** There is no issue.

.... Phil

At it is, glass tube is inserted into a an aluminum tune which is one electrode, the other electrode is a 12 gauge wire centered in the tube. Agreed, no current in emulsion.

Yes, aluminum tube (acting as one electrode) rather large, it is a holder for centrifuge tubes. The glass tube does fit snugly in the aluminum tube. The other electrode is centered in the tube and is in contact with the solution. I'll see if I can get a measurement of the capacitance.

Yes, tight spacing.

Just shortening wires, doing what I think will eliminate corona currents.

Thanks Phil, I hope they can run the DC vs AC experiment this week. Mikek