I can´t imagine that an electrode covered in oil would be able to detect conductivity of water, which implies it would need to detect subtle changes in impedance or more exactly in capacitance.
What are the ratios of water to oil that you are interested in? I would imagine detecting 1 part in 100 would be tricky, and in any use of oil I can imagine such number would be an awful lot and render the oil useles.
I have done some basic experiments and tests on capacitive oil/water probes.
From memory oil has a dielectric constant of about 2, and water has a dielectric constant of about 60.
Since they explain their system's operation as: "7... All we can say is that it's - magic.... ", there's no way to tell. They do say it's "non-conductive". It's obviously self-powered.
The capacitive water/oil sensors use the properties of the oil/water fluid mix to alter the frequency of an oscillator. This is a fairly mature sensing scheme. And obviously designed for very harsh environments, i.e., not potting soil.
You may have already done this, but I did a google on "capacitance as water concentration in oil" and got a good number of relevant hits.
A soil moisture detector measures the resistance of the salty water in soil. When the soil is dry then there is no salty water so the resistance is much higher.
Condensation of water in the air that is in oil probably is pure water with no salts so its resistance would be near infinity, an insulator. Therefore a soil moisture detector will not detect it.
A soil moisture detector measures the resistance of the salty water in soil. When the soil is dry then there is no salty water so the resistance is much higher.
Condensation of water in the air that is in oil probably is pure water with no salts so its resistance would be near infinity, an insulator. Therefore a soil moisture detector will not detect it.
audioguru / strantor
The dielectric constant of oil changes when water is added (very slightly solvable but oil/water emulsion will also change the dielectric constant as well. Here us one example of how it can be detected with a pair of plates - the changing properties of the fluid changes the capacitance of the plates.
In tankers, to detect and measure water in oil (not as an emulsion but - we hoped - consistently settled in the bottom), we used a device that looked like a common sounding tape with a somewhat bulky "bob" at the end.
Submerged in the tank, it sounded a beeper when reached the water interface. Never had the time to look into the details. Maybe it was in line with Jim's design. With the vessel in calm water, repeatability was very good.
I don't need awesome accuracy or resolution. Ultimately I don't need to measure the content of water, just detect if water is present above a threshold value. Arbitrarily I pick 10%; if my volume of oil consists of 10% or more water, then output a digital high for an idiot light.
I have done some basic experiments and tests on capacitive oil/water probes.
From memory oil has a dielectric constant of about 2, and water has a dielectric constant of about 60.
Here is a capacity (and inductance.) meter design with a built in display that work quite well (I have built one.) It would suit the range of capacitance values you are interested in. It is the same as the design that was in the EPE magazine March 2010
10% water in oil seems like a fairly high threshold. At that point, I suspect those methods based on physical displacement of the oil would be preferred. You could also look at light scatter for physically suspended water. At lower concentrations, you might want to consider IR. In the near IR, water shows a strong absorption around 1500 nm:
Surprisingly, the publication citation isn't shown on the Google link. To get it, you have to join for "free." I wasn't up for that, but here is the search header from Google:
**broken link removed**
And to tease your appetite, here is another figure showing results:
That apparently is dissolved water. The same IR source could be used for measuring light scatter too.
It might and is an easy test to do. While scatter and turbidity are similar in some ways, scatter offers some different dimensions. I assumed that water droplets would have different reflective properties than suspended solids, like dirt. Thus an off center detector, or even a forward detector might provide differentiation between dirt and water or other immiscible liquid.
Idea #2: can you add a small amount of water soluble dye?
What kind of oil is it & what is it used for? Refined vegetable oil, mineral oil, used motor oil, crude oil, ... Lubricating oils change quite a bit over time. So do heat-transfer oils depending on service conditions. The expected changes during the life of the oil could be an issue.
Also, is there reservoir tank for the oil. A 10% volume change when water is present may allow a simple float switch to be used.
What kind of oil is it & what is it used for? Refined vegetable oil, mineral oil, used motor oil, crude oil, ... Lubricating oils change quite a bit over time. So do heat-transfer oils depending on service conditions. The expected changes during the life of the oil could be an issue.
Also, is there reservoir tank for the oil. A 10% volume change when water is present may allow a simple float switch to be used.