detecting moisture in sOIL

[jpanhalt]

I looked at one of the MSDS's. New Jersey lists it as requiring mineral oil precautions. But that does not mean it is a mineral oil. It is also said to be "inherently" biodegradeable, which may mean it is not a mineral oil. Do you know off hand whether it is a mineral oil? Also, color is not described, but the name makes me wonder. Is the oil highly colored, like purple? If it is not a mineral oil, the near IR method may not work at all. We would need to check the IR properties of the oil.

The fact it is in a motor is good news. I was thinking it was in some sort of tank, but a motor will keep the oil-water mix agitated. I have a small, 5 HP gasoline-powered sprayer for the farm. You can easily tell if any seals are leaking from the water side into the oil side (mineral oil). As mentioned above, turbidity may be a simple answer.

John

 

[gophert]

strantor
What type of connections are you going to use to get sensor connections in/out of the motor? This is a common issue in the chemical insudtry when high pressure vessels are used. There was an incident at one of our suppliers where the hot, agitated, low pressure (60psi) liquid developed solvent properties that greatly exceeded the suppliers test for the ceramic potting compound used on a sensor. Once the potting compound dissolved, the liquid vapor blasted out and carried through the conduit and back into the control room.

Be careful.

 

[strantor]

strantor
What type of connections are you going to use to get sensor connections in/out of the motor? This is a common issue in the chemical insudtry when high pressure vessels are used. There was an incident at one of our suppliers where the hot, agitated, low pressure (60psi) liquid developed solvent properties that greatly exceeded the suppliers test for the ceramic potting compound used on a sensor. Once the potting compound dissolved, the liquid vapor blasted out and carried through the conduit and back into the control room.

Be careful.

The first thought that comes to mind is an NPT plug with holes drilled for the probe wires to pass thru, then filled with 3M 82-F1 potting. But I haven't devoted much thought to it or confirmed suitability of the potting. It will not be under any pressure. Well... it will not be under any differential pressure; internal oil pressure inside will be the same as water pressure outside, maybe a few PSI higher to discourage water intrusion.

I was laying in bed last night doing some reading on my phone about submersible well pumps. Those are water filled, not oil filled, but the rotor bearings are in oil filled cavities. They use water sensors based on conduction. I can't remember if they use high voltage or not, but it seems like a "megger" type insulation resistance monitor (in sensor form) would work.

 

[strantor]

So a few possibilities have emerged:

• sensing change in capacitance
• sensing change in visual properties
• sensing change in resistance
• employing dark arts

I will try to evaluate these in the order above until one of them works. I will let you guys know what I find. FYI I will not begin testing for some time, maybe a few weeks.

 

[gophert]

strantor

Not to belabor the issue, but, how do you plan to keep the pressure differential zero? Hydraulic fluids change volume with temperature and, once filled into the motor, you either have a vacuum when it cools (from not running, if filled with hot oil) or pressure when it heats up - even 10C is a big deal when the device is completely filled with oil (no head space). Paraffins/mineral oil have linear expansion of 100 ppm/degree C whereas steel is about 10 ppm /degree C). On a volume basis, that is a cube factor (30x more in volume change). You can estimate the heating of the oil by measuring the difference in current draw when the motor is run empty vs. full of oil. That is the energy pumped into the oil. The heat loss to the water is more difficult to calculate but wall thickness of the pump housing, an estimate of pump surface area and the heat transfer co-efficient of steel should give a rough estimate if the pump will gain additional heat or lose heat to the surroundings when it is 10 degrees C warmer than the water (compressibility of your hydraulic fluid will be needed to determine the pressure increase).

In other words, if the pump is filled cold, runs and warms up at any pace while underwater, you have no fear of getting mineral oil in the pump until it shuts down and cools. That is because the seals on the pump will likely leak the first time it warms up and will suck in water through those blown seals when it cools down.

 

[jpanhalt]

The pump in my hydronic heating system (Taco brand) is totally in water. No oil-filled bearings are used. Product description refers to the impeller shaft as ceramic and bearings as carbon. Water is the lubricant. It cannot pump against much of a head, but it is an example of an effectively oil-less pump that except for its electrical connections might be submerged.

John

 

[gophert]

impeller shaft as ceramic

Interesting. Is the impeller ceramic as well (or just the shaft)? Nice way to avoid cavitation-related erosion.

NOTE: Better to design around cavitation but not always possible.

 

[jpanhalt]

The pump is quite old . I have owned the home for 6 years and there was no evidence the pump had been replaced since at least 2003 (previous owner left maintenance records). Taco states that the impeller is non-metallic, which makes me believe it is an FRP or just composite rather than ceramic. They are quite expensive, so I am hoping I will not get a chance to take it apart.

John

 

[strantor]

strantor

Not to belabor the issue, but, how do you plan to keep the pressure differential zero? Hydraulic fluids change volume with temperature and, once filled into the motor, you either have a vacuum when it cools (from not running, if filled with hot oil) or pressure when it heats up - even 10C is a big deal when the device is completely filled with oil (no head space). Paraffins/mineral oil have linear expansion of 100 ppm/degree C whereas steel is about 10 ppm /degree C). On a volume basis, that is a cube factor (30x more in volume change). You can estimate the heating of the oil by measuring the difference in current draw when the motor is run empty vs. full of oil. That is the energy pumped into the oil. The heat loss to the water is more difficult to calculate but wall thickness of the pump housing, an estimate of pump surface area and the heat transfer co-efficient of steel should give a rough estimate if the pump will gain additional heat or lose heat to the surroundings when it is 10 degrees C warmer than the water (compressibility of your hydraulic fluid will be needed to determine the pressure increase).

In other words, if the pump is filled cold, runs and warms up at any pace while underwater, you have no fear of getting mineral oil in the pump until it shuts down and cools. That is because the seals on the pump will likely leak the first time it warms up and will suck in water through those blown seals when it cools down.

There will be a rubber bladder membrane over the non-business end of the motor, separating oil from water. The motor will be pumped full of oil until all possible air bubbles are purged and the membrane expands slightly, giving a slight internal pressure. As the motor descends to the bottom of the ocean (500ft max) any entrained air bubbles will collapse and oil will contract per temperature, decreasing volume slightly, the membrane will move inward. Testing will show how much membrane collapse occurs, I do not know yet. This is an experiment; I am modifying an **broken link removed** to go subsea.

Your expertise especially applicable here, John & Gopher; I've ordered 60A Nitrile/BUNA sheet, 1/16" thickness as my first choice of material for the membrane. I selected it because what I read said it has good oil and seawater resistance. Is that a good choice? Is there a more optimal material?

 

[strantor]

The pump in my hydronic heating system (Taco brand) is totally in water. No oil-filled bearings are used. Product description refers to the impeller shaft as ceramic and bearings as carbon. Water is the lubricant. It cannot pump against much of a head, but it is an example of an effectively oil-less pump that except for its electrical connections might be submerged.

John

There are a few varieties of motor for submerged applications. The one you have described does not *sound* to me like it is designed to go down to 500ft of seawater. I have not encountered this type of motor in my research of subsea motors. Maybe it could work, but I am pursuing proven methods for subsea applications first.

 

[strantor]

What do you guys (especially JimB ) think about this capacitive sensing (proximity) evaluation module for water sensing in oil (for proof of concept, not permanent installation)? I appreciate your circuit Jim, and I may build it yet, but first I want to eliminate any more plug-&-play options.

 

[tcmtech]

The pump in my hydronic heating system (Taco brand) is totally in water. No oil-filled bearings are used. Product description refers to the impeller shaft as ceramic and bearings as carbon. Water is the lubricant. It cannot pump against much of a head, but it is an example of an effectively oil-less pump that except for its electrical connections might be submerged.

John

I have a bunch of those cartridge type water circulator pumps as well. Pretty simple design really. On every one I have ever taken apart there is a thin stainless steel shell that slides in between the motor stator and the rotor (weak press fit and usually easy to pop out) that separates the two parts so the windings are actually on the outside of the water jacket not the inside.

Interesting. Is the impeller ceramic as well (or just the shaft)? Nice way to avoid cavitation-related erosion.

On all of mine they have either been some sort of fiber reinforced plastic, stainless steel or brass depending on the size and age. I don't think vacuum cavitation is a big issue being they are pretty small pumps with tiny impellers with a rather limited capacity to create any degree of suction. The biggest hammer down ones I have are around 285 watts and even then I doubt they can pull much more than 10 inches of vacuum.

 

[tcmtech]

I'm late to party but what I know of water detection is mostly from diesel equipment fuel systems and used oil burning.

On many of the diesel trucks they just use a simple two or three prong platinum or similar metal conductivity or capacitance sensor that sits down at the lowest point in a filter housing.

As far as I know they basically doing a lower voltage megger test of the fluids or a conductivity/capacitance comparison between the three electrodes.

Now as for water entering an oil obviously seawater will change the conductivity of the oil far more so that condensation water so that's one way to detect it on the go or from the bottom.

Also depending on whether the oil is hydrophobic or hydroscopic will affect things too. A good hydrophobic oil will only hold the bulk of the water in emulsion if it's being continually stirred up which although would make on the go monitoring a bit more difficult the moment the fluid mixture gets into an area of the system where there is not enough motion to keep it emulsified it will start to settle out and thus be easy to detect with a simple conductivity sensor located at a low point in the system.

Either way conductivity or capacitance based monitoring with reference to a known clean baseline value should be able to pick up water contamination fairly easily.

As I see it either off the shelf or roll your own it should not be a hard thing to work with.