Fluid Interface Detection and Level Control in 3 Phase Oil Water Gas Separators


Thread Starter


We have been using electric level controllers that utilize a SPDT dry contact electric switch, with limited success. The switch is activated by a displacer at the fluid interface. We believe the switch itself is not adequate for our application.

We have recently begun testing another displacer-type electric level controller that utilizes NAMUR sensors. So far this has been effective, but we are worried about the impact of buildup in the nipple area of the displacer arm.

Our fluid properties are as follows:
- Produced water: 1 < SG < 1.2
- Oil: mean SG is 0.75
- Gas: Primarily methane (> 85%)
- We have considerable iron sulfide buildup along with sand production

Does anyone have experience with electric level controllers in a similar environment? Does anyone have experience with Guided Wave Radar or some other type of level measurement device that does not have any moving parts?

All general advice and input is also much appreciated.

Thank you in advance.
Hello John,

3-phase separator interface measurement is one of the difficult applications for interface level measurement. In my experience there is no reliable instrument in the industry except nucleonic type level measuring instrument at present. However, nucleonic type for this application is very costly instrument.

Typical technologies used for this application are Capacitance, Guided Wave Radar (GWR) and Magnetostrictive level instruments. But for successful application of these instruments many things to consider.

Main problem in this applications are emulsion, material build-up, dielectric property change of the fluid and density change.

If there is density variation of the fluid and material build-up, displacer will not work. Dielectric constant change will badly affect capacitance and also capacitance needs careful calibration at field to take care of the tank geometry and actual capacitance variation after installation. Material build-up also affects capacitance. GWR and magnetostrictive works well if there is thin emulsion layer. Thick emulsion layer affects both the technologies. Many such points and installation factors to be considered for a successful application.


Yes, Guided wave radar is best possible way to measure the level without any deposition along it. We are using for Ammonia tank and its highly reliable. For more than ten years not a single time brought out from tank and highly reliable, accurate. You can go for it.

We have considerable iron sulfide buildup along with sand ,this type of problem are often seen in displacer type measurement and leads to malfunction.
Do you know how emulsion layers effect the ability of GWR to detect interface? Also, I don't think the dielectric constants of our methane gas and 60 API oil (condensate) are sufficiently different to detect and interface with GWR. Thoughts?
>Our fluid properties are as follows:
>- Produced water: 1 < SG < 1.2
>- Oil: mean SG is 0.75
>- Gas: Primarily methane (> 85%)

The Oil Water interface is dead easy.

A toroidal conductivity sensor at the interface point will tell if the interface is high or low, and you get a short intermediate range equivalent to the hole size. Put the probe at the desired interface point and control the level using differential gap technique. It will also read the emulsion layer if you have one.
In an emulsion layer the toroidal probe will measure from the conductive layer (highest conductance) falling off up through the emulsion as the drops of oil surrounded by water (conductive) changes to become drops of water surrounded by oil (non conductive). At about half way through the emulsion the conductivity suddenly drops to zero.

In a non emulsion the conductivity is a function of the portion of the hole that is filled with the conductive fluid.