Optical Method Characterizes Flow of Oil and Water

Hank Hogan

The fact that oil and water don’t mix causes problems for the petroleum industry. The production of offshore oil is accompanied by the production of free water, with the ratio of the two varying with the oil field, its maturity and other factors. In getting the oil out, both water and petroleum are pumped through pipes over long distances.

Moving oil and water down a conduit sounds easy, but the lack of mixing makes it complicated. Prasanta Kumar Das, a professor of mechanical engineering at the Indian Institute of Technology in Kharagpur, said that the liquids distribute themselves into different geometric configurations, called flow patterns or flow regimes, that affect pumping.

Oil and water don’t mix, as can be seen in these pictures of water and kerosene flowing in a pipe. Depending on the relative flow rate of the liquids, they can stratify in a smooth (a) or a wavy (b) manner. At even higher flow rates, the liquids separate into three layers (c). Courtesy of Arun Kumar Jana, Indian Institute of Technology.

“Different hydrodynamic and transport parameters like pressure drop, coefficient of heat transfer and mass transfer, reaction rate and so on are dependent on the flow regime,” he said.

In characterizing the flow phenomena, academic and industrial researchers have been hampered by the choice of methodologies available to them. Many measurement approaches for this application involve conductivity probes, which are intrusive and which can become fouled by an oil film and yield erroneous results. Better probes and circuit designs can improve conductivity-based techniques, but they always will be intrusive.

Das and his colleagues at the institute thus set out to develop their own technique, turning to optical methods. After some work, they settled on a fairly simple scheme, which they describe in the Feb. 24 online edition of Industrial & Engineering Chemistry Research.

At low flow rates, the oil is dispersed in the water as distinct droplets, resembling a flow of bubbles.

They used a 2-mW diode laser to produce a 2-mm-diameter beam of 660-nm light that they sent through a transparent pipe containing the flowing liquids. On the opposite side, they detected the light that emerged using a photodiode, counting on some basic principles of optics for their measurement.

As a light beam passes through different media, there may be reflection, refraction, absorption and scattering, Das said. “These phenomena were considered in designing the new measurement technique.”

When passing through the flowing fluids, the beam traversed two liquids arranged in bubbles, drops, waves and other structures, leading to an array of interfaces, each with its own optical impact. The ensemble of configurations was characteristic of the flow regime.

As the flow rate increases, a chaotic situation arises in which the oil and water transition between a state in which the oil is dispersed in water and in which the water is dispersed in oil (a). If the oil flow is further increased, it becomes pluglike and occupies the center of the pipe (b).

Converting the light falling onto the photodiode into a voltage and plotting it against time provided insight into the phase distribution. For a better appraisal of the flow, the group analyzed the signal using probability density function and wavelet methods.

In a demonstration of the technique, the researchers investigated the flow of kerosene and water through a pipe, varying the velocity of each from 0.05 to 1.5 m/s. They found that the optical probe produced a characteristic signal when water and kerosene flow were both low. Under those conditions, the fuel was distributed as visually observable discrete droplets in the water, and the probability density function found a single high-value peak with a large spread.

Increasing the kerosene flow eventually resulted in no visible drops and only a bluish color. The probe indicated chaotic distribution, with shifts from water in kerosene to kerosene in water and back again. The color remained the same at still higher kerosene flows, but the probe picked up what appeared to be an annular flow where the kerosene occupied the center of the pipe and the water surrounded it.

Das said that a patent has been filed on the approach and that the commercial prospects are promising for a variety of applications.

He explained that, “It would be good for use both as a research tool and as an instrument to monitor an industrial process.”

Contact: Prasanta Kumar Das, Indian Institute of Technology, Kharagpur, India; e-mail: pkd@mechiitkgp.ernet.in.