Image-based analysis of intermittent three-phase flow

During the last decades, a rising interest was earned by two-phase oil-water and three-phase oil-water-gas flows, in particular for long-distance transportation of transitional and heavy oils. A technique to reduce the pressure drop is water injection into the oil stream, aimed at creating the so-called core annular flow (CAF). In the presence also of a gaseous phase, which is very common at the exit of an oil well, bubbles perturb and deform the oil core and water annulus. A technique is proposed here to measure the gas bubble velocity by image processing of frames from videos of the flow and by an analysis in a combined space-time representation. The main advantage of this procedure is that it does not require cross-correlation between different signals. Within certain ranges of the governing parameters such slug flow can be analysed as an equivalent gas-liquid flow, considering oil and water as a single liquid. Under this assumption the measured velocity can be used to estimate the in situ gas volume fraction. The results are in very good agreement with the predictions of the drift-flux model and the Armand correlation.