The objective of the current work is to present the development of a Particle Tracking Velocimetry (PTV) algorithm for the analysis of oil drops behaviour in two-phase oil–water dispersions within a centrifugal pump impeller. The drop tracking was realized through high-speed camera images in a transparent pump prototype, which enabled the visualization of oil drops dispersed in water in all the impeller channels. The PTV algorithm is based on deep-learning techniques for image processing. The drops are detected by a combined U-Net and Convolutional Neural Network (CNN) method, with the former generating a binary image and the latter detecting valid oil drop contours. After detection, the Labelled Object Velocimetry (LOV) is adopted to calculate the instantaneous oil drop velocity. A synthetic image generator based on a Generative Adversarial Network (GAN) is then developed to assess the results from the U-Net, CNN, and LOV models. Additional validation studies are performed using the results from Perissinotto et al. (2019a). The results reveal that the presented deep-learning PTV algorithm is robust and provides consistent and reliable data for the dispersed oil phase in two-phase oil–water flows.
Tag: Oil drop
Experimental Investigation of Oil Drops Behavior in Dispersed Oil-Water Two-Phase Flow within a Centrifugal Pump Impeller
In oil production, one important artificial lift method involves the commonly used centrifugal pump. The use of this pump in the petroleum industry, however, is hindered by some unfavorable operational conditions. Operating centrifugal pumps with gas and viscous fluids, such as dispersions, may lead to a degradation of their performance. The objective of this paper is to analyze oil-water dispersions in a pump impeller, in order to investigate the behavior of oil drops, which may influence the pump working. Thus, experiments were carried out at different pump rotation speeds and water flow rates. Researchers used a facility with a pump prototype that enabled them to visualize the flow in all the impeller channels. Images, captured through a high-speed camera, revealed a unique flow pattern of oil drops dispersed in water. Processed with computer codes, the images indicated that the oil drops were, in general, spherical or elliptical, and only a few broke up in the impeller. The interaction with water caused the oil drops to rotate, deform, and deviate, thus moving in random paths. Size distributions suggested that the drops became smaller as the impeller rotation speed and water flow rate increased. This behavior was due to the turbulence-induced shear stress and kinetic energy. The oil drops’ equivalent diameters ranged from 0.1 to 6.0 mm; velocities took values measurable by units of m/s; accelerations reached hundreds of m/s2; and forces had magnitudes of thousandths of N. Researchers observed a clear dependence between flow conditions and drop dynamics. Carried by the water flow, the oil drops on the suction blade moved faster than those on the pressure blade of a channel. The drop dynamics were significantly influenced by the presence of adverse pressure gradients and water velocity fields.