An online flow pattern identification system for gas-oil two-phase flow using electrical capacitance tomography

An online and non-invasive gas-oil two-phase flow pattern identification system based on electrical capacitance tomography technique is proposed. Image processing is used to analyze the reconstructed images. Fuzzy pattern recognition is introduced to determine the flow patterns. The experimental results show that the identification accuracy of homogeneous flow, stratified flow, annular-flow and slug flow is 94.8%, 96.4%, 85.2% and 92.0% respectively and the online flow patterns identification system is effective.