Flow Status Monitoring for Oil-Gas-Water Three-Phase Flow via Slow and Steady Feature Analysis and Empirical Mode Decomposition

As a dynamic and random complex process, oil-gas-water three-phase flow widely exists in petroleum production. The three-phase flow process has numerous flow regimes and the flow status is complex and changeable. To extract the key features of oil-gas-water three-phase flow and monitor the flow status, a method based on slow and steady feature analysis (S²FA) and empirical mode decomposition (EMD) is proposed. Empirical mode decomposition is performed on the measured data of the continuous wave ultrasonic Doppler (CWUD) sensor to get flow velocities of different phases. Four S²FA-based monitoring models are then established based on the data measured by multiple sensors. The flow status of oil-gas-water three-phase flow is monitored with concurrent monitoring of steady states and process dynamics. Two models are used to monitor the gas-liquid relationship, while another two models are used to monitor the oil-water relationship. The application to monitoring oil-gas-water three-phase flow process demonstrates the feasibility and efficacy.