Gas Holdup Measurement Of Oil-Gas-Water Three-Phase Flow Based On Ultrasonic Dual Mode Sensor

Oil-gas-water multiphase flow generally exists in the process of oil and gas exploitation.It is of great significance to carry out the measurement and research on the flow parameters of the produced liquid profile in oil Wells for optimizing the development of oil fields and improving the oil recovery.Because there are two different physical properties of the dispersed phase,complex interface effect and relative velocity,the interface between phases in time and space of randomly changes,which makes the oil-gas-water three phase flow a complex industrial process.Different flow conditions will form different flow structure that will seriously affect the flow characteristics and mass transfer,heat transfer performance of oil-gas-water three phase flow.Flow parameters measurement of oil-gas-water three phase often depends on the understanding of the flow structure.It is difficult to give mathematical description of the highly random,irregular and unstable flow structure,which brings great challenges to the flow parameters measurement of oil-gas-water three phase flow.The innovated results of this study are stated as follows:1.A distributed ultrasonic method for measuring liquid film thickness in oil-gaswater three-phase slug flow is proposed.In this method,three twin-crystal ultrasonic probes with an angle of 120°are embedded in the same section of the small pipe diameter measurement pipe.By measuring the ultrasonic pulse transit time,the fluctuation of liquid film thickness at different positions could be obtained in real time.The nonlinear characteristics of liquid film thickness fluctuation are analyzed using the rescale range permutation entropy.This study provides a novel strategy for the measurement of liquid film structure in multiphase flow.2.A physical model of liquid film thickness in slug flow is proposed.Firstly,according to the flow continuity equation and Bernoulli equation,the existed liquid film thickness model is improved under the stagnant flow,which effectively improves the prediction accuracy of liquid film thickness.In addition,a new physical model for the liquid film thickness in slug flow is established by means of infinitesimal analysis of liquid film and continuity equation under co-current flow.The validity of the model is verified by dynamic experiments.3.A method for measuring the gas holdup of oil-gas-water three-phase slug flow is introduced.Firstly,a prediction model of gas holdup by ultrasonic method is established under different gas bubble sizes.The pulse transmission ultrasonic sensor is used to measure gas holdup in the liquid slug by analyzing energy attenuation of ultrasonic pulse after passing through the liquid slug.The liquid film thickness around Taylor bubble is measured by distributed reflection ultrasonic sensor,and the gas holdup of the gas slug in the slug flow is predicted.Finally,the complete gas holdup prediction results of the slug flow are obtained through the sensor information fusion architecture.The oil-gas-water three-phase slug flow gas holdup measurement system based on the ultrasonic sensor is constructed.4.A deep learning network for predicting gas holdup of oil-gas-water three-phase flow is designed.The network first extracts high-dimensional short-time domain features through deep convolution module,then uses feature attention mechanism to assign weights to different features,and utilizes Long Short-Term Memory network to learn weighted features extracted from the front end.Finally,the gas holdup of oil-gaswater three-phase flow is forecasted.This study realizes the accurate prediction of multiphase flow parameters based on soft measurement.5.A new nonlinear dynamic analysis method for oil-gas-water three-phase flow is proposed based on complex network.Firstly,the vector visibility graph method is proposed inferring complex network from multivariate data and successfully characterize the nonlinear characteristics in the flow pattern evolution of oil-gas-water three-phase flow.Then,the limited penetrable visibility-graph motif entropy is introduced to analyze the ultrasonic sensor signals of the oil-gas-water three-phase flow,which effectively reveals the multi-scale structural instability of the oil-gas-water threephase flow.