A Differential Pressure Sensor Coupled with Conductance Sensors to Evaluate Pressure Drop Prediction Models of Gas-Water Two-Phase Flow in a Vertical Small Pipe.

Abstract

In the process of production logging to evaluate fluid flow inside pipe, logging tools that force all flow to pass through a small measuring pipe are commonly utilized for measuring mixture density. For these logging tools, studying the fluid flow phenomenon inside the small diameter pipe and improving the prediction accuracy of pressure drop are beneficial to accurately measure mixture density. In this paper, a pressure drop prediction system is designed based on a combination of an eight-electrode rotating electric field conductance sensor (REFCS), plug-in cross-correlation conductance sensor, and differential pressure sensor. This combination overcomes the limitation of the existing pressure drop prediction model that the inlet flow velocity needs to be known. An experiment is conducted in a flow loop facility with 20 mm inner diameter small pipe. The responses of the combination sensors are collected. The REFCS is used to identify flow pattern and measure water holdup. During which five flow patterns are identified by recurrence plot method, i.e., slug flow, bubble flow, churn flow, bubble-slug transitional flow, and slug-churn transitional flow. The mixture velocity of two-phase flow is determined by the plug-in conductance sensor. The differential pressure sensor provides a differential pressure fluctuation signal. Five models of prediction of pressure drop are evaluated. The mixture friction factor of gas-water two-phase flow is obtained by a fitting method based on the measured parameters and flow pattern identification using the optimal model. Then, the pressure drop can be predicted according to the measurement results of a conductance sensor and fitting relationship. The results of pressure drop prediction show that the model proposed by Ansari et al. presents a higher accuracy compared with the other four differential pressure models with the absolute average percentage deviation (AAPD) of less than 2.632%. Moreover, the accuracy of pressure drop prediction of the Zhang et al. model is improved by using the mixture friction factor.