Prediction Model for a Gas-Water Two-Phase Flow Pressure Drop Based on Pattern Classification

Abstract

Due to the variability of flow pattern of gas-liquid two-phase flow and complexity of flow mechanism, it is very difficult to seek a single model which is able to predict pressure drop and fit for any flow conditions. Currently, the conventional models of two-phase flow pressure drop which are widely used in engineering practice are mainly based on the experimental data of circular pipe flow, so their applicable conditions are limited, mostly close to oil well producing conditions. As far as producing gas well with formation water is concerned, gas-water ratio is usually much higher than that of producing oil well; the physical property of water in some aspects is different from that of oil and velocity slip between gas phase and liquid phase is serious when the existing model of two-phase flow pressure drop is used to predict pressure drop of the conditions of producing gas well, a large error occurs Therefore, it is necessary, based on the experiment data of gas-water two-phase flow, to research the flow mechanism and discover the regular existing in the process of fluid property changing. On the basis of the current two-phase flow pressure drop model, it is necessary and important to explore modified pressure loss model applicable for producing gas well with water, to improve predictability of the pressure drop of gas wells, and to provide the theory and technology guidance for development of gas reservoir with water.