Application of Imaging Algorithms for Gas-Water Two-Phase Array Fiber Holdup Meters in Horizontal Wells

The flow dynamics of low-yield horizontal wells demonstrate considerable complexity and unpredictability, chiefly attributable to the structural attributes of the wellbore and the interplay of gas-water two-phase flow. In horizontal wellbores, precisely predicting flow patterns using conventional approaches is often problematic. Consequently, accurate monitoring and analysis of water holdup in gas-water two-phase flows are essential. This study performs a gas-water two-phase flow simulation experiment under diverse total flow and water cut conditions, utilizing air and tap water to represent downhole gas and formation water, respectively. The experiment relies on the measurement principles of an array fiber holdup meter (GAT) and the response characteristics of the sensors. In the experiment, GAT was utilized for real-time water holdup measurement, and the acquired sensor data were analyzed using three interpolation algorithms: simple linear interpolation, inverse distance weighted interpolation, and Gaussian radial basis function interpolation. The results were subsequently post-processed and visualized with 2020 version MATLAB software, generating two-dimensional representations of water holdup in the wellbore. The study findings demonstrate that, at total flow of 300 m³/d and 500 m³/d, the simple linear interpolation approach yields superior accuracy in water holdup calculations, with imaging outcomes closely aligning with the actual gas-water flow patterns and the authentic gas-water distribution. As total flow and water cut increase, the gas-water two-phase flow progressively shifts from stratified smooth flow to stratified wavy flow. In this paper, the Gaussian radial basis function and inverse distance weighted interpolation algorithms exhibit superior accuracy in water holdup calculations, effectively representing the fluctuating features of the gas-water interface and yielding imaging outcomes that align more closely with experimentally observed gas-water flow patterns.