Three-phase gas-oil-water flow in undulating pipeline

Abstract

Three-phase gas-oil-water (G-O-W) is common in wells and pipelines producing from mature reservoirs, imposing serious flow assurance problems. Therefore, prediction of flow behavior and characteristics is critical for optimal and sustainable production. In this study, experimental three-phase G-O-W slug flow in undulating pipelines were acquired and analyzed with respect to average liquid holdup, phase distributions and slug characteristics. The slug characteristics considered in this study include, translational velocity, slug length and frequency along the system. The objective of this study is to experimentally investigate the slug flow characteristics to improve the current understanding of G-O-W flow behavior in undulating pipelines and the effect of water cut on slug flow characteristics. Further objective is to modify and validate an in-house two-phase flow simulator and commercial transient multiphase flow simulator OLGA™ with the experimental data. This study contributes to the improvement of the existing slug tracking models and the development of new models for the proper design and safe operation of three-phase flow in pipeline systems.Several results have been found in this study, for example, seven three-phase slug flow patterns were identified based on oil-water mixture in the upstream horizontal section of the undulating pipe section. These flow patterns were analyzed and compared with slug dissipation in the downhill section of the undulating pipe section. When these flow patterns were compared with two-phase slug dissipation behavior, no water cut effect was observed. For moderate and high flow rates, slugs with different oil-water mixing status had differences in slug frequencies and lengths. However, the evolution of liquid slug length distributions for 20% and 80% water cuts in the upstream horizontal section and upward inclined section did not show any significant dependence on water cut.Furthermore, the modified in-house two-phase model predicted the slug flow evolution along the undulated pipeline with a reasonable accuracy. OLGA™ predictions of average total liquid and water holdups, slug frequency and length data showed a fair match especially with the high flow-rate conditions. The models validation study shows that the effect of water cut on slug flow evolution in undulating pipelines was limited to the effect on dispersion characteristics.