Water Cut Effects on Gas-Oil-Water Slug Flow in Horizontal Pipes

Abstract Gas-oil-water three-phase flow is a common occurrence in the oil and gas industry, particularly in mature fields where water production increases and gas evolves as pressure declines. However, research on three-phase flow remains limited. This study presents an experimental investigation particularly into the effects of water cut on three-phase slug flow characteristics. This experimental work was conducted at Colorado School of Mines using a 45-ft-long horizontal test section and a high-speed camera to capture flow behaviors. A novel algorithm was developed to quantitatively analyze two-dimensional grayscale values from these recordings, offering new insights into slug flow dynamics and enabling the precise quantification of slug frequency, translational velocity, slug body length, slug unit length, and the length of the liquid film region. Key findings illuminate the intricate influence of water cut on these parameters. Slug frequency exhibited an evident variation, initially decreasing and then increasing as water cut rose from zero to one. Slug translational velocity was observed to initially rise before decreasing at higher water cuts, with experimental data aligning well with classic predictive models. Slug body length, liquid film region length, and slug unit length generally showed an initial increase followed by a decrease with respect to water cut. The pressure drop in three-phase slug flow demonstrated a complex dependence on water cut, decreasing for water cuts between 0% and 60% due to the higher volumetric fraction of lower-viscosity water and decreasing slug frequency. However, at 80% water cut, the slug body transitioned to a dispersed oil-in-water pattern, leading to a rise in pressure drop suspected to result from elevated bulk viscosity and increased slug frequency. This study advances the fundamental understanding of gas-oil-water slug flow by providing a discussion on how water cut influences slug flow characteristics, offering insights into the underlying causes of pressure drop variations. The findings are highly valuable for modeling three-phase slug flow characteristic parameters, serving as key inputs for some widely used mechanistic slug flow models.