Pressure drop of slug flow in horizontal pipes with different pipe diameters and liquid viscosities

Accurate prediction of the horizontal slug pressure drop is valuable for optimizing the multiphase pipeline transport process and ensuring the stable operation of the pipeline systems and treatment facilities. Experimental slug pressure drop research is conducted based on differential pressure signals in the short (0.065 m ID, 215D) and long (0.046 m ID, 35 957D) horizontal pipes. A slug pressure drop database in horizontal pipes is constructed over a wider range of pipe diameters (0.030–0.3 m), liquid viscosities (0.0101–6.25 Pa⋅s), and pipe lengths (6–1657 m). The slug pressure drop increases with decreasing pipe diameter, while the distribution range of slug pressure drop vs the superficial gas velocity rises with increasing liquid viscosity. The 26 pressure drop prediction models are evaluated for their ability in the horizontal slug pressure drop database with different pipe structures and fluid properties, and the best pressure drop correlations are selected for the low, medium, and high liquid viscosity ranges. Accordingly, a combined pressure drop prediction approach of horizontal slug flow based on liquid viscosity classification is proposed with a Pearson's correlation coefficient of 0.987, and the average absolute percentage error is reduced from 44.98% to 30.14%. Based on the Lockhart–Martinelli method, a prediction correlation for horizontal slug pressure drop is developed over a wider range of pipe diameters and liquid viscosities, and industrial near-horizontal pressure drop data are successfully predicted with 89.6% of the data in the error interval of ±20%.