Prediction of slug length distribution in horizontal large-diameter gas/liquid pipeline systems

Abstract

Large-diameter (ID ≥ 0.1524 m) pipeline network systems are widely used in the food industry for grain transportation, in the chemical and nuclear industries for steam/water thermal distribution systems, in the oil and gas industry for crude oil transportation and exploitation, and in urban water supply and drainage systems. Due to the favorable operational and geometrical conditions in surface gas/liquid pipelines, slug flow often prevails, dominating the horizontal and slightly inclined flow pattern map. Formed liquid slugs at large-diameter pipeline entrances and dip can grow, accelerate, and overtake other slugs, resulting in high-momentum and long slugs ranging in length from a few meters to a few hundred meters. The discrepancy of current mechanistic models to predict slug length in large diameter pipes, and the scarcity of large diameter slug length data motivated this work to propose a combined mechanistic and empirical approach to predict full slug length distribution in large diameter surface pipeline systems. A first-principle mechanistic mean slug length model is improved by optimizing its closure relationships using a large-diameter lab and field slug length database. Furthermore, an empirical large-diameter slug length standard deviation dimensionless model is proposed and statistically validated. The proposed mean and standard deviation slug length models are combined to construct a probability distribution and maximum slug length models. Validation of the proposed model revealed an absolute average error of 27% and 31% for the mean and maximum slug lengths, respectively, vastly outperforming all existing models.