Revisiting Slug Length Distribution in Large-Diameter Pipelines Available to Purchase

Modeling slug length in large-diameter pipelines is crucial for the efficient design and safe operation of petroleum production systems. This study aims to develop a mechanistic model that predicts slug length distribution, including the maximum slug length and its occurrence probability in large-diameter horizontal pipelines. A comprehensive slug length database that contains experimental and field data is collected from the current literature for large-diameter pipes (ID ≥ 6 inches). The mechanistic unit-cell model for slug flow is applied by solving mass and momentum conservation equations to predict the average slug length, which requires a set of closure relationships. Thousands of available closure relationship combinations are evaluated to minimize the prediction error against the compiled database. The best combination of closure relationships is found to be of Jepson and Taylor (1993) for slug frequency, Wong (2003) for slug liquid holdup, Lizarrage-Gracia et al. (2017) for elongated bubble drift velocity, and Bendiksen (1985) for the elongated bubble distribution coefficient. In addition, an empirical model is proposed for slug length standard deviation that is a function of superficial liquid and gas Reynolds numbers. The proposed models of average slug length and standard deviation are used to generate a full lognormal distribution, from which the maximum slug length is predicted given its occurrence probability. A validation study of the proposed slug length distribution model revealed an absolute average error of 28% and 38% for the average slug length and the maximum slug length, respectively. The proposed model vastly outperformed all existing slug length models.