Statistical features of the flow evolution in horizontal liquid-gas slug flow

• Published in: Experimental thermal and fluid science Vol. 119; p. 110203
• Main Authors: Rodrigues, Rômulo L.P., Cozin, Cristiane, Naidek, Bruna P., Marcelino Neto, Moises A., da Silva, Marco J., Morales, Rigoberto E.M.
• Format: Journal Article
• Language: English
• Published: Philadelphia Elsevier Inc 01-11-2020; Elsevier Science Ltd
• Subjects: Experimental characterization; Flow pattern; Flow rates; Flow velocity; Frequency distribution; Information processing; Maximum likelihood estimates; Petroleum industry; Pipes; Probability density functions; Signal processing; Slug flow; Statistical analysis; Translation; Two-phase flow; Weibull distribution; Experimental characterization; Two-phase flow; Slug flow
• ISSN: 0894-1777; 1879-2286
• DOI: 10.1016/j.expthermflusci.2020.110203

•Experimental investigation on the slug flow evolution in horizontal pipes.•Resistive sensors are used to obtain slug flow parameters.•Effect of the superficial velocities in the PDF shapes of the slug parameters.•Distribution types were defined based on maximum likelihood estimate (MLE).
The occurrence of the slug flow pattern is frequently detected in the petroleum industry. Theoretical and experimental works on the complex behaviour of this flow pattern are abundantly found, and their aim is to advance the state-of-the-technique on the behaviour of this kind of flow. The goal of the present work is to provide experimental data collected in five probing stations with resistive sensors mounted on a horizontal air–water slug flow in a 25.8-mm ID pipe. A signal processing technique provides the translational velocities, frequencies and void fractions that arise from different gas and liquid volumetric flow rates. Based on the obtained results, the effects of the liquid and gas superficial velocities were correlated with the probability density functions (PDFs) of the parameters variations observed in the slug flow development along the pipe. A quantitative criteria based on maximum likelihood estimates is used to determine the Weibull distribution for the translational velocity and the lognormal distribution type for the slug frequency distribution type.