Experimental and computational analysis of vertical jet fires of methane in normal and sub-atmospheric pressures

• Published in: Fuel (Guildford) Vol. 265; p. 116878
• Main Authors: Rengel, B., Àgueda, A., Pastor, E., Casal, J., Planas, E., Hu, L., Palacios, A.
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.04.2020; Elsevier BV
• Subjects: Atmospheric and sub-atmospheric pressures; Atmospheric pressure; Building codes; CFD fire modelling; Computer applications; Computer simulation; Fire codes; Fires; Flame geometry; Fluid flow; Industrial plants; Jet flow; Methane; Methane jet fires; Qualitative analysis; Reynolds number; Atmospheric and sub-atmospheric pressures; CFD fire modelling; Methane jet fires; Flame geometry
• ISSN: 0016-2361; 1873-7153
• DOI: 10.1016/j.fuel.2019.116878

•Methane jet fires varied as the ambient pressured decreased.•The lower the ambient pressure, the larger the lift-off distance.•The lower the ambient pressure, the shorter the radiant flame length.•Reasonable predictions of flame surfaces were usually found with FDS and FireFOAM. Accidental jet fires occurring in industrial facilities can involve severe consequences as they can trigger domino effect. The assessment of the flame-geometry descriptors of the jet can contribute to prevent flame impingement on plant equipment, hence reducing inventory loss and structural collapse. This paper reports the geometrical features of vertical methane subsonic jet flames at normal and sub-atmospheric pressures: 1.0 atm, 0.9 atm, 0.8 atm, 0.7 atm and 0.6 atm. Differences on flame shape are evaluated, and linear correlations of the main geometrical parameters of interest (i.e. lift-off distance, radiant flame length, and equivalent diameter) are defined as a function of the Reynolds number. Moreover, the predictive capabilities of FDS, FireFOAM and FLACS-Fire codes are assessed when determining the geometrical features of jet fire experiments. Based on a qualitative and a quantitative comparison between simulation results and experimental data, the main strengths and weaknesses of each code are identified. Recommendations on suitable grid sizes are delivered.