Process Safety and Environmental Protection, Vol.135, 101-112, 2020
Investigation on the overpressure of methane-air mixture gas explosions in straight large-scale tunnels
To investigate the overpressure of methane-air explosions in straight large-scale tunnels, the computational fluid dynamics (CFD) code of Flame Accelerator Simulator (FLACS) was used and validated against experiments conducted at three different scales, and the effects of the volume concentration of methane in air, the blockage ratio (BR), the tunnel length, and the cross-section were studied. When analysed using the GaussAmp mathematical model, the maximum peak overpressure appears at a volume concentration of 10.30 % of methane in air. Blockage ratios (BR) of 0.15 and 0.3 resulted in the combustion of methane-air mixtures with the volume concentration of 6.5 % and 14.0 % of methane in air, producing a fatal overpressure of 21 kPa. When the BR increases up to 0.75, both the lean and rich mixtures cause a peak overpressure of over 60 kPa. Combustion of the same methane-air mixture produces the same overpressure, which decays approximately linearly at the same slope owing to a smooth wall roughness before travelling near the outlet, independent of the specific tunnel length. A method to characterise the cross-sections was proposed, and the maximum peak overpressure of different lengths of methaneair mixtures in different cross-sectional tunnels was found, presenting various regimes from a hump shape to a wave-like uplift and bowl shape. The cross-section parameters determine the degree of confinement and further control the maximum peak overpressure in the modelled tunnels. An exponential asymptotic model can be used to conveniently obtain the maximum peak overpressure. These phenomena indicate that approximately square-shaped cross-sections should be selected to avoid an extremely high overpressure in large-scale tunnels with the potentially significant accumulation of methane-air mixtures. (C) 2019 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.