Exposure of a liquefied gas container to an external fire

Effect of tank diameter on thermal behavior of gasoline and diesel storage tanks fires

Studies on fire behavior are extremely important as they contribute in a firefighting situation or even to avoid such hazard. Experimental studies of fire in real scale are unfeasible, implying that reduced-scale experiments must be performed, and results extrapolated to the range of interest. This research aims to experimentally study the fire behavior in tanks of 0.04m, 0.20m, 0.40m, 0.80m and 4.28m diameter, burning regular gasoline or diesel oil S-500. The following parameters were here obtained: burning rates, burning velocities, heat release rates, flame heights, and temperature distributions adjacent to the tank. Such parameters were obtained for each tank diameter with the purpose of correlating the results and understanding the relationship of each parameter for the different geometrical scale of the tanks. Asymptotic results for larger tanks were found as (regular gasoline and diesel oil S-500, respectively): burning rates 0.050kg/(m²s) and 0.031kg/(m²s), burning velocities 4.0mm/min and 2.5mm/min, heat release rates per unit area 2200kW/m² and 1500kW/m², normalized averaged flame heights (H_i/D, where H_i is the average flame height, D is the tank diameter) 0.9 and 0.8. Maximum temperatures for gasoline pools were higher than for diesel oil pools, and temperature gradients close to the tanks were also higher for the former fuel. The behavior of the maximum temperature was correlated as a function of the tank diameter, the heat release rate of each fuel and the dimensionless distance from the tank.

A new method for assessing domino effect in chemical process industry

A new methodology is presented with which the likely impact of accident in one process unit of an industry on other process units can be forecast and assessed. The methodology is based on Monte Carlo Simulation and overcomes the limitations of analytical methods, used hitherto, which were inherently limited in their ability to handle the uncertainty and the complexity associated with domino effect phenomena. The methodology has been validated and its applicability has been demonstrated with two case studies.