Jet fire assessment in complex environments using computational fluid dynamics

Abstract

Fires are one of the major hazards threatening the safety of equipment and humans working in process units. Jet fires are one of the most dangerous types of fires. By estimating the size and geometry of their flames, it is possible to determine the appropriate area for the placement of equipment, buildings and fuel tanks. For this reason, the ability to predict the shape and dimensions of a jet fire is of particular importance. Unfortunately, the existing models and software in this field (such as Phast) have drawbacks such as the lack of consideration of barriers and equipment that make them unreliable. One method that can be used to predict the length, temperature, and thermal radiation of a jet fire and also consider its geometry is computational fluid dynamics (CFD). In this study, a vertical propane jet fire is simulated using the different models of turbulence and radiation, and the results are compared with the empirical results from the literature. The obtained results show that the discrete transfer models, Montecarlo and P-1, are more capable of simulating the propagation of heat radiation, while the Rosseland radiation model is not suitable for this purpose. The SST, BSL and BSL RS models are more suitable for simulating the turbulence and separation phenomenon in propane jet fires, while the k-ε and RNG k-ε models are not appropriate. Considering the corresponding relative errors of the different models, the discrete transfer model and the SST model are proposed for use in future works. The effect of geometry and equipment on the path of the flame were investigated while mainly focusing on the flame area (zoomed in), thereby reducing the scope of the work domain. The results show that the presence of a firewall near the flame or in the direction of the flame can control and significantly reduce the production of radiation.