Coupling of integral methods and CFD for modeling complex industrial accidents

Abstract Safety enhancement of operations in the chemical and petrochemical industry requires for advances in the tools aimed at supporting risk estimation and evaluation. In conventional risk studies, consequence assessment is carried out through simplified tools and conservative assumptions, often resulting in overestimation of accident severity and worst-case scenarios. Computational Fluid Dynamics (CFD) may overcome the limitation of simplified approaches supporting the study of the dynamic evolution of accidental scenarios and, eventually, the consequences analysis of major accidents. However, the complexity of the problem makes the simulations too computationally demanding; hence an interesting approach is to couple simplified tools based on integral models and CFD. This work is aimed at modeling a safety critical scenario, i.e. domino effect triggered by fire. An integral model is adopted to reproduce a large-scale pool fire, thus simulating the radiative heat received by an exposed pressurized vessel. The behavior of the latter is then modeled through CFD, to investigate the heat-up process and the consequent pressure build up. Potential benefits and limitations of coupling distributed and integral models to support consequence assessment studies are discussed