Analysis of gas dispersion and ventilation within a comprehensive CAD model of an offshore platform via computational fluid dynamics

Turbulence Model Selection for Heavy Gases Dispersion Modeling in Topographically Complex Area

“…This model has been used by many authors for the application of atmospheric dispersion. (Sklavounos & Rigas 2004;Tominaga & Guo-cheng 2012;Gavelli et al 2008;Kashi et al 2015a) It is a two-equation model from the Reynolds-averaged Navier-Stokes (RANS) group, where the first transported variable is turbulent kinetic energy, k, and the second transported variable is turbulent dissipation, ε. The ε equation governs the scale of the turbulence, and the k equation describes the energy in the turbulence.…”

“…The robustness of the mesh has been studied intensely for the same experiment in our previous publication (Kashi et al 2015a). For discretization purpose, second-order scheme is used.…”

“…To demonstrate these variations, different Schmidt values used by different authors are listed in Table 5. (Gousseau et al, 2011) 0.40 (Nakibouglu et al, 2009 0.63 (Lien et al, 2006) 0.70 (Li and Stathopoulos, 1997;Vardoulakis and Cai, 2011;Mokhtarzadeh-dehghan, Akcayoglu and Robins, 2012) 0.72 (Giannissi, Venetsanos and Markatos, 2015) 0.90 (Delaunay, 1996;Kim and Baik, 2003;Santiago, Martilli and Martilli, 2007;Baik, Park and Kim, 2009) 1.00 (Antonioni et al 2012;Kashi et al 2015a) 0.7-1.4 (Yang and Zhang, 2017) In this study, a sensitivity analysis has been done to find the most suitable Schmidt number. In order to do so, a gas dispersion case from the Kit-Fox experiences has been evaluated through the variation of the Schmidt number using the RNG k-ε turbulence model (the reason for selecting the RNG k-ε model is explained later in this paper).…”

“…However, in recent years, the use of the CFD approach in modeling gas dispersion scenarios has been more common as computers are becoming faster and more capable. Meanwhile, CFD codes provide more accurate results in comparison to traditional gas dispersion models (Tominaga & Stathopoulos 2007;Kashi et al 2015a;Gavelli et al 2008;Giannissi et al 2015).…”

“…It has been stated in various research papers (Hu et al, 2011;Tominaga and Guo-cheng, 2012) that gas dispersion simulations in a complex environment are similarly feasible and very effective with ANSYS CFX® codes. Furthermore, the relevant software is crucial to define the problem correctly and precisely at the beginning of the simulation along with the use of accurate CAD models, sufficient mesh quality, setting proper boundary layers, choosing the best turbulence model, and setting correct parameters for achieving more accurate results (Amorim et al 2013;Katz & Sankaran 2011;Tominaga & Stathopoulos 2013;Kashi et al 2015a;Hoi et al 2006).…”