257 
a 
Slice Cut View 
Ship” 
790m 
258 
259 Figure 4: Domain of modelling: (a) External shape of the Domain and meshing resolution, (b) Slice cut of the Domain in the ship 
260 location, vertically to the bottom boundary, (c) Zoom of the Domain’s slice cut, downstream the ship. 
261 
2.4 CFD Modelling 
262 The case studies examined in the current work have been simulated with the use of micro- 
263 scale RANS modelling, concerning gas dispersion from pollutant sources (Toja-Silva et al., 2017), 
264 using OpenFOAM code for simulation of flow and plume dispersion. OpenFOAM is an open- 
265 source CFD software, which can be easily adapted to predict gaseous pollutants dispersion in the 
266  atmosphere, by modifying its libraries (solvers, turbulence models, etc.). The solver simpleFoam 
267 is suitable for incompressible flows, like the one considered in the current study. The post- 
268 processing work is performed in ParaView, which is also open-source software. Para View can 
269 read simulation results produced by OpenFOAM and export the spatial concentration data that are 
270 needed for this study. 
271 
272 
273 
274 
275 
276 
277 
Buoyancy effects have been considered due to the plume rise phenomenon. The Boussinesq 
approximation constitutes a suitable technique to model temperature differences, regarding 
ambient and exhaust gas outlet temperatures. Thus, the solver that was selected, based on the 
previously described characteristics, is “buoyantBoussinesqSimpleFoam”, which combines both 
incompressibility and buoyancy. However, gas dispersion expressed by the convection-diffusion 
passive scalar transport equation is not included and was manually added to the solver’s script by 
means of Equation 2 (Elfverson & Lejon, 2021; Toja-Silva et al., 2017). 
12