CFD simulation of atmospheric dispersion

Numerical simulation of atmospheric dispersion is used to study the propagation of gases, smoke or dust, and thus to assess the impact of any type of atmospheric pollution. The approach can be applied to a restricted area, known as the near-field, or to a particularly large region. Numerous phenomena can be modelled, such as the chronic or accidental aspect of emissions, or chemical interaction with the atmosphere. A CFD approach also makes it possible to take precise account of the topography and/or urban planning of the site, and the associated turbulent fluctuations.

This type of modelling has become commonplace in the context of air quality monitoring. The example below illustrates this with the modelling of near-field odour dispersion near a factory.

CFD simulation of near field atmospheric dispersion for odour nuisance

Background​

  • Urban development project near a waste management centre.

  • Key physics concepts : atmospheric dispersion, multi-phases flows, atmospheric boundary layer.
cfd simulation atmospheric dispersion - mesh

Issues

  • Air quality assessment for future residents.

  • Simulation of near-field atmospheric dispersion.

  • Is the odour exposure limit value likely to be exceeded in certain areas and under certain weather conditions?

  • Which weather conditions are the most unfavourable?

  • What does CFD offer compared to a study with a Gaussian model and annual statistics?

Contributions to the project

  • Development of a model with construction of a detailed 3D model of the area of interest.

  • Selection of unfavourable weather patterns.

  • Performing simulations. Stationary approach, without taking puffing phenomena into account.

  • Analysis of odour concentration in the project area for the different configurations.

  • Contribution of CFD: detailed results at a fine scale (~1m), with precise impact of buildings.
 
 
cfd simulation atmospheric dispersion - streamlines
cfd simulation atmospheric dispersion - pollutant concentration

Technical environment

  • Linux and Windows.

  • Programming with Fortran and C++.

  • Meshing with Salome and OpenFOAM.

  • CFD with Code_Saturne.

  • Visualisation with ParaView.

  • Personal computing cluster.