Mastorakos, Kim, Britter, Garmory

The current state of air pollution modelling typically neglects the correlations between chemical species. This may be a good approximation for low Damkohler number turbulent reacting flows, but is an increasingly questionable assumption for small-scale problems, such as the region close to emission sources. As a way to introduce micro-mixing into air pollution calculations, we use the Conditional Moment Closure (CMC) method that we also use for turbulent flames. The results reproduce the expectation that for species that react slowly in the atmosphere (e.g. CH4, CO, some VOC's), neglecting correlations is not a bad assumption. However, for species affected by fast chemistry (radicals such as OH, HO2, NO3 etc.), introduction of micro-mixing has a very large effect on the calculated mean concentrations (Fig. 1). Further work is underway to interface the CMC model with practical Air Quality Models.

Figure1.  The difference between mean concentrations calculated with CMC and with the conventional method used in AQM of neglecting species concentration fluctuations, for a power station plume evolving in a uniform wind field.

Work started in 2005 emphasizes the use of the Stochastic Fields model with detailed chemistry and, in conjunction with Prof. Rex Britter's group, the implementation of these models for practical calculations. The Stochastic Fields, or Eulerian Monte Carlo, method developed by Valiño is a relatively new transported PDF method for the modelling of turbulent reacting flows. It is developed from the modelled PDF transport equation, but uses the random development of a number of Eulerian fields extending over the spatial domain of the flow rather than Lagrangian particles moving through the flow. We have used this method to model a laboratory reacting plume experiment, in which we saw very close predictions to the experimental data and we were able to show that effects of reactant segregation and micromixing need to be taken into account in some atmospheric reaction problems. One of the advantages of the Stochastic Fields method is that it can easily be linked to existing, proven Eulerian grid based CFD solvers. We are currently using this technique to apply the Stochastic Fields method to the problem of reacting flows in jet engine plumes at airports. Some preliminary results are shown below.

Figure2.  Horizontal contour plots of NO2 concentration (ppm) in jet engine plume for arbitrary inlet conditions by (a) Stochastic Fields method and (b) using no micromixing models.