AIR QUALITY ESTIMATION AND ERROR ANALYSIS

Abstract

Air quality in urban areas is being influenced by many factors, mostly anthropogenic. A proper estimation of pollutant distribution is very important for making decisions for air quality improvement. The application of CFD techniques provides the most detailed information about pollutant dispersion. If we consider pollution a problem, the application of CFD techniques is not a solution to the problem, but an estimation tool, which supplies environment protection experts with the necessary data.
In this paper, the measured traffic intensity is used for defining initial and boundary conditions for the CFD model. To this aim, the COPERT software package was used. Since it is well known that wind characteristics have the largest influence on pollutant dispersion, they were carefully treated. There are two main wind directions in the City of Niš area, so two groups of simulations were done, for each main wind direction separately. The PHOENICS software package was used for pollutant dispersion estimation.
The two simulations were done considering the traffic intensity distribution, one for the morning traffic minimum, and the other for the afternoon traffic maximum, taking the measured wind data into account at the same time. The results of the numerical simulation show good agreement with the measured wind data and CO2 concentrations.
Three turbulence models were compared: the standard k-ε model (SKE), the RNG k-ε turbulence model (RNG KE) and the Chen-Kim modification of the k-ε turbulence model (CHEN-KIM KE). Error analysis was performed on the basis of the comparison of the u and v velocity components and CO2 concentration with the measured results.