Flow and Pollutant Dispersion in Urban Arrays for the Standardization of Cfd Modelling Practise

The study of the effect of obstacles on flow and dispersion in the atmospheric boundary layer is one of the most important topics in atmospheric dispersion research. Computational Fluid Dynamics (CFD) methods are increasingly used to predict concentration fields near buildings in an operational context, but extensive validations are still needed. In this paper, we analyse the effect on flow and dispersion due to the presence of a building array placed in the atmospheric boundary layer. The main aim is that of improving our current knowledge of the application of CFD methods to new case studies and contributing to the standardisation of CFD modelling practise in a wider context. Our analysis is based on the experience gained by the comparison of CFD numerical simulations, obtained with the commercial CFD model FLUENT, with wind tunnel data sets from the MUST experiment. INTRODUCTION Flow patterns around buildings have a strong influence on pollutant dispersion from sources placed within the urban area. The prediction of ground-level pollutant concentrations is important for the assessment of the impact of existing sources on people health and the environment. Traditionally, information about flow and pollutant concentrations has been obtained using field and wind tunnel experiments. In the MUST (Mock Urban Setting Test) (Yee, E. and C.A Biltoft, 2004) experiment, a large outdoor field study which has been reproduced in wind tunnel, it was attempted to simulate an urban boundary layer by the construction of a regular array of shipping containers in near-neutral atmospheric conditions. Recently, Computational Fluid Dynamics (CFD) has become an attractive tool to predict concentration fields near buildings, but at present there are not yet best-practice methodologies for using CFD as an operational tool. In this paper, we simulate the MUST experiment using the CFD code FLUENT (FLUENT 6.2, 2005). CFD simulations are performed using both the k–ε and the Reynolds-Stress turbulence models. The present work is part of our current research within the COST 732 Action (2005-2009), devoted to the study of the effect of obstacles on flow and dispersion in the real urban environment and to the standardization of CFD modelling practise for atmospheric applications. The work may be considered a verification of the efficiency of the Protocol and the Best Practice Guidelines (BPG) of the COST 732 Action and an assessment of their applicability. Here we present and discuss CFD simulations results for the mean velocity components and for turbulent kinetic energy using 0° and -45° approaching flow conditions. Pollutant dispersion results within the same array is presented and discussed in Di Sabatino, S. and R. Buccolieri (2007). METHODOLOGY Description of wind tunnel experiments (MUST) The wind tunnel data set used in this paper contains flow and dispersion data measured within an idealized urban roughness. The wind tunnel experimental setup originates from the Mock Urban Setting Test (MUST), an extensive field test carried out on a test site of the US Army in the Great Basin Desert in 2001. 120 standard size shipping containers were set up in a nearly regular array of 10 by 12 obstacles. The wind tunnel measurements within a scaled