Development of air pollutants dispersion algorithm in urban industrial park for air pollution simulation

Abstract

Environmental Impact Assessment analysis is generally restricted to neighbourhood scale air pollution simulation using the Gaussian Plume model (GPM). This approach expected to enhance the resolution of ground level concentration in the conventional GPM based software up to building scale by using Computational Fluid Dynamics (CFD) model alongside the GPM. The aim of this study was to develop an air pollution prediction algorithm for air pollutants release from industrial stacks. It was used to estimate, simulate and control air pollution in urban industrial park using integrated GPM and CFD model. The GPM was used for regional air pollutant level prediction to find high pollutant concentration zones. Whereas, CFD model was used for detailed simulation on respective polluted areas. In order to achieve this, a building detection algorithm from satellite image based on building footprint detection and height estimation from shadow thickness has been proposed to reduce pre-processing effort of the present CFD solver. The present CFD algorithm were based on Fractional Step Method for efficient steady state solver and Prandtl Mixing Length turbulence model for low cost turbulence calculation. The accuracy of the CFD algorithm has been tested and verified against benchmark problems (less than 3% error for lid-driven cavity problem, less than 8% for flow over isolated cube). It was discovered that CFD algorithm developed in this study is sufficiently accurate as other wind flow models with slight over prediction in wind speed by 1.04 m/s (15.6% are below 10% error) and able to predict the wind direction correctly within 60° angle (37.5% are within 15° angle) compared to measurement data. Air pollutant release from major stacks in Pasir Gudang Industrial Park was studied using GPM and high NO2 concentration zone (1800 µg/m3) was found in Taman Air Biru. Results suggest that 24-hour averaged SO2 and PM10 maximum ground level concentration are well within Ambient Air Quality Standard (AAQS) limits with 8.9 µg/m3 (8.4%) and 11.4 µg/m3 (7.6%) respectively. Meanwhile, 24-hour averaged NO2 concentration exceed AAQS limit with 270.3 µg/m3 (360%). The detailed CFD simulation of wind distribution and pollutant dispersion process within the area was presented. Present CFD model (1800 µg/m3) over predicted 1-hour averaged NO2 ground concentrations by a factor of 3 compared to the present GPM (700 µg/m3) but it provides more information on wind distribution as well as pollutant dispersion process. A new atmospheric dispersion solver has been developed that is able to simulate pollutant dispersion on both regional scale using GPM and building scale using CFD model.