Simulations of near-source wind development and pollution dispersion over complex terrain under different thermal conditions

TL;DR

This study develops a CFD model to simulate wind and pollution dispersion over complex terrain, accurately predicting wind speeds and SO2 concentrations under various meteorological conditions.

Contribution

The paper introduces a new CFD-based approach for modeling buoyant pollution dispersion over complex terrain with validated predictions against real measurements.

Findings

Wind speed predictions within 10% error at key locations

Accurate trend prediction of wind speeds across different elevations

Good agreement between modeled and observed SO2 concentrations

Abstract

A computational fluid dynamics (CFD) model that solves the steady-state Reynolds-Averaged Navier-Stokes (RANS) equations for buoyant compressible pollution dispersion under different meteorological conditions is developed. A 6.4 km by 6.4 km computational domain over a complex terrain with a height of 1 km above the ground surface is created. Meteorological data from multiple available sources are utilized to obtain boundary conditions of wind speed, air temperature, turbulent kinetic energy (TKE), and its dissipation rate. To evaluate the model, a monitoring network of four anemometers is deployed. Model predictions are compared with measurements of wind speed and the concentration of SO2 emitted by a local coke plant. Comparisons show that the predicted wind speeds are reasonably close to the measured mean wind speeds and the average error is within 10 percent at one location where relative fast wind speeds are recorded. The CFD model also predicts the correct trend of varying wind speeds across multiple sites of different elevations. The model also provides good predictions of SO2 concentrations for multiple cases, considering the complex nature of the terrain and meteorological conditions.