Pollution dispersion from a fire using a Gaussian plume model

A deterministic approach was used in this work to assess the PM2.5 pollutant dispersion in the air during a fire event. The pollution data were recorded with high time resolution by a monitoring station located 4 km South-West from the fire. The pollutant emission due to the fire was described as an equivalent stack having the height of the observed cloud of generated smoke. The pollutant dispersion was modelled by means of a Gaussian plume dispersion model. To this purpose, the unknown equivalent emission mass flow rate at the stack in the model was found out using the available experimental data of PM2.5 measured on the ground far away, considering the changing of the air stability between nighttime and daytime and the variable wind direction. Model results highlighted that the predicted maximum pollutant concentration was larger of an order of magnitude than the data value recorded at the monitoring station and exceeded the law limit value. A sensitivity analysis on the wind speed and the atmospheric stability conditions was performed as well to identify the worst case scenarios in case of a fire event. The main conclusion is that a dense network of measurement stations with high time resolution is necessary to properly monitor an area or to provide validation data for any predicting dispersion model in case of a pollutant release.