Application of Numerical Experimental Design in Consequence Analysis of Chlorine Release Using Breeze Incident Analyst Software

Although various studies have applied numerical experimental design and response surface methodology (RSM) in combination with consequence modeling tools for gas dispersion prediction, there remains a need for a more generalizable and robust approach. Such an approach should account for diverse industrial settings, both offshore and onshore, and variable environmental conditions including atmospheric parameters, wind, and topography. This study integrates numerical experimental design with BREEZE Incident Analyst software to model chlorine gas dispersion in an industrial environment. The effects of five independent variables, which are wind speed, leakage hole diameter, release height, ambient temperature, and surface roughness, on the downwind distance to a specific toxic threshold level were investigated. Significant interactions were identified, while the release height was found to have a minimal influence on gas dispersion. Well-validated predictive models were developed using RSM, followed by optimization to identify the worst-case scenario, defined as the maximum downwind distance to the ERPG-2 level (3 ppm for chlorine). The optimization results indicated that the worst-case scenario occurs under conditions of 1.5 m/s wind speed, 0.076 m leakage diameter, 38 °C ambient temperature, and urban surface roughness. This study demonstrates the effectiveness of integrating numerical experimental design with consequence modeling to enhance the accuracy and reliability of gas dispersion predictions while minimizing uncertainties.