Simulating Pollutant Dispersion from Accidental Fires with a Focus on Source Characterization.

Area of Science:

• Environmental Science
• Chemical Engineering
• Risk Assessment

Background:

• Oil and gas facilities store flammable compounds posing fire and explosion risks.
• Assessing health risks and air quality necessitates evaluating fire consequences.
• Atmospheric dispersion models are crucial tools for simulating pollutant distribution.

Purpose of the Study:

• To present the setup and findings of a modeling study on a hypothetical oil refinery fire.
• To evaluate environmental consequences by comparing simulation results with regulatory limits.
• To identify key variables influencing fire and pollutant dispersion through sensitivity analysis.

Main Methods:

• Utilized Lagrangian (SPRAY) and puff (CALPUFF) dispersion models.
• Focused on estimating uncertain source variables for input data.
• Conducted a sensitivity analysis to determine influential parameters.

Main Results:

• Ground concentration values remained below long-term regulatory limits.
• Dispersion model and source type significantly influenced simulation outcomes.
• Source area emerged as the most critical variable, causing up to +/- 60% variability in ground concentrations.

Conclusions:

• Model selection and emission physics algorithms affect sensitivity to input variables.
• Choosing the most appropriate model for specific incidental sources is essential for accurate real-world behavior approximation.