CFD Simulations of An Ethylene Flame and Propylene Flares

Air quality models (CAMx, CMAQ) often significantly under-predict observed peak O3 with the current VOC emission inventories in HGB/BPA. Even though this could be due to the inaccuracy of the photochemical models used, it is very likely that some VOC emissions may be under-reported or unreported due to the assumption of robust destruction efficiencies of industrial flares.  From rigorous combustion chemistry, incomplete combustion of hydrocarbons is a likely source of free radicals such as perhydroxyl, radical producing species such as formaldehyde, and highly-reactive VOCs (HRVOCs) such as ethylene and propylene.

CHEMKIN CFD and FLUENT were used for our simulations. The simulation of laboratory McKenna burner is aimed at validating the turbulence model parameters and a 50-species reduced mechanism based on a combination of GRI 3.0 and USC mechanisms for the combustion of C1-C3 light hydrocarbons. The simulation of Tulsa flare tests is aimed at predicting the combustion efficiencies (CE and DRE) and hydrocarbon emissions based on validated CFD models. The simulations can establish the needed data base for finding the relationship between flare efficiencies/emissions and flare operating/meteorological conditions.