CFD Modeling of Air Emissions from Controlled Flare Tests

Although significant progress has been made in developing technologies capable of measuring flare efficiencies and emissions in situ, essential models necessary to interpret field data or to predict trends in flare performance over a range of conditions remain elusive. In this study, the issue of simultaneously addressing hydrocarbon combustion efficiency and soot emission was examined. Existing industrial-scale controlled flare test data for which both soot and DRE/CE/VOC emission data are available were modeled using computational fluid dynamics (CFD) approach. Two new combustion mechanisms, LU 3.0 and LU 3.0.1, which can handle flaring of C₁-C₄ light hydrocarbons and contain soot precursors in order to utilize the Moss-Brooks soot model built in FLUENT were developed.  Both mechanisms were validated with key laboratory performance data (ignition delay, laminar flame speed, and adiabatic flame temperature).  CFD simulations with LU 3.0.1 (being more stable in CFD runs) were then conducted to model the performance of Carleton University (CU)’s lab data as well as air-assisted flare tests in 2010 TCEQ flare campaign in Tulsa, Oklahoma using both the Eddy dissipation concept (EDC) and non-premixed probability density function (PDF) chemistry-turbulence interaction models. The validated methodology provides a convenient tool to analyze the potential trade-offs in emissions at different operating conditions, and ultimately to develop optimal control strategies to maximize environmental performance.