Tuesday, November 10, 2015: 10:35 AM
Salon I (Salt Lake Marriott Downtown at City Creek)
For the fabrication of effective, robust, and safe Fischer-Tropsch Synthesis (FTS) microchannel reactor, it is required to meet both high heat removal and C5+ productivity. In this research, axial symmetric thermal non-equilibrium porous media reactor model was developed in computational fluid dynamics (CFD) and C++ user defined function (UDF) for implanting given reaction kinetics. To manipulate the local temperature profile, which represents high heat removal, catalyst was diluted, i.e., mixed with various amount of inert bead to pack the reactor channel with discrete zone. Meanwhile, the C5+ productivity was simultaneously optimized to satisfy the economic feasibility, as well as the safety. For optimizing several objectives containing discontinuities and inconsistent function simultaneously, machine learning regression of CFD data set was performed to get differentiable function set. Out results suggested the optimized multiple catalyst dilution zones that decreased maximum temperature less then limited temperature (~5oC) and maximized the C5+ yield set as pareto efficient curve.