A computationally efficient multi-scale simulation of a multi-stage fixed-bed reactor for methanol to propylene reactions

• A heterogeneous model was solved efficiently by hybrid of Matlab and Comsol.
• Recycling the undesirable olefins into reactor inlet promotes propylene yield.
• The simulated reactor agrees well with an industrial reactor of 500 KTY.
• Low propylene selectivity is due to large catalyst diameter and long contact time.

A multi-scale model incorporating interphase and intraparticle mass and heat transfer was established for a multi-stage fixed-bed reactor for methanol to propylene (MTP) process with recycle of the undesired olefins other than propylene. By converting the catalyst dimension and reactor dimension into a pseudo two-dimension and solving the resulting model by a hybrid method of Matlab and Comsol, the computation efficiency is 5 times higher than the conventional one that solves separately the single catalyst dimension at different reactor position. The model was validated by experimental data obtained from a lab-scale isothermal fixed-bed reactor. The calculated results show that propylene selectivity and methanol conversion achieve 62.6% and 99.99%, respectively, with space velocity of 0.741 gMeOH/gcat/h, which agree well with the practical data from a commercial six-stage Lurgi reactor with a capacity of 500 KTY propylene. It was found furthermore that the intraparticle diffusion resistance is notable and the contact time of reactant fluid is too long, resulting in a low propylene selectivity of the present commercialized MTP reactor, thus decreasing catalyst size, stage number and contact time will effectively promote propylene selectivity.