Process Design of Methanol to Propylene in Moving Bed

The methanol to hydrocarbons (MTH) reactions based on zeolite catalysts provide a promising non-oil light olefin producing route. Silicoaluminophosphate zeolite (SAPO-34) and aluminosilicate zeolite (ZSM-5) are two common catalysts for MTH process. Compared to SAPO-34, high Si/Al ratio ZSM-5 achieves higher propylene yield over ethylene and lower coking rate, thereby as to effectively alleviate the strong market demand for propylene. The Methanol to propylene (MTP) reaction over ZSM-5 has been widely studied. However, the size of catalyst adopted in most studies is in the range of 200-450 microns, where MTP reactions are controlled by internal diffusion, thus results are not intrinsic. Effect of internal diffusion can be ignored when catalyst size is less than 150 µm, in which case propylene selectivity and propylene to ethylene mass ratio (P/E) reaches the maximum. However such small catalyst inevitably causes excessive pressure drop, channeling and poor experiment reproducibility.

To avoid such problems, a porous ZSM-5/SiC foam structured catalyst was prepared, in which ZSM-5 coating thickness was about 15 µm, the crystal size was 100 nm and the Si/Al ratio was 200. MTP reactions on the structured catalyst were investigated with two kinds of feeding: methanol only and methanol alkene co-feeding. With methanol feeding, reaction conditions i.e. (temperature, methanol partial pressure, water/methanol ratio, methanol WHSV) remarkably affect the products distribution and catalyst coking rate. Increasing the reaction temperature and methanol WHSV favors the increase of not only the propylene yield but also the coking rate, thus the catalyst deactivation rate. Deactivated catalyst by coke, even whose methanol conversion declined to 80%, still well catalyzes alkene methylation, oligomerization and cracking. At the same time, hydrogen transfer on coked catalyst is greatly suppressed due to the modification of coke on zeolite acidity and pore structure, which reduces the yield of alkanes and aromatics. It was demonstrated that small catalyst size, reaction severity and conversion of the recycled alkene over coked catalyst are the important factors to be addressed for the successful design of the MTP process. However, catalyst will deactivate in about 200 h under high severity reaction conditions, which means that a fixed bed process is no longer applicable.

Furthermore, a novel three-step moving-bed process employing core-shell ZSM-5/SiC catalyst is proposed for MTP industrial development. Spherical catalyst composed of 15-100 µm ZSM-5 shell and 1.5-2.0 mm SiC core are used. In the first moving bed, a majority of methanol is converted to dimethyl ether over fresh/regenerated catalysts. The used catalysts are sent to the second step. In the second moving-bed, the mixture of dimethyl ether and unreacted methanol react to form hydrocarbons under high severity condition, and deactivated catalysts are transferred to the third moving bed. After separation of propylene, the hydrocarbon mixture is recycled back to the third moving bed for alkene equilibrium conversion to propylene. Catalysts at the outlet of the third moving bed are sent to the regeneration unit to remove the coke. In such a three-step moving-bed process recycled hydrocarbons are fed back to the third reactor alone rather than the second reactor with methanol and dimethyl ether together, which takes full advantage of high propylene yield when methaonl and dimethyl ether react at high severity and low alkanes and aromatics yields when alkenes oligomerization and cracking take place on coked catalyst. Meanwhile the propylene produced in the second and third step can be separated in time, which reduces the propylene expense due to secondary reaction. Thereby the maximum propylene yield can be achieved with lower circulation rate of hydrocarbons.