Effective reclamation of vent gas in ethylbenzene dehydrogenation by coupling multi-stage circle absorption and membrane units

2016 
Abstract Ethylbenzene catalytic dehydrogenation can generate a large amount of hydrogen. Its vent gas is attractive with hydrogen content higher than 75 mol%. Nevertheless, styrene remaining in the vent gas is highly active to polymerize into oligomers, the gradual accumulation of which is harmful to hydrogen separation equipments. In this work, an integral separation process including absorption and membrane units is established to overcome this challenge. Absorption unit running with fresh ethylbenzene as the absorbent is proposed to deeply remove styrene, the residual content of which is reduced to less than 10 ppm. Thereafter, the generation and accumulation of styrene oligomers is restrained strictly to protect hydrogen separation equipments. Furthermore, the custom-designed absorption tower with a multi-stage cycle structure is used to reasonably improve styrene content in the used ethylbenzene at tower bottom, in the interest of reducing styrene recovery expense. For a 100 kt/a styrene plant with the vent gas flow rate up to 3180 N m 3 /h, a 5-stage tower is designed, which is 0.5 m in diameter, 5.0 m in height, and about 3.0 m in the total packing height of 25-mm metal Intalox saddles. Flow rates of liquids in the tower and the recycle routes are 1.19 m 3 /h and 1.05 m 3 /h, respectively. Finally, compressor and membrane for hydrogen recovery are optimized simultaneously with gross profit as the criterion. For compressor, its duty and outlet pressure are 520 kW and 1.85 MPa, respectively. For membrane separation unit, its area and permeation stage cut are 460 m 2 and 0.84, respectively. According to the simulation research, this integrated system can behave excellently with an annual recovery of 162 ton styrene, 281 ton aromatic hydrocarbons and 2.1 × 10 7  N m 3 hydrogen, with a gross profit up to 5.1 × 10 6  US dollars per year.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    32
    References
    8
    Citations
    NaN
    KQI
    []