Organic Sulfur Removal From Gasoline By Pervaporation Membrane

Gasoline desulphurization by pervaporation (PV) membrane process is a newly-emerged desulphurization technology. And this process has the following advantages: low investment and operating cost, depth desulphurization, modular design, easily magnification and construction etc. The technology has attracted the attention of petrochemical field. The aim of this paper is to propel the newly-emerged desulphurization technology into practical application by using the self-developed membranes. The related investigations on membrane material, membrane modification, composite membrane, desulphurization mechanism, process simulation and scale-up experiments were conducted to solve the key problem for the industrialization of gasoline desulphurization technology with membrane method.In this paper, the theory of solubility parameter was used for the preliminary selection of membrane materials. The solubility behavior of typical gasoline components in common membrane materials was obtained by UNIFAC model for the further selection of membrane materials. Then the optimum membrane material polyethylene glycol (PEG) for desulphurization was determined by PV experiments. The separation performance and solvent resistant performance of PEG was improved by detailed study on crosslinking modification. Sulfur enrichment factor increased with the increase of amount of crosslinking agent and crosslinking time, while total permeation flux decreased. Unfavorable results occurred with exorbitant crosslinking resulted form excessive crosslinking agent and overlong crosslinking time. Since sulfur enrichment factor was for key consideration, 16% of crosslinking agent amount and 60min of crosslinking time were more practical.Desulphurization mechanism of PEG membranes was investigated by the study of solubility and diffusion behavior of typical gasoline components through PEG membranes. The study on transport behavior of typical gasoline components in PEG showed that thiophene components had higher solubility speed than other gasoline components, which is the key to fulfill the separation of thiophene/hydrocarbon mixture (i.e. gasoline). From the calculated results, it was easier for thiophene components to adsorb on and diffuse through membrane. chromatography (GC) analysis results corresponded to the above conclusion. Meanwhile, a full predictable model for the pervaporation process was proposed based on solution-diffusion mechanism. The sorption and PV experiments showed that UNIFAC model was a sound way to predict the solubility of solvents in membrane and the established model can predict the PV performance within certain error range.The study on the PEG/PVDF composite membrane showed that the pre-wetting method could effectively confine the intrusion of PEG solution to porous PVDF support layer in coating process. The composite membrane had a clear-cut boundary surface between the dense active layer and the porous support layer by SEM photos. It was found that the PV performance of the composite membrane changed slightly when the thickness of active layer varied from 4.25μm to 33.26μm. Effects of operation conditions on the PV performance revealed that flux decreased with increasing permeate pressure while increased with the increasing operating temperature. Sulfur enrichment factor increased firstly and decreased then when temperature rose. The peak value occurred at about 373K. Under the feed flow rate of 100mL/min, the membrane showed better and stable performance. Flux increased while sulfur enrichment factor decreased as the feed sulfur content increased. The PV performance of composite membrane was better than homogeneous membranes. PV experiments indicated that the membrane, with the crosslinking agent amount of 18% and solids content in active layer solution of 16%, had a stable performance for FCC desulphurization. The sulfur enrichment factor came to 3.6, and the total permeation flux was 2.7 kg /(m2.h).The process simulation was conducted build a bridge between experimental studies and potential applications of pervaporation for gasoline desulphurization on a pilot scale. Process simulation model was developed. By applying the simulation, the influence of interatage heating manner and process parameters, e.g. feed temperature, permeate pressure, feed sulfur content level and retentate sulfur content level on the required membrane area and retentate rate was investigated. Interstage heating was necessary for the scale-up PV process in gasoline desulphurization application. The interstage heating manner with constant membrane area difference was more practical, especially when a number of modules are used for large treatment capacity. It is found that the temperature increase can compensate partly for the decreased membrane performance brought by permeate pressure increase. The required membrane area and retentate rate decreased with the decreasing permeate pressure and increasing feed temperature. The decreased with the decreasing permeate pressure and increasing feed temperature, since such changes of operating parameters can improve membrane flux. Also, the permeate pressure of 650Pa was advised as reasonable value for the process scale-up in this simulation since it was the boundary point before retentate rate changed significantly. The required membrane area increased with the increasing separation difficulty, while retentate rate decreased under the same conditions.Based on the simulation results as well as discussions on considerable base data, scale-up plant was designed and established. The optimal operating conditions for scale-up plant were 373-383K of feed temperature, 5-10KPa of permeate pressure in vacuum shell and 18-21L/h of feed flow rate .The scale-up experiments under above conditions showed that the sulfur content of FCC gasoline was reduced from 710μg/g to 70μg/g after about 20h of continuous running, when the retentate rate was above 70%. Meanwhile, it was found that the desulphurization capacity for membrane sheets along the feed flow reduced, which correspond to the simulation results. Octane number has attracted more attention for gasoline desulfurization technology. Whether or not there is octane number loss for desulfurization process with PEG membranes is a necessary investigation from economic considerations. According to above GC hydrocarbon group analysis, the octane numbers of retentate stream were a little higher than those of feed. However, the octane numbers for feed, retentate stream and permeate products changed little as a whole.