Influence of hydrocarbon species on the removal of thiophene from FCC gasoline by using a spiral wound pervaporation module

Pervaporation is an emerging technology in the field of thiophene removal from the Fluid Catalytic Cracker (FCC) gasoline. The performance of pervaporation process for thiophene removal depends on different types of hydrocarbon species present in the FCC gasoline due to their distinctive transport properties in the membrane, as well as due to their distinct physical and thermal properties. In membrane modules, the presence of different hydrocarbon species may also affect the variations in flow variables such as the temperature and concentrations in the feed and permeate channels. This in turn may further influence the performance of the module. The present study shows how different hydrocarbon species present in the gasoline influence the performance of a spiral wound module. First, the experimental data for a number of binary mixtures are obtained for a variety of hydrocarbons and thiophene. This data is then analysed by using a mathematical model for the spiral wound module. The membrane transport parameters are determined by using a parameter estimation technique. The results indicate that the thiophene (C4H4S) removal from different binary mixtures may be sorted out based on the type of hydrocarbon species present in the binary mixtures as linear alkane>alkene>branched alkane>aromatic compounds. Results for a ternary mixture show that the same mathematical model may also be used for predictions of the module performance by assuming the ternary mixture to be equivalent to a binary mixture of two species: first is thiophene, and second species is a pseudo-species, having the average physical properties of two other hydrocarbons present in the ternary mixture. The same approach may be extended to a multicomponent feed of a real FCC gasoline, containing a variety of hydrocarbons and sulphur containing compounds. This approach may provide sufficiently accurate predictions with much lesser efforts in comparison to more rigorous, and much more tedious, multicomponent modelling.