Development and verification of a two phase reactor model for industrial light naphtha isomerization reactors

Abstract

Isomerization process is one of the most important chemical processes used in the petroleum industry. Its product is named as the “isomerate” which is the major element of the gasoline pool that compensates for the octane rating and volume loss upon reduction of benzene and butane derivatives. In addition, environmental regulations limit the sulfur and benzene levels in gasoline, which are expected to become even more stringent with future restrictions on maximum vapor pressure and olefins content. UOP PenexTM process is one of the most preferred applications for isomerization. It converts low-octane light naphtha to higher-octane, branched isomers. In addition to isomerization reactions, benzene reduction, ring opening and cracking reactions also occur in reaction mixtures. A typical Penex reactor involves catalytic cracking and conversion on a fixed bed of chlorinated Pt/Al2O3 catalyst and increases the octane number from 60-70 to 85-87. The reactions take place in the presence of hydrogen with low partial pressure, at operating pressures around 32 bars and in a temperature range of 130-160°C. Isomerization involves moderately exothermic reversible reactions; hence, it is controlled by thermodynamic equilibrium that is more favorable at low temperatures. This implies that the reaction temperature has to be maintained in an optimal range between the thermodynamic and kinetic limits. A descriptive reactor model with sufficient complexity is essential in order to optimize the process by maximizing the isomerate up to a given octane number and quantity and related process improvement. In this work, the Penex reactors at Tüpras Izmir Refinery are handled rigorously, using a steady-state, co-current, down-flow, two-phase reactor model where chemical reaction rates are described by power law kinetics. Kinetic parameters are estimated by nonlinear regression that uses an objective function based on the minimization of the sum of square errors between plant data and calculated values. The results show that, i-C5 product ratio is typically maximized at ca. 160°C, slightly above the required ca. 130°C for obtaining maximum 2,2-dimethylbutane product ratio. It is also seen that C5 paraffins isomerize more readily than C6 paraffins. Moreover, the benzene in the feed is found to be completely converted in the 100-200°C range studied.