Dynamic modeling: The influence of the improved solid elutriation correlation on the two‐phase mathematical model of polyolefin polymerization in a gas‐phase fluidized bed reactor

Abstract

Dynamic modeling of the polyolefin polymerization process in the gas phase is relatively complex due to its high nonlinearity. This nonlinearity is induced by several factors including the reaction mechanism, the heat and mass transfer mechanism, the flow properties in both solid and gas phases, the type of reactor, and the correlations between the operating conditions and the chemical and physical properties of the polyolefin. In this study, polymerization of the polypropylene was observed in a fluidized bed reactor to potentially improve the solid elutriation correlations proposed previously, to increase the reliability and the accuracy of this two-phase model. By embedding the improved solid elutriation correlations, the production rate was reduced relative to earlier proposed solid elutriation correlations due to the loss of the product at the top of the fluidized bed reactor. The composition of the reactants and the catalyst flow rate also influenced the production rate, yet the inlet temperature did not.The current two-phase model is validated with the pilot plant data with the deviations lower than the previously validated model with reference to the production rate, the concentration of propylene and hydrogen, and the temperature of the reactor. K E Y W O R D Sfluidized bed reactor, gas phase, polyolefin, propylene, solid elutriation, two-phase model

The influence of catalyst flow rate and superficial gas velocity on the modified two‐phase model for propylene polymerization in a gas‐phase fluidized bed reactor

“…This solver was chosen based on the stiffness characteristics of the ODEs. [ 9,41 ] The output profiles generated by this improved two‐phase model were compared with the output profiles generated by the mathematical model implemented by Atan and Hussain [ 41 ] to evaluate heat loss during the polymerization process and with the experimental data published by Shamiri et al [ 37 ] to validate the current implemented model. For simulation purposes, Table 4 presents the operating conditions and the physical properties of this polymerization process.…”

“…To begin with, to ensure the validity and the reliability of the current, modified two‐phase model, the data generated from this model was compared with the pilot plant data published by Shamiri et al [ 37 ] To demonstrate the current model's improvements, its data were compared with the data generated by the model described by Atan and Hussain. [ 41 ] Before validating this model, the operating conditions and physical properties had to correspond with those described by Shamiri et al [ 37 ] and Atan and Hussain. [ 41 ] The data validation by Atan and Hussain [ 41 ] for production rate, propylene consumption, and hydrogen consumption executed was presumed to remain valid.…”

“…[ 8,9,36–38 ] Further to this, the solid entrainment at the top of the reactor was also embedded in the two‐phase model to predict product loss during polymerization. [ 2,39–41 ]…”

“…The present study improves the two‐phase model by considering heat loss through the fluidized‐bed wall during the polymerization process. The mass balance equations for propylene and hydrogen in the emulsion and bubble phases are derived from Atan and Hussain [ 41 ] and set out as follows: …”

“…[ 33,34 ] Shamiri et al [ 8,9,35–37 ] and Ho et al [ 38 ] also improved upon the two‐phase model by assuming that the emulsion phase did not maintain minimum fluidization, that polymerization reactions took place in both the emulsion and bubble phases, and that the size of particles formed in the reactor was constant. Abbasi et al, [ 2 ] Shamiri et al, [ 39 ] Salahuddin et al, [ 40 ] and Atan and Hussain [ 41 ] improved the two‐phase model by adding more details to the assumptions, respectively considering the presence of solid elutriation at the top of the fluidized bed reactor for the polymerization of ethylene, copolymerization of ethylene and propylene, and polymerization of propylene. Lastly, Bumphenkiattikul et al [ 42 ] studied the CFD‐based model by considering the presence of heat loss through the fluidized bed wall for the purpose of scaling up the reactor.…”