In this investigation, we demonstrate our methodology in developing a comprehensive computer simulation model for the low-density polyethylene process in a tubular reactor using Polymers Plus. We use the perturbed-chain statistical associating fluid theory to describe the thermodynamic properties of the system. A comparison with literature data shows that the selected equation of state does a very good job in describing the physical properties and phase equilibria of the system. A detailed reactor model was proposed on the basis of transport literature that provides insight into the various resistances to heat transfer that arise during polymerization, and a comprehensive free-radical kinetic model was developed that describes the various individual mechanisms of the polymerization of ethylene and the properties of the polymer product. Results from the proposed simulation model were used in comparison with plant measurements from an Equistar Chemicals plant, in both correlative and predictive modes, for several polymer grades. In all cases considered, very good agreement was observed between simulation results and plant data on reactor temperature profiles, polymer properties, and production rates.

In this work, we investigated various approaches for the modeling of the high-and low-pressure separator units downstream from a low-density polyethylene tubular reactor using the Polymers Plus software package. First, we examined the performance of thermodynamic equilibrium by using the perturbed-chain statistical associating fluid theory (PC-SAFT) equation of state. Experimental data taken from the open literature were used to obtain the model parameters.Comparison with data from an Equistar plant showed that the PC-SAFT simulations agreed very well with the low-pressure separator residual-ethylene solubility measurements. There were, however, significant discrepancies between the model and the plant data for the highpressure separator, indicating that the high-pressure separator is not operating at equilibrium conditions. A further investigation was performed where a physical mechanism based on a bubble formation model was evaluated and a mathematical correlation using dimensionless numbers developed. The resulting model yielded high-pressure separator predictions that agreed adequately with plant data.