Premium
Modeling of Polyolefin Polymerization in Semibatch Slurry Reactors: Experiments and Simulations
Author(s) -
Casalini Tommaso,
Visscher Frans,
Janssen Erik,
Bertola Francesco,
Storti Giuseppe,
Morbidelli Massimo
Publication year - 2017
Publication title -
macromolecular reaction engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.37
H-Index - 32
eISSN - 1862-8338
pISSN - 1862-832X
DOI - 10.1002/mren.201600036
Subject(s) - materials science , polymer , homogeneity (statistics) , polyolefin , polymerization , thermodynamics , particle size , slurry , particle (ecology) , molar mass , ostwald ripening , chemistry , mathematics , composite material , physics , nanotechnology , statistics , geology , oceanography , layer (electronics)
In this work, a well‐known single‐particle model (the multigrain model) is applied to simulate size growth and morphology evolution of polyethylene particles obtained through a slurry phase, catalytic polymerization in semibatch reactor. The model explicitly accounts for diffusion limitations within the polymer particle and between the particle and the solvent, as well as void fraction variations due to the particle growth. Catalyst behavior is described assuming a two‐site model, a good compromise between accuracy and simplicity, along with a conventional kinetic scheme, whose kinetic parameters are determined by fitting the model predictions to the experimental data. Polymer molecular weight is evaluated from population balances, solved through the method of moments. The semibatch reactor is modeled by means of fundamental mass conservation equations. Simulations show that the proposed model is able to describe not only the polymer properties of interest but also the evolution of the average particle size, thus providing a comprehensive overview of the system behavior. The model is further validated through regression‐free simulation of additional sets of experimental data, including materials with a bimodal molar mass distribution: the good agreement also found in this case confirms the robustness of the model and the estimated parameter values.