MODELING MIXING EFFECTS IN FLUIDIZED-BED, ETHYLENE-POLYMERIZATION REACTORS WITH CONDENSING AGENTS

CHIOVETTA, MARIO G.; BORTOLOZZI, RAÚL A.

Montreal, Canada

Congreso; 8th World Congress of Chemical Engineering; 2009

École Polytechnique de Montréal and McGill University,

8th World Congress of Chemical Engineering 1128 PE4 10:30 Tuesday AM 514b A set of modeling options are studied in the case of a bubbling fluidized-bed reactor for the production of polyolefins, in particular, the synthesis of polyethylene using supported catalysts. Modeling of a typical, commercial-scale reactor is performed employing a two-tear scheme. The first one, a macroscopic approach, is used to produce an initial estimation of the temperature in the gaseous mixture exiting the top of the reactor. The range for this temperature is limited by the reactor conditions producing polymer degradation. From this limit, the maximum reactor throughput without the addition of vaporizing components can be approximated. Later, these first-level predictions can be applied to calculate the amount of condensates that must be added to boost the reactor polymer production for a given operating temperature profile. The second modeling level consists of detailed simultaneous particle and reactor representations, proposing the geometric set-up where mass and energy balances are performed and the polymerization reactions solved. The fluidized bed is modeled as a heterogeneous system with a bubble gas phase and a solid-particle emulsion, envisioned as a pseudo-continuous phase. The catalyst active sites are considered located within a growing, solid, ever changing particle composed of the support, the catalyst and the produced polymer. Particle morphology and geometry are analyzed during particle residence time to follow the changes in the reactants transport conditions. Solving for the above mentioned balances, temperature and monomer concentration profiles along the bed height are calculated. Several modeling options for the division of the reactor in sections are studied. Results for each of these modeling options are presented and compared with the existing data from operating reactors. The model is used to establish which options are adequate for properly representing the known reactor behavior.