SIMULATION OF PROPYLENE POLYMERIZATION - AN EFFICIENT ALGORITHM

Multigrain type models for propylene polymerization suffer from the drawback that they take excessive computer time to predict the degree of polymerization (DP) and the polydispersity index (Q) of the polymer produced. An efficient algorithm using adaptive grid-point spacing in a finite-difference technique has been developed to overcome this defect. Using this algorithm, we have studied the effects of catalyst deactivation as well as multiplicity of catalyst sites, on the rate of polymerization, and on the values of DP and Q. The improved algorithm can easily be used to model industrial reactors where additional physicochemical effects are present.