Dynamic model development for residence time distribution control in high-impact polypropylene copolymer process

Abstract

Two steps are usually involved in producing high impact polypropylene copolymer in a continuous process. The first is to polymerize propylene using Ziegler–Natta type or more recently metallocene catalyst, and the second is to add ethylene–propylene rubber (EPR). A narrow residence time distribution (RTD) of polymer particles is an important process design and operational target in this process. The broadening of RTD will lead to a widening of copolymer composition distribution in which results in inferior product properties. There are several techniques to obtain a narrower RTD, including by controlling size of particles leaving a well-mixed reactor. In this paper, a dynamic population balance model to track polymer particle age and size is presented. The model is applied to a well-mixed slurry reactor linked to a classifier. Polymer particle growth kinetics including catalyst deactivation with particle age are incorporated. The model agrees well with steady-state analytical solutions and process data for particle age and catalyst efficiency. The classifier acts to significantly narrow particle age distribution. However, the system is very sensitive to small change in feed catalyst size, catalyst deactivation can cause build up of small polymer particles trapped in the system if the feed catalyst size is too low. As a ‘soft sensor’, the RTD information from the validated model provides valuable insight for operational support. Further applications include strategies of reactor control design and optimization of grade transition policy.