Early Detection of Agglomeration in Fluidized Bed Polymerisation Reactors

Polyethylene (PE) and polypropylene (PP) make up for a large part of the total worldwide polymer production. Originally, both polymers were often produced in solution processes. However, fluidized bed reactors for polyethylene and polypropylene production gain increasing industrial interest because of the relatively low production costs, the ability to produce a wide range of polymer grades, and the environmental advantage of the absence of solvent. Nevertheless, fluidized bed production of polyolefins does have a serious drawback: agglomeration of particles is a major problem in this process (e.g., Burdett et al., 2001). The initial formation of particle clusters will deteriorate the otherwise uniform heat distribution in the bed. In the resulting hot pockets further agglomeration is promoted due to increased reaction rate and particle stickiness. Without proper control this process leads to excessive growth of particles in the agglomerate. This ultimately results in complete defluidization, forcing undesired plant shutdown and subsequent expensive cleaning of the reactor. To control agglomeration, operators commonly rely on average pressure drop or temperature measurements. Unfortunately, these techniques are insensitive to the actual agglomeration process: the detection is often too late.