Manufacturers of polyethylene products constantly seek ways to improve their production processes. However, this poses major safety and financial risks because of the exothermic chemistry of ethylene. For this reason, simulation of polyethylene reactors is preferred over experiment. Many polymer-specific simulation packages have been developed but have limited usefulness due to their simplifying assumptions. Computational Fluid Dynamics (CFD) does not suffer from the same pitfalls; however, it is much more computationally costly. This is because a transport equation is needed for every species in the chemical mixture, including every chain length of polymer. The Method of Moments was developed to circumvent this burden by representing the chain length distribution with only a few terms. The primary flaw of this method is that the Molecular Weight Distribution (MWD) of the polymer is necessarily lost. Although a distribution can be reconstructed from its moments by assuming it conforms to a canonical shape, this fails to reproduce the bimodal structure of industrial polyethylene. To better approximate the MWDs, the distribution can be divided into classes based on their degree of branching, reconstructed assuming a canonical shape, and then summed to achieve the full distribution. This methodology was implemented into a Low-Density Polyethylene CSTR-like model in CFD. This model was then compared to an industry standard polymer simulation software, and results only differed by a maximum of about 2%. Additionally, the MWD reconstruction methodology was able to produce multimodal distributions. The number of polymer classes, assumed distribution shape, and summation methodology all impacted the shape of the distribution and can be tailored to the process under consideration. Once this methodology is implemented into a plant-scale CFD reactor model and validated with an industrial sample, process intensification studies can be conducted accurately.
