Developing a parametric system model to describe the product distribution of steam pyrolysis in a Dual Fluidized bed

Artikel i vetenskaplig tidskrift, 2023

Steam pyrolysis is a thermochemical process that converts carbon-based materials into valuable gases. In general, the products of the reaction are syngas (H2,CO,CO2), low-molecular-weight hydrocarbon gases (methane, ethylene, and propylene), pyrolytic gasoline and oils, monoaromatic and polyaromatic species (tar), and carbonaceous residues (char) with ashes. However, the intricacy of the reactions comprising the process, the diversity of the product species, and the constraints linked to the sampling and measurement equipment, create a highly complex system. In this work, a method for data representation is presented based on a special Parametric System Model (PSM) that portrays product species measurements in a way that provides relevant information and valuable insights into the process. The method incorporates generic knowledge of the chemical nature of the reactions to create a constrained system in which the data can be expressed in parametric terms with meaningful statistical functions. The evaluated data were obtained from a high-temperature steam pyrolysis process performed in the 2–4-MW Dual Fluidized Bed reactor at Chalmers University using polyethylene as feedstock. The quantities of the hydrocarbon species detected in the gas product were taken for the PSM as a probabilistic system that can be described with a set of distribution functions. The carbon, hydrogen and oxygen balances were taken into account to build a constrained set of equations to find the parameters of the functions. The resulting model was proven to be useful as a prediction tool to quantify unmeasured carbon group species and to estimate process variables, such as the oxygen transport of the bed material. Also, it was demonstrated the potential of the model as a method to identify and estimate inconsistencies in the measurements, which improve the quality of the characterization data. The modeĺs outcomes find application in providing critical information for the control and evaluation of pyrolysis process and downstream operation of biorefineries.