Mathematically modelling pyrolytic polygeneration processes using artificial intelligence

Pyrolytic polygeneration has been gaining interest in industry and academia as an organic waste treatment and valorisation endeavour that is capable of producing valuable carbon aerogels, pyrolysis gas and bio-oils. The outputs of pyrolysis exercises are highly dependent on feedstock composition and pyrolysis conditions. Due to the variable nature of organic waste material and due to limitations in variable pyrolytic conditions being actively implemented in process systems, it is difficult to understand and anticipate the outputs that would be produced from a given class of materials, using a set of pyrolytic conditions. In this work, we present a series of mathematical models that allows pyrolytic polygeneration processes to be practically calibrated, fine-tuned and predicted in an industrial context. By utilising pyrolysis experimental data from the literature, we model pyrolytic processes and generate governing expressions that apply to numerous pyrolytic polygeneration processes for cellulose-rich streams. We furnish this work with three case studies which demonstrate the utility and merit of our modelling approach. In these case studies, we predict the different conditions required to produce desired yields of biochar, pyrolysis gas and bio-oils and desired carbon aerogel properties.