(403c) From Degradation Kinetics to Process Evaluation: Challenges and Solution Approaches in Designing Chemical Recycling Processes

The chemical industry is more and more confronted with social and political expectations to increase process efficiency and to provide new process technologies, allowing for the utilization and implementation of previously unused material streams such as plastic waste. For that purpose, circular economy principles in terms of circularity strategies are considered increasingly: besides the utilization of circularity strategies such as product re-design (which can be driven e.g. by means of simplified recycling approaches) and the avoidance and/ or substitution of crude oil derivatives and other fossil resources, the focus is strongly on recycling processes. In the chemical industry, such processes aim at the reintegration of material flows into specific production processes, e.g. the integration of pyrolysis oils in steam crackers as a naphtha substitute [1] or the re-integration of monomers, obtained from chemical recycling processes, in polymer production [2, 3]. Regarding plastics in particular, chemical recycling technologies are currently strongly focussed: The complementary reintegration of material flows from both mechanical and chemical recycling processes aims at substituting crude oil derivatives and allowing for closed material cycles, thus reducing waste volumes and fossil resource depletion in the genuine sense of a circular economy.

To date, there are three main limitations and challenges in the design and engineering of chemical recycling processes: first, there is a lack of systematic methodologies for the implementation and combination of both experimental data - especially regarding kinetics of the conversion steps - and model-based approaches for the design and scaling of reactors and entire processes for chemical recycling technologies (challenge #1). This is a crucial limitation since the availability of a design tool for reactors and entire processes is crucial for knowledge-based engineering. Secondly, the varying composition of the chemical recycling products such as pyrolysis oils, which is a consequence of different waste sources and process conditions, requires robust downstream processes to ensure successful reintegration of the obtained products and material streams. For both process robustness, and chemical recycling downstream design, no knowledge-based approaches are available (challenge #2). And finally, an environmental assessment of possible reintegration pathways is essential to assess the sustainability at the earliest possible stage of process development; this aspect is challenging because it must typically be carried out at a very early stage of process development, in which data availability and data quality is often not given (challenge #3). Based on polystyrene as an example, this contribution presents different solution approaches to overcome the above-mentioned challenges in the design and evaluation of chemical recycling processes.

To define and identify suitable operating windows of processes, and to implement experimental data especially obtained from plastic waste degradation investigations, process simulation tools using computer-based modelling approaches are often employed. Based on validated process simulations, these processes can be scaled, optimized, and evaluated regarding, e.g., their energy consumption and environmental impacts without the need for costly and time-consuming experimental trials. In chemical recycling processes, the knowledge-based definition and identification of operating windows considering different and varying types of feedstocks is of particular importance. Process simulation can help identifying such operating windows by using computer-based modelling to understand and predict the behavior of both conversion steps and downstream operations. But simulating especially the reaction networks of chemical recycling processes is challenging due to the complexity of the reaction system. For example, pyrolysis processes involve a complex reaction network, resulting in the formation of various reactive intermediates, that lead to a large variety in the final product composition. The sheer number of possible reaction pathways makes it difficult, if not impossible, to solve the stoichiometry of the entire reaction system. Consequently, the reaction systems must be treated as black boxes in which the feedstock reacts to the actual products. To achieve a knowledge-based process simulation, the reaction model must contain kinetic data representing the actual behavior of the reaction system. The derivation of those kinetic data sets for the degradation of polymers is addressed in numerous studies ([4–6], among others), mostly using pure polymer fractions or defined polymer mixtures, some already gathering kinetic data from actual waste fractions. Those waste fractions, especially post-consumer wastes, are not homogenous. The composition, the particle sizes and the grade of impurity differ a lot through the year influenced by the season, the collection area, and the sorting facilities. Therefore, further investigation on how these varying feed-parameters affect the reaction is needed, which is addressed in this contribution: Kinetic data sets are derived and compared for real waste polystyrene fractions collected from sorted light packaging waste using simultaneous thermal analysis measurements, providing a solution approach for challenge #1. Two polystyrene samples are taken before and after a cleaning process resulting in two fractions with a defined grade of impurity. Each is sorted by particle size to investigate the influence of the impurities on the degradation and product yield of the reaction. The results are used to overcome the stated limitations giving suitable kinetic data sets for process simulations and development.

The resulting varying composition of the pyrolysis oil due to the different waste composition requires robust downstream processes to ensure successful product reintegration. Thus, to provide a solution approach for challenge #2, different pyrolysis oil compositions obtained from waste streams with varying styrene content are investigated in a downstream process simulation using sensitivity analysis. In the Aspen Plus^® based process simulation, process robustness is evaluated in terms of the sensitivity of overall energy demand and product quality towards distillation process parameters: reflux ratio, distillate-to-feed ratio, and side-stream control. Depending on the pyrolysis oil composition two different reintegration pathways for the main product styrene as well as the by-products are considered in the distillation sequence.

Finally, a prospective life-cycle assessment (LCA) is conducted comparing the chemical recycling pathways to waste incineration at industrial scale. The chemical recycling route is shown to be favorable over the waste incineration considering the impact categories global warming potential (GWP) and fossil fuel depletion potential (FFDP). The prospective LCA also identifies environmental trade-offs by including the impact categories freshwater eutrophication (FE), marine eutrophication (ME) and terrestrial acidification (TE). The FE and ME potential is much lower for waste incineration, although this is mainly due to the high eutrophication potential of the current electricity mix. Finally, the prospective LCA results of the possible reintegration pathways are presented to assess the sustainability at the earliest possible stage of process development and discuss methodological limitations and possibilities based on this example as a solution approach for challenge #3.

In summary, this contribution gives detailed insights into major challenges and solution approaches for the design of chemical recycling processes, mainly focusing on the integration of experimentally obtained reaction kinetics, model-based process development and prospective LCA. Based on that, the applicability of the proposed methods and approaches are evaluated with respect to the forthcoming shift towards a circular plastics economy.

[1] M. Kusenberg, G. C. Faussone, H. D. Thi, M. Roosen, M. Grilc, A. Eschenbacher, S. de Meester, K. M. van Geem, The Science of the total environment 2022, 838 (Pt 2), 156092. DOI: https://doi.org/10.1016/j.scitotenv.2022.156092

[2] A. Merkel, L. Plessing, G. A. Luinstra, M. Grünewald, P. Biessey, Chemie Ingenieur Technik 2023, 95 (8), 1323 – 1331. DOI: https://doi.org/10.1002/cite.202100210

[3] M. Holtkamp, M. Renner, K. Matthiesen, M. Wald, G. A. Luinstra, P. Biessey, Resources, Conservation and Recycling 2024, 205, 107558. DOI: https://doi.org/10.1016/j.resconrec.2024.107558

[4] W. Zhang, J. Zhang, Y. Ding, Q. He, K. Lu, H. Chen, J. Cleaner Prod. 2021, 285, 125042. DOI: https://doi.org/10.1016/j.jclepro.2020.125042

[5] J. Nisar, G. Ali, A. Shah, M. Iqbal, R. A. Khan, Sirajuddin, F. Anwar, R. Ullah, M. S. Akhter, Waste management (New York, N.Y.) 2019, 88, 236 – 247. DOI: https://doi.org/10.1016/j.wasman.2019.03.035

[6] X. Ren, Z. Huang, X.-J. Wang, G. Guo, J Therm Anal Calorim 2022, 147 (2), 1421 – 1437. DOI: https://doi.org/10.1007/s10973-020-10411-9