Thermal conversion can transform the carbon-based waste into valuable chemicals to be further used in the petrochemical industry for a polymeric carbon circular economy. This work's aim was to identify chemical correlations between the thermal-cracking products and the feedstock polymer composition when using highly blended waste streams. The challenges addressed were to: (i) access a pool of experimental data on the monomer recovery potential of real-life, highly blended waste streams; (ii) estimate the polymer constituents of the mixed waste streams; and (iii) formulate a generic and systematic method to identify correlations between feedstock constituents and cracking products. Different post-consumer waste streams were investigated, including cardboard, automotive shredder residues, cable stripping waste, and textile waste. The cracking experiments were performed in a 2–4MWth industrial-scale Dual Fluidized Bed system at 800 °C using steam as fluidization agent. The polymeric constituents of the feedstocks were estimated using a numerical convex optimization method. To identify correlations between the feedstocks and products, a carbon bond-based classification was introduced. The experimental monomer yield ranged from 0.08 kg/kgf to 0.3 kg/kgf (f = feedstock) for the evaluated materials, corresponding to a carbon feedstock conversion rate between 14 % and 44 %. High yields of valuable monomers were obtained for the materials with the highest polyolefin content. The olefin monomer production correlated positively to the amount of aliphatic carbon in the original material and negatively to the carbon contents of the aromatic rings. From the trends observed, it was concluded that a framework based on carbon bond types is a promising approach to identify such correlations, which could serve as predictive tools for monomer recovery based on material's composition and overall process conditions.
