Development and application of a predictive modelling approach for household packaging waste flows in sorting facilities.

Household packaging waste sorting facilities consist of complex networks of processes to separate diverse waste streams. These facilities are a key first step to re-enter materials into the recycling chain. However, so far there are no general methods to predict the performance of such sorting facilities, i.e. how efficiently the heterogeneous packaging waste is sorted into fractions with value for further recycling. In this paper, a model of the material flow in a sorting facility is presented, which allows changing the incoming waste composition, split factors on the sorting units as well as the setup of the sorting facility. The performance of the sorting facility is judged based on the purity of the output material (grade) and the recovery of the input material. A validation of the model was performed via a case study on Belgian post-consumer packaging waste with a selection of typical waste items that can be found in this stream. Moreover, the model was used to predict the possible sorting qualities of future Belgian post-consumer packaging waste after an extension of the allowed waste packaging items in the waste stream. Finally, a sensitivity analysis was performed on the split factors, which are a key data source in the model. Overall, the developed model is flexible and able to predict the performance of packaging waste sorting facilities as well as support waste management and design for recycling decisions, including future design of packaging, to ensure proper sorting and separation.