Parameter study of a ballistic separator in waste treatment for optimized mechanical processing and improved recovery of valuable materials from mixed commercial waste

Abstract

Industry-scale experiments have to be performed to investigate the challenges of processing mixed waste, for example, to identify potential improvements in machines and to optimize waste treatment plants. Conducting such experiments involves a high level of experimental effort, e.g. caused by the inhomogeneity of mixed waste. A parameter study of a ballistic separator is performed, to discuss the influencing machine factors on the material and to uncover optimization potential. Therefore, a methodical experimental plan is developed, to provide representative and meaningful results keeping the effort to a minimum. Due to data distortion caused by the inhomogeneity of the input material, a material batch is used in this study, representing the heterogeneity of the material but keeping the material-specific influence factor to a minimum. In total 22 test runs are carried out in randomized order. The selected models, based on the responses of the data gained, provide significant results regarding the separation behavior of a ballistic separator, since all $p$-values (< 0.0001) are below the threshold ($\propto =$ 0.1). Considering the factors paddle angle, paddle speed and air supply, each has an influence on the separation into 3D and 2D material stream. A Pareto front is shown to achieve the optimum setting for the respective yield and purity. For example, good purity (> 93 %) can be achieved by setting the factors paddle angle between 53–71 %, paddle speed over 83 %, and air supply to a maximum, resulting in a lower yield (< 30 %).