(356c) A Data-Driven Approach to Systems Identification for Scrap Metal Shredders in the Context of Circular Economy

Metal recycling is imperative for environmental and sustainable development. Recycling technologies have attracted significant attention, getting increasingly mature [Reuter et al. (2013)] and expected to assume a fundamental role in the transition towards circular economy [George et al. (2012)]. Shredders are essential processes in the metal scrap industry. Their efficiency is critical to sustainable recycling as shredders account for the largest energy consumers and the most expensive units to operate in the plant. While such processes are based on relatively simple principles and in operation for a long time, their feedstock (scrap) is difficult to characterize and relate to standard control schemes; operation is carried out empirically with an apparent scope for energy savings once the system is better understood. Shredding aims in the separation of various metals found in end of life (EOL) products. Separation units separate the various materials composing the feedstock that has been shredded. The underlying assumption is that fragments of the original EOL products are sufficiently small so that each one consists of a single material. Energy use relates to the use of hammers that break metals into fragments (shredded products) that are usually separated with magnetic separators followed by eddy-current separators. Handpicking is also used at various stages of the process. Products may vary and are typically classified as (1) “Shredded”, (2) Zorba, (3) CuFe, (4) Inox and a by-product (5) Fluff. Feeds, however, include a much wider range of materials and EOL products that include whole cars, car engines, water heaters etc. In our case, they are organized in 21 different scrap categories (ISRI Scrap Specifications Circular, 2022).