Data to Value in the Recycling of Lithium-Ion Batteries

Batteries are highly integrated products composed of mixed materials that are highly affected by energy-intensive production and processing, security of supply, and raw material criticality.

Due to the energy-intensive production and processing of battery raw materials, valuable materials with high qualities are concentrated in batteries. With the steadily increasing demand for batteries, more and more of the high-quality and critical materials are needed. To ensure that these materials remain in the material cycle, effective and efficient recycling processes for end-of-life batteries are required. In the European Union, the EU Battery Regulation gives specific target values for efficiency and material recovery levels in battery recycling.

The strategy of lithium-ion battery recycling is firstly distributed into individual procedures including mechanical treatment, thermal treatment, hydrometallurgy and pyrometallurgy. These individual procedures are then combined into several conceivable process routes. Due to the resulting interdependencies, the optimization of the whole battery recycling process chain is a challenging task. To overcome this challenge, the structured collection, combination and analysis of data from the recycling process chain can form the basis for gaining a deeper understanding of the interactions.

In this work, a practical roadmap for value-added digitalization in lithium-ion battery recycling is proposed. It combines the framework of cyber-physical systems with strategies for data mining and artificial intelligence.

An exemplary methodological toolbox for a step-by-step data-based approach is illustrated. Firstly, the physical world needs to be investigated to define the problem. In this case, a mechano-hydrometallurgical battery recycling is investigated. It consists of a battery-to-blackmass process as mechanical treatment that recovers the valuable battery active material mixture. To remove the final binder and electrolyte residues, the black mass is fed into a pyrolysis oven. The pyrolyzed blackmass is then treated in a hydrometallurgical COOL-process to recover metal solutions with Lithium, Nickel, Cobalt, and Manganese, respectively. The aim is to uncover interdependencies along the whole process chain.
Secondly, data engineering methods are applied at the input transition level to acquire, harmonize and integrate data from the physical to the cyber world. In the cyber world, data analytics and business intelligence approaches are performed that prepare and interact with artificial intelligence models. As a result, transparent KPIs and active decision support towards the physical world can be provided to a human in-the-loop. Furthermore, modeling as well as physical and cyber processes should be automated to support targets like efficiency, safety, and sustainability.