Enabling data-driven process dynamic modeling for extractive leaching and chemical precipitation

To address the limitations of static models and gain insight into the processes of extractive leaching and chemical precipitation, a data-driven dynamic modeling strategy is proposed using a Lithium-ion battery recycling case study. The data correlations among pH, temperature, redox potential, conductivity and system state are investigated. Predictive models are then developed to describe the system state online and are employed as surrogate models for time-intensive offline chemical analyses. This enables further process optimization, such as time-saving measures and improved process efficiency through dynamic parameter studies. The proposed strategy serves as a guideline for dynamic modeling and integrates big data methodologies into chemical engineering. • A new strategy for dynamic modeling chemical extractive processes is developed. • The linearity and non-linearity between sensor data and system state are studied. • Data-driven models serve as surrogates for time-intensive chemical analyses. • Leaching states can be accurately predicted by artificial neural network models.