3221
A Novel Approach for Data-Driven Design of
Battery Recycling Processes
Nima Emami, Milica Todorović
Department of Mechanical and Materials Engineering, University of Turku, Finland
Luis A. Gomez-Moreno, Anna Klemettinen, Rodrigo Serna-Guerrero
Department of Chemical and Metallurgical Engineering, Aalto University, Finland
ABSTRACT: With the growing demand of raw materials to enable the ongoing electrification transitions,
robust battery recycling technologies will also become necessary. Our work explores a novel approach to design
and optimize recycling processes parameters by using a data-driven analysis and optimization of processing
and materials recovery for the various battery components. By applying a big data approach to HSC-Sim
process simulation software, thousand of possible scenarios were simultaneously explored in all unit operations.
The result is a distribution of recovery and grade of materials potentially obtained under the given range of
operating variables, thus defining under which conditions the process design is capable of producing the target
products. We demonstrate that large-scale data analysis can be used to guide decision making and refine the
process towards improving materials recovery targets.
INTRODUCTION
The global push to phase out fossil fuels has drawn the
attention to Li-ion batteries (LIBs) as integral part of
renewable energy grids and transportation (Kittner, Lill,
and Kammen 2017). The demand for reliable energy stor-
age to complement intermittent renewable energy sources
is crucial for facilitating the transition to the electrifica-
tion of systems (Diouf and Pode 2015). In this context,
the increasing production of LIBs in the coming decades
will pose challenges to waste management as they reach
their end-of-life (EoL) (Srivastava et al. 2023). As a result,
adopting a circular economy approach for battery materi-
als becomes essential, and researchers and practitioners are
actively searching technologies to improve recycling pro-
cesses (Gutierrez 2021).
Presently, industrial and pilot-scale LIB recycling typi-
cally involve pyrometallurgical and hydrometallurgical
routes or their combination, often aided by some degree of
sorting and mechanical pre-treatment (Velázquez-Martínez
et al. 2019 Brückner, Frank, and Elwert 2020 Makuza et
al. 2021). Pyrometallurgical routes operate at high tem-
peratures and are thus considered energy intensive, but
they require minimum pretreatment and are flexible on
the composition of batteries used as feed. Processes such as
smelting, roasting, incineration, reduction and pyrolysis are
employed in this route (Makuza et al. 2021). In contrast,
hydrometallurgical extraction is carried out at low tem-
peratures, although they require more complex mechani-
cal pre-treatment processes for discharging, dismantling,
and reducing the particle size of the battery waste mix.
Leaching, precipitation, solvent extraction and chromatog-
raphy ion exchange are some of the typical process units
in hydrometallurgy (Saleem, Joshi, and Bandyopadhyay
2023). Mechanical pre-treatment encompasses a multitude
A Novel Approach for Data-Driven Design of
Battery Recycling Processes
Nima Emami, Milica Todorović
Department of Mechanical and Materials Engineering, University of Turku, Finland
Luis A. Gomez-Moreno, Anna Klemettinen, Rodrigo Serna-Guerrero
Department of Chemical and Metallurgical Engineering, Aalto University, Finland
ABSTRACT: With the growing demand of raw materials to enable the ongoing electrification transitions,
robust battery recycling technologies will also become necessary. Our work explores a novel approach to design
and optimize recycling processes parameters by using a data-driven analysis and optimization of processing
and materials recovery for the various battery components. By applying a big data approach to HSC-Sim
process simulation software, thousand of possible scenarios were simultaneously explored in all unit operations.
The result is a distribution of recovery and grade of materials potentially obtained under the given range of
operating variables, thus defining under which conditions the process design is capable of producing the target
products. We demonstrate that large-scale data analysis can be used to guide decision making and refine the
process towards improving materials recovery targets.
INTRODUCTION
The global push to phase out fossil fuels has drawn the
attention to Li-ion batteries (LIBs) as integral part of
renewable energy grids and transportation (Kittner, Lill,
and Kammen 2017). The demand for reliable energy stor-
age to complement intermittent renewable energy sources
is crucial for facilitating the transition to the electrifica-
tion of systems (Diouf and Pode 2015). In this context,
the increasing production of LIBs in the coming decades
will pose challenges to waste management as they reach
their end-of-life (EoL) (Srivastava et al. 2023). As a result,
adopting a circular economy approach for battery materi-
als becomes essential, and researchers and practitioners are
actively searching technologies to improve recycling pro-
cesses (Gutierrez 2021).
Presently, industrial and pilot-scale LIB recycling typi-
cally involve pyrometallurgical and hydrometallurgical
routes or their combination, often aided by some degree of
sorting and mechanical pre-treatment (Velázquez-Martínez
et al. 2019 Brückner, Frank, and Elwert 2020 Makuza et
al. 2021). Pyrometallurgical routes operate at high tem-
peratures and are thus considered energy intensive, but
they require minimum pretreatment and are flexible on
the composition of batteries used as feed. Processes such as
smelting, roasting, incineration, reduction and pyrolysis are
employed in this route (Makuza et al. 2021). In contrast,
hydrometallurgical extraction is carried out at low tem-
peratures, although they require more complex mechani-
cal pre-treatment processes for discharging, dismantling,
and reducing the particle size of the battery waste mix.
Leaching, precipitation, solvent extraction and chromatog-
raphy ion exchange are some of the typical process units
in hydrometallurgy (Saleem, Joshi, and Bandyopadhyay
2023). Mechanical pre-treatment encompasses a multitude