Control co-design and life cycle assessment of battery energy storage systems

This dissertation addresses critical challenges in battery energy storage systems through comprehensive studies on control co-design and life cycle assessment. Experimental analyses on cylindrical and pouch battery cells employing various cooling methods are used to develop high-fidelity finite element models for battery thermal management systems. Gaussian process-based surrogate models are implemented to reduce the computational demands of these simulations, significantly improving optimization efficiency. A control co-design approach is introduced to jointly optimize plant and control design parameters, reducing cooling system energy consumption while maintaining thermal safety. To handle uncertainties in real-world scenarios, a reliability-based design optimization framework integrated with control co-design is proposed, achieving a 90% reduction in cooling system energy consumption while ensuring robust performance. The work further develops a generative AI framework for battery pack layout design under geometric constraints and a multi-fidelity physics-informed convolutional neural network for efficient temperature field prediction. Finally, the dissertation conducts life cycle assessment studies of hydrometallurgical and direct recycling processes for cathode active materials and evaluates environmentally favorable solid-state battery manufacturing pathways, highlighting strategies that reduce environmental impact by more than 50% and support sustainable battery design and manufacturing.

PhD thesis, University of Illinois Urbana-Champaign