Solid-state batteries provide an excellent opportunity for innovation, thanks to their ability to combine higher energy density, greater power, and increased safety, but require new materials and processes. This report will focus on Saft’s position in the technological development of all-solid-state batteries (ASSB) and our progress in working with sulfide electrolytes. The topics of discussion will include screening for sulfide and silicon materials, creation of protective layers, as well as areas of growth for prototype development.

We will examine custom battery packs engineered for racing vehicles and explore how these innovations are being adapted for use in the aerospace industry, particularly in electric vehicle vertical takeoff and landing (eVTOL) aircraft. This discussion will cover their implications for safety, battery pack design, and the circular battery pack recycling system.

The proposal highlights the use of advanced AI and virtual testing methods to enhance battery performance and quality for automotive applications in much shorter time. By integrating state-of-the-art simulation tools and virtualization solutions, this approach allows for rapid, cost-effective development and optimization of battery systems. Utilizing predictive modeling and iterative design optimization, engineers can explore diverse scenarios, accelerating the development process and significantly improving quality and cycle-life of Li-ion batteries.

The diagnosis and prognosis of deployed batteries is complex because the cells might never experience controlled conditions during operation as both the charge and discharge duty cycles could be sporadic. To circumvent this issue new methodologies must be implemented and thoroughly validated. This work presents a new methodology for diagnosis that used real observed solar irradiance, modeled clear sky irradiance, a load usage model, and synthetically-generated battery data from a battery digital twin to diagnose the degradation of commercial Li-ion batteries connected to photovoltaic systems.

In the energy storage systems containing multiple lithium-ion cells, thermal runaway may propagate from cell to cell and grow into a large fire or an explosion. Several module-level migration strategies have been tested in this work to determine their effectiveness. These strategies included introduction of air gaps and solid physical barriers between cell clusters as well as injection of fire suppression agents such as Novec 1230 and water mist. Certain barriers were found to be effective in slowing down the thermal runaway propagation. The suppression agents were found to be effective only at the concentrations and application rates significantly higher than those recommended for traditional fires

Current method of useful life is based on the SoH monitoring of LiB and perform various methods to estimate from the trend of SoH degradation, and usually these LiB cells are already connected. This talk presents a method where the estimation is done on the first charge-discharge cycle of LiB cell, allowing for good selection of cells for pack building. This method is verified experimentally using commerical LiB cells of different rating from different suppliers. It can also help to spot used cells that may be sold as fresh cell. It can also apply to connected cells during their operations.

This talk presents low data machine learning models for capacity degradation trajectory and EOL of commercial lithium-ion batteries (LIBs) using the data from the initial historical cycles, employing a recently developed machine learning approach based on a capacity degradation network (CD-Net). Additionally, the machine learning models are employed to predict the knee point in the degradation curve showing applicability to accident scenario. Finally, data is shown in real world operations.

This talk will provide a high-level exploration of some of the key challenges and opportunities in the model-based control of both lithium-ion and lithium-sulfur batteries, including both solid and liquid electrolyte Li-S batteries. Much of the talk will focus on emerging opportunities for test trajectory optimization and machine learning for both battery types, with a focus on Li-S batteries.

Lithium-metal batteries hold promise for heavy-duty transportation due to their high energy density, but before they can be adopted, battery-management systems (BMS) must be developed to monitor and control their operation. This talk introduces a method to parameterize a physics-based model of these cells for BMS application that uses simple tests performed on electrochemical-impedance-spectroscopy (EIS) equipment.

State-of-the-art BMS rely on accurate battery models and specialized algorithms to obtain useful estimates of the battery state in order to ensure proper performance and safe operation. Most practical models are simplifications, and thus must trade off high accuracy for computational efficiency. This talk examines the implications of using various simplified models in the performance of key BMS tasks.

This presentation covers understanding lithium-ion battery degradation, how to model it, and how close those models are getting to usefully predict lifetime. We will describe our efforts to model lithium plating, SEI layer growth, positive electrode (cathode) decomposition, unequal degradation in silicon carbon composite electrodes, particle cracking, electrolyte consumption and cell dry-out, and how multiple degradation mechanisms are coupled with each other and contribute towards accelerated degradation (the knee point/cliff-edge/etc).