Given the degrading environment, countries across the globe have strongly emphasized the development and deployment of renewable energy and energy storage systems across various sectors such as residential, commercial, industry, rural households, charging stations, etc. Given with this pace of deployment, there will be unforeseen problems in next decade on how to diagnose, perform online monitoring and regular maintenance, and recycle millions of these renewable energy and energy storage systems whose life span is over (spent systems). Main challenges that exist includes the heterogeneity of the systems (different standards, size, and shape) adding complexity to recycling procedures, laboratory-based models having assumptions for online prediction and monitoring, and non-uniform reusability and recycling procedures. Due to this, the process adopted for recycling and reusability of the spent systems is mainly manual. There is a strong need for designing and developing the automation, artificial intelligence (AI), and robust optimization procedures for making these recycling and reusability procedures smart and efficient.
Some of these mentioned problems and challenges are addressed by papers in this Special Section issue, which includes the development of advanced AI methods such as auto-long short term memory deep learning networks, human-augmented Gaussian process regression, genetic programing, kriging, radial basis function surrogates, etc. In addition, parameter identification methods such as extended Kalman filter and unscented Kalman filter are proposed for real-time and online prediction of state of charge (SoC). AI-based methods and parameter identification algorithms are applied on large data set obtained from laboratory, repository resources (NASA), field drive cycles such as dynamic stress tests, federal urban driving schedule, etc. to build battery remaining useful life (RUL), state of health (SoH) (degradation), and SoC prediction models. The models developed based on these algorithms are found to perform better than those developed from the traditional methods (response surface methodology, polynomial regression, etc.). Furthermore, these models are useful during the recycling procedure for efficient and real-time sorting of batteries based on RUL, SoC, and SoH. The function of these estimated indexes (RUL, SoC, and SoH) of batteries shall facilitate expert decisions such as the recovery of materials from cells, reusability of cells for secondary applications such as grid storage, telecom networks, or for remanufacturing. In addition, there are studies on the comparison and development of charging protocols (constant current and constant current-constant voltage) and new anode materials by use of inorganic precursors that provide an accurate measurement of SoC, SoH and enhances battery life, respectively.
In the second phase, for efficient recycling, a lighter weight and higher structural and thermal performance battery is desirable. For this purpose, the papers in this Special Section proposed computational field dynamics simulations for computing thermal and fluid performance, machine learning methods such as support vector regression and neural networks and multiobjective optimization frameworks (non-dominated sorted genetic algorithm II) for minimizing total volume of the battery pack, and increasing its structural and thermal performance (minimizing maximum temperature, standard deviation between cells, total pressure, and power consumption) simultaneously. In the context of reusability of batteries, one application of lead-plated tin bronze as a negative plate for lead-acid battery is illustrated. The performance of the proposed battery was found to be similar to that of new one. The study simplifies the grid recovery and opens a great attractive direction for developing lightweight, high energy technique of lead-acid batteries.
There still exists a huge scope for carrying out further work such as on the reusability of batteries for secondary applications or remanufacturing and development of semi-automatic methods involving digital twin, robotics, and cyber physical systems for industrial battery pack recycling.
We like to express our warm gratitude and thanks to Professor Wilson K. S. Chiu, Editor-in-Chief, and the ASME production team, all contributing authors, reviewers, and our assistants for their efforts that have made this Special Section issue a unique success. We believe that this Special Section issue will be a valuable contribution to energy storage literature and open new frontiers for the researchers.
