Emerging Trends on Batteries for EVs
Energy Systems such as renewable energy and energy storage or hybrid energy systems (solar battery, wind battery, battery-charger-grid, etc.), or sustainable energy systems (smart buildings, electric vehicles (EVs), etc.) are the major contributors that are essential for life. Processes involved in four phases such as design, manufacturing, operation & service, and recycling of these energy systems have a greater impact not only on its performance and socio-economic indexes but also on its sustainability.
Among the energy systems, a strong emphasis on research and development has been made on EVs for promoting the commercialization of cleaner transportation. EVs are complex systems that comprise the components such as a battery, motor, inverter, converter, etc. would be housed on a chassis. The battery is one of the most important components because it needs huge resources (raw materials sourced from other countries, advanced fabrication, trained engineers; etc.) for its manufacturing. The battery used in EVs is expensive and its life span is also limited to 3-5 years which is paramount to the importance of its regular monitoring and enhancing its life. In the present scenario, the customers purchasing EVs tend to have range anxiety i.e. how long can its vehicle travel before it needs to be charged, where to charge its vehicle, and how to determine when to replace the battery. Thus, the commercialization of EVs has not gained much pace despite heavy subsidies and policies being framed by governments .
Solutions to this also lie in hands of EVs and battery manufacturers to accelerate the technology developments for the design of advanced batteries. Technology development must be a holistic design. Therefore, research breakthroughs should be achieved across the vertical and also considered from the cell level to module to pack level and finally recycling. At each level, the multidisciplinary fundamentals from thermal, materials, electrical, mechanical, and electrochemical will make the study of the existing issues much more complex. Now, a brief introduction on the key issues in EV battery across the cell level, module level, pack level, and its recycling will be provided respectively in this article.
Battery Cell level:
To reduce the dependence of sourcing materials from other countries, the new advanced chemistry whose materials are needed for cathode/electrolyte/anode/separator, the present generative designs for battery electrodes using topology optimization and shape grammar would be emphasized producing batteries with higher energy density. Efficient computing search and analysis of the materials database including CT scans, electron microscopy images, etc. would be desirable to select the best-optimized materials for the cell. For accurate monitoring like the voltage and temperature of the cell in real-time, a research attempt has also been made to embed the fine thermocouple in the cell . The state estimation algorithms based on AI such as deep learning networks would need to be further improved to be able to capture and predict the dynamics of the cell in real-time. However, these attempts have been limited to laboratory scale and, the practical prospects when these cells are discharged at very dynamic rates would need more efforts by way of thorough studies in the future .
Battery Module level:
It is interesting to note that though the cells are of the same manufacturer brand, same specifications, and involves the same processes during its mass production, yet there is the observance of divergent behavior among the cells in respect to the capacity, voltage, internal resistance, etc. In an ideal scenario, during the manufacturing phase, the intelligent optimal control of the assembly and fabrication processes used for the battery cell must be done to ensure its uniformity and higher performance. But, in a real case, this does not happen. There is always a slight difference in operating conditions in the manufacturing line of the cell .
One method is to use MRI/CT/ scan technology to single out the defective cells by observing the pattern of assembly of each component of the cell such as cathode, anode, and electrolyte . This can perhaps reduce the chances of occurrence of the defective cells into the production batch regardless of ensuring the uniformity of the behavior among the cells in the production batch. Therefore, concerning the enhancement of uniformity degree of behavior among the cells, the intelligent assembly of the cells would be carried out by using advancing AI approaches to form a battery module. The procedure of an intelligent assembly involves cells sorting by the way of data collection such as cell internal resistance, voltage, cyclic voltammetry data, capacity, etc. The cells are then grouped and based on analysis of data undertaken by using AI approaches such as neural networks and support vector machines having similar performance. Perhaps, this approach of an intelligent assembly of cells is also referred to as the pre-design cell balancing approach used during the formation of a battery module. However, over the period during the actual operation (charge-discharge) of the battery module, the cells in the battery module become unbalanced and need online balancing in real-time. Therefore, the battery management system (BMS) like the hardware would be integrated and used in conjunction with the battery module to test its performance parameters such as the current, voltage, and temperature. The parameter values of each of the cells in the module are recorded by BMS to ensure that they are in the given safe operating range. The BMS will be equipped with a cell balancing function based on passive balancing and using a shunt resistor to dissipate the excess voltage from the cell of higher voltage into heat. This method of passive balancing is economical and easy to use. Most of the BMS used in the battery modules will use passive balancing. However, other cell balancing methods would involve active balancing using capacitance or inductance, or transformer. This method requires complex circuit design in BMS and consumes much more time than the passive balancing because this method will make the capacitance need to get charged first to store the energy and then dissipate to another cell until the equilibrium is reached .
One more important function of the BMS includes the online estimation of the state of charge (SoC), state of health (SoH), state of energy, etc. of the cells in the battery module which is essential for prognostics and diagnostics. This is essential parameter had to be determined when charging the battery or replacing the battery. The microprocessor of the BMS will determine SoC/SoH whose items undertaken a series of current, voltage, and temperature. Primarily, the BMS uses the coulomb counting method to measure SoC at a given time for a given cell in the module. However, this method is not accurate as it is based on an integration of current entering the cell over time. As for all of the parameters, the value may not be accurately measured during each time interval, therefore, the error will accumulate with time resulting in incorrect SoC measurement of the battery module . Recently, the trends for SoC/SoH estimation are on the development of models based on the physics (electrochemical), electrical circuit (equivalent circuit models), or data-driven models (deep learning networks, genetic programming, support vector machines, etc.). Another emerging trend is the design of fusion-based models which is a type of combination of physics-based models and data-driven to complement the advantages of each . The primary advantage of data-driven models is that they can analyze a large amount of data in real-time. As for the positive support from the Notion of Digital Twin, recent researches have been proposed for the development of cloud-based BMS that can store the large data in real-time, and analysis can be done through the advanced approaches to provide the feedback (information such as SoC/SoH) to all of the electric car driver and EV controller when performing the necessary tasks with a safe operating range and longer battery life .
Besides, cell balancing, estimation of SoC/SoH, and temperature sensing would also be very important aspects. When the battery module is operated at higher temperature conditions, or it is overcharged, the cells in the module may get overheated which may result in a progressive breakdown of anode and cathode leading to the thermal runaway. Therefore, the BMS function is to sense the temperature of each cell in real-time and perform the cut-off of the specific cell from the circuit or cut-off the entire circuit itself. Highly advanced batteries include BMS integrated with thermal management system (air/liquid/phase change materials) which provides the SoC, maximum temperature values, and temperature standard deviation values of the battery module to the cooling system in real-time. In this way, the cooling system dynamically adjusts the velocity and temperature of the fluid (air/liquid) at the inlet that minimizes the maximum temperature of the cells below the threshold limit and also minimizes the temperature deviation among the cells in the battery module leading to the uniform behavior . Although there is also energy consumption in the battery module due to the usage of a cooling fan/pump for liquid flow, experts have found ways to optimize these objectives including energy consumption simultaneously.
Battery Pack Level:
The next important aspect to be considered is the battery pack level. The battery pack comprises several battery modules in a series-parallel configuration designed to meet the required power rating of the vehicle. The most crucial research point of the battery pack is its efficient energy management. Every battery module has its own master BMS which will communicate with BMS of other battery modules on the available SoC, SoH, and maximum available energy. Therefore, a controller would be needed to determine the switching conditions of the battery module with remaining to be connected to meet the required power.
Battery Module/Pack Recycling:
The battery module comprises the cells in series-parallel configurations, a complex wiring harness system, cooling system, BMS, etc. The battery module needs to be recycled after that its life span is over. However, it is found that there could be still a significant number of cells in the module whose life span is not over and can be reused. For the remaining cells, whose life span is over, the materials such as cobalt, nickel, and manganese can be recovered. The recycling process is a complex one and involves (a) procedure of disassembly of module casing (b) separation of wiring and BMS, residual energy prediction in the battery, etc. (c) Sorting of the cells in the module for reusability (second life; hybrid energy storage), recovery of materials, etc.
In the present scenario, the process of recycling is mainly manual and involves numerous chemical processes for the recovery of materials which could be risky and increases the chances of exposure to toxic chemicals within the battery. Also, given the pace of deployment of 2 Wheelers, 3 Wheelers, and 4 Wheelers in the market, there could be million tons of battery modules in the market by 2027 whose life span may be over. If the recycling procedure is not an efficient one, there could be endangering to the soil, human lives, and environment .
At present, not many experts are focusing on revolutionizing the existing recycling procedure with an automated/semi-automated one. This implies that all of the designed robotic arms holding machine tools would be required for dismantling of the battery casing, and other physical components separation from the cells. An intelligent camera vision interfaces with a temperature infrared machine to capture the cell images and their characteristics (temperature, voltage) to determine the residual energy of the cells in the module. An optimized procedure for such automated/semi-automated recycling is the robot-human interaction for different stages of recycling of battery modules. However, there are numerous challenges for designing an automated/semi-automated procedure for battery module recycling because that the battery modules with different shapes/sizes and the cells with different manufacturers should be considered at the same time. That is to say, the internal physical structure of the cells configuration, BMS, wiring systems, etc. will be different for the different manufacturers. No fixed standards are being used for the battery modules by the different manufacturers. In addition, there are hardly any concrete market players who are willing to take responsibility for battery module recycling. Neither the battery manufacturer nor the EV manufacturer and nor the consumers. In this scenario, the government may have to introduce incentive-based policies to encourage battery manufacturers, EV manufacturers, and consumers to play an equal role in recycling these battery packs.
As for considering the emerging issues when battery packs were used in EVs across different levels from cell level to module to pack and to recycling. Processes involved in the four phases of these systems are highly complex, interactive, and multidimensional, therefore, the challenges arising in each of the four-phase can be addressed by an application, development, and deployment of Digital Twin Technology. Digital Twins is an integral part of Industry 4.0, enabling digitization across various systems. This technology is an integration of emerging technologies such as artificial intelligence, the Internet of Things, big data, cloud computing, blockchain, virtual reality technologies, collaborative platforms, open standards, etc. These underlying technologies allow a digital twin to be continuously fed by real-world data so that it can effectively and efficiently offer real-time insights about the energy system’s present and future performance and problems. The key information from each of the analyses can be shared among the cell designers and cell manufacturers, EV designers and manufacturers, EV consumers, and the recycling industries. In summary, Digital Twins technologies can efficiently be useful for the entire life cycle management of batteries from design to end-of-life use.
The following are hot emerging trends in Digital Twin applications that could play a key role in a strategic transformation of battery systems. Researchers and experts at various academic and industry levels are encouraged to adopt such areas.
(a)Database development and its integration for modeling and design of batteries
(b)Application of technologies such as 5G communication, edge computing, cloud computing, and artificial intelligence in EVs/battery monitoring.
(c)Big data analytics, modeling, and simulation for digital twins (physics-based, ML/AI, etc.).
(d)Topology optimization and shape grammar for the sustainable and generative design of battery electrodes.
(e)Automation including robotics and AI for recycling of EVs batteries.
I would like to convey my special thanks to Prof. Chin-Tsan Wang, who is currently working as a Director of Science and Technology Division at TECC, India for his valuable feedback and suggestions for improving this article.
. Zeng, Xiaoqiao, Matthew Li, Deia Abd El‐Hady, Wael Alshitari, Abdullah S. Al‐Bogami, Jun Lu, and Khalil Amine. "Commercialization of lithium battery technologies for electric vehicles." Advanced Energy Materials 9, no. 27 (2019): 1900161.
 Li, Z., Zhang, J., Wu, B., Huang, J., Nie, Z., Sun, Y., ... & Wu, N. (2013). Examining temporal and spatial variations of internal temperature in large-format laminated battery with embedded thermocouples. Journal of Power Sources, 241, 536-553.
 Cannarella, J., & Arnold, C. B. (2014). Stress evolution and capacity fade in constrained lithium-ion pouch cells. Journal of Power Sources, 245, 745-751.
 Diao, W., Xue, N., Bhattacharjee, V., Jiang, J., Karabasoglu, O., & Pecht, M. (2018). Active battery cell equalization is based on residual available energy maximization. Applied energy, 210, 690-698.
 Zhu, J., Li, W., Wierzbicki, T., Xia, Y., & Harding, J. (2019). Deformation and failure of lithium-ion batteries treated as a discrete layered structure. International Journal of Plasticity, 121, 293-311.
 Uzair, M., Abbas, G., & Hosain, S. (2021). Characteristics of Battery Management Systems of Electric Vehicles with Consideration of the Active and Passive Cell Balancing Process. World Electric Vehicle Journal, 12(3), 120.
 Naguib, M., Kollmeyer, P., & Emadi, A. (2021). Lithium-Ion Battery Pack Robust State of Charge Estimation, Cell Inconsistency, and Balancing. IEEE Access, 9, 50570-50582.
 Liu, Q., Kang, Y., Qu, S., Duan, B., Wen, F., & Zhang, C. (2020, July). An Online SOH Estimation Method Based on the Fusion of Improved ICA and LSTM. In 2020 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia) (pp. 1163-1167). IEEE.
 Garg, A., & Panigrahi, B. K. Multi‐dimensional digital twin of energy storage system for electric vehicles: A brief review. Energy Storage, e242 DOI: 10.1002/est2.242.
 Jishnu, A. K., Garg, A., Shaosen, S., Su, Y., & Panigrahi, B. K. (2021). A novel procedure combining computational fluid dynamics and evolutionary approach to minimize parasitic power loss in air cooling of Li‐ion battery for thermal management system design. Energy Storage, 3(1), e210.
 Yun, L., Linh, D., Shui, L., Peng, X., Garg, A., Le, M. L. P., ... & Sandoval, J. (2018). Metallurgical and mechanical methods for recycling lithium-ion battery packs for electric vehicles. Resources, Conservation and Recycling, 136, 198-208.
© 2021 Ministry of Science and Technology, Taiwan. All Rights Reserved.