A Nondestructive and Cost-Effective Approach
Editor’s Note: The paper on which this article is based was originally presented at the 2021 IEEE International Symposium on Product Compliance Engineering – Asia (ISPCE-ASIA), held in Taipei, Taiwan in November/December 2021. It is reprinted here with the gracious permission of the IEEE. Copyright 2021, IEEE.
Introduction
Following the growth in the fire incidents of electric vehicles (EVs) and energy storage systems (ESSs) after years of operation, the health monitoring system of EVs and ESSs are still a concerned topic. While it is relatively easy to measure health condition of a cell under static condition, the health condition measurement over cell packed into a system and under operating condition is rather difficult or time consuming with static measurement methodologies.
However, the deterioration of one cell in a series block will reduce the performance of the whole block, and the deterioration will lead to economic concerns like life span depreciation or mileage cost, so it is very important to develop a health monitoring system without the interruption of actual operation and disassembly of battery pack into modules and cells.
Dynamic Behavior Description of Cell Performance in Lab Environment
Considering the importance of cell condition under heavy duty and long-life demand in EV application, IEC/ISO has published performance standard IEC 62660-1:2018 for the cell and ISO 12405-4:2018 for the pack. Both standards emphasize the performance of cell under dynamic charge and discharge behaviors, not only in battery electric vehicles (BEVs) but also hybrid electric vehicles (HEVs).
Those dynamic profiles have taken conditions into consideration, such as:
- Regenerative braking;
- Road driving, accelerating and deacceleration, and at different period; and
- Mode switching from discharge rich at high state of charge (SOC) to charge rich at low SOC.
Testing profile simulating real life operation are created accordingly as can be shown in Figure 1 and Figure 2 from IEC and ISO standards.
Battery Health Condition by Electro-Impedance Spectroscopy
As most research study referenced, EIS was considered as a comprehensive description of battery structures as shown in Figure 3. Typical EIS was conducted by measurement of applying an AC potential to an electrochemical cell and then measuring the current through the cell under various frequency, usually from frequencies as low as 1mHz to as high as 1MHz.
The responsive frequency can be considered as a description of cell electrochemical structure, as there are many layers of materials between electrodes, and the external potential is like a tuning fork set with in different vibration frequencies. Each layer material has different characteristic natural frequency and will be in resonance when the voltage frequency is the same. The amplitude of the characteristic frequency peak can be analogous to the thickness or the mass of the material. When the material is thicker, the response is stronger.
However, a single spectrum at one time for a cell does not constitute any meaning, but when comparing spectrums across cell when the layer thicknesses changed under different operating conditions, the response changes between extremes condition, e.g., SOC 0% to SOC 100%, will help users to estimate the condition under the measurement to the original unused condition.
Performance Measurement Under Operating Conditions
Establishment of Cell Data by Leveraging BMS
As a voltage of single lithium-ion battery or cell is only 3 volts, creating an output at 12 volts, 48 volts, 96 volts or even above for large powers more than 5 KWs without DC/DC voltage transformation technologies will need to combine cells into series blocks. However, as electrochemical cells do have internal resistance differences, and the voltage difference is greater in the end of blocks. To avoid overvoltage of cells in the block, based on the module safety standard requirement like IEC 62133, UL 2594, or UL 2580, each cell in the same series is required to integrate overvoltage prevention mechanisms or monitoring systems, as shown in Figure 4.
Accordingly, most of manufacturers install voltage sensor onto each cell or cell block of the same series and collect the voltage data into the BMS system within the module or pack or send out to downstream storage or analyzing devices like programmable logic controllers (PLC), industrial personal computer (IPC), or cloud computing or storage services by wired communication interface like controller area network (CAN) bus, Modbus, Ethernet, or wireless like WIFI, Bluetooth, or 4G.
Through the data retrieved from BMS system plotted with time, the slope of charge/discharge response or the shape of voltage recovery, can then be applied to the cell EIS database for further analysis.
Though not perfectly controlled like laboratory environment with standardized power supply and electrical loads, the voltage sensor, current sensor, and temperature sensor with power controlling system (PCS), battery charger/discharger, and the motor or actually load, and BMS do constitute a similar testing system also providing continuous and meaningful data.
Establishment State of Health Profiles by EIS-Like Database
As can be imagined, EIS is typically conducted on single cell at laboratory conditions, and a complete scan from 1 mHz to 1MHz will take several hours to complete. It is not economically feasible to conduct EIS over each cell in a pack, and impossible without disassembly and also under operation.
However, each charge and discharge under different potential and time in real operation is like a pulse of EIS scan, and then we can reconstruct an EIS like spectrum after voltage normalization though may not be continuous in frequencies.
By accumulation of spectrum across different cells and time against real behavior or performance, user or manufacturer can define a database of cell health condition, similar to study as shown in Figure 5.
Data Processing and Report Preparation
When the database of SOH for the interested cell or electrochemical structure was established, the voltage, current, SOC, and time information can be retrieved by BMS then send to edge devices like computers or mobile phones, or directly to cloud for comparison. An overall distribution of SOH for cells in the pack can be drawn as visualized in 3D or 2D objects, similarly to the diagrams shown in Figure 6.
Conclusion
The SOC information of a cell is important as the baseline information, but EIS or dynamic behavior information is more comprehensive and important for the description of SOH. Through the data processing retrieved from BMS under real time operation and comparing with the database established with existing SOH profile, the SOH distribution for the cells in the pack can be described and demonstrated without disassemble the battery pack back to cells under dynamic and operating conditions.
The similar approach can be applied to ESS as well as long as BMS data are available by communication interfaces and regardless of battery chemistries.
References
- https://ec.europa.eu/research/participants/documents/downloadPublic?documentIds=080166e5a1622586&appId=PPGMS
- Energies 2012, 5, 138-156; doi:10.3390/en5010138, Standardization Work for BEV and HEV Applications: Critical Appraisal of Recent Traction Battery Documents
- https://en.x-mol.com/paper/article/1399478987489001472
- https://circuitdigest.com/article/battery-management-system-bms-for-electric-vehicles
- https://www.battpedia.com
- https://battpedia.com/battery/battpedia-battery-aging-simulator
