Lithium-ion (Li-ion) batteries have become the energy storage system of choice in an array of applications, e.g. as a starter battery in conventional automobiles, but also, and above all, in electric vehicles, in medical applications, professional tools, mobile robots, and UPS. In all of these cases, the battery’s state of health has a direct effect on the capacity of the overall system. In terms of electric cars, the main selling points, – first and foremost the vehicle range but also good acceleration, depend – on the battery. In safety-relevant applications, such as backup systems or mobile medical applications (e.g. defibrillators), it is essential to know that the battery will supply the required energy when it is actually needed.
Determining a battery’s state of health
Besides its present state of charge (SoC), the real determining factor here is the age of the battery. Based on complex chemical reactions inside the battery, there is a gradual degradation in performance over time and the battery’s state of health (SoH) consequently suffers. The SoH reflects the ratio between the current maximum practical capacity and the theoretical capacity of a battery, i.e. a 100Ah battery with a SoH of 80% has a residual capacity of 80Ah. It is very difficult to determine or to predict how quickly a battery or the individual cells of a battery pack will age. On the one hand, the capacity cannot be measured directly; and on the other, the aging process is influenced by a number of factors, e.g. by the individual condition of the battery, the charging behavior, and the temperature.
Determination of the SoH is however essential to evaluate the battery life. Depending on its actual application, the end of battery life is reached with a SoH of between 70% and 80%. The battery then frequently swaps its ‘first life’ for a ‘second life’, i.e. it is used in an application that demands less capacity. For instance, electric car batteries are used as stationary energy storage systems for PV units during their second life. The remaining maximum capacity of the battery in the respective application is referred to as the remaining useful life (RUL).
Complex procedures offer unreliable predictions
Since it is not possible to simply measure the remaining capacity to determine the SoH and RUL, relatively complex and often inaccurate procedures are in use at present: Before installing the battery, a vast array of data is collected in the lab to characterize the respective battery type. Algorithmic calculations are used to create a lookup table or an empirical model that describes the battery at defined working points and in various applications. The data are saved in the battery management system and the end of battery life is merely predicted by comparison with the stored data. The actual state of the battery in operation is in fact no longer measured. Needless to say, the base data for the battery management system thus remains very inaccurate.
A coulombmeter, which measures the charge flowing in and subtracts the charge flowing out, is often used to determine the capacity. The data is then compared with the model to draw conclusions about the SoH and the RUL. However, even this method provides relatively inaccurate values, i.e. the determined end of battery life may vary considerably from the actual situation.
The result: To ensure the guaranteed battery life, manufacturers have to install more battery cells than necessary in the instrument or the vehicle as a safety buffer. Alternatively, they have to scale down the specified values that depend on the state of the battery, e.g. the vehicle range and the warranty period of the battery when installed in an electric car. In both cases, it means: The battery capacity is not utilized fully.
Fully utilizing batteries
To enhance utilization of the battery significantly, the Professorship for Measurement and Sensor Technology at Chemnitz University of Technology has developed a procedure that allows precise diagnosis of a fully operational battery in just a few minutes. It additionally provides reliable online information about the SoH and RUL of the battery. RUTRONIK is supporting these research activities as part of partnerships for master's and bachelor's theses and by supplying electronic components and development tools. As the official distribution partner and supplier of Samsung SDI Li-ion batteries, RUTRONIK obviously has close links to the battery manufacturer and, as such, is an ideal research partner that ensures indispensable knowledge transfer in all aspects of battery cells and battery management systems.
Accurate values with impedance spectroscopy
The Professorship for Measurement and Sensor Technology develops measuring systems based on impedance spectroscopy. This enables the measurement and assessment of battery internal processes such as charge transfer, electrode degradation, and diffusion. To do so, the battery is excited with varying alternating current supply potential. The resulting battery voltage and the excitation current can be used to calculate the impedance, allowing you to draw conclusions about the state of the battery.
Since impedance with currently used Li-ion cells can be less than 1mOhm, the measuring methods and the applied hardware need to meet special requirements. Due to the extremely low impedance values, but also as a result of low frequencies and a further frequency range, expensive, precise measuring instruments are necessary along with high-performance devices with large memory capacity to generate accurate, dynamic signals. It is for this reason that the method has only been applied in laboratory conditions to date where the process is usually monitored by an engineer.
From the laboratory to embedded measuring system
To also enable the application of impedance spectroscopy in mobile systems, scientists at Chemnitz University of Technology have optimized the methodology for generating the necessary signal to such a degree that a chip with limited memory capacity and relatively small processing power can map the procedure without the need for additional signal generators. The battery itself or energy from another stack is used as the source of power, thus reducing the related hardware costs enormously. Due to the large frequency range, multi-spectral methods had to be applied to cut the measuring time. All calculations can be carried at the same time as the measurement thanks to innovative algorithms. For instance, it was possible to reduce the memory capacity of the controller to less than 500kByte for intermediate storage of the measured data. In addition, the measuring period was shortened to roughly five minutes. This allows you to repeat measurements in defined cycles during operation, e.g. in certain operating conditions. These features also ensure the methodology meets the development requirements for controllers in the automotive sector.
The prototype hardware developed at the Professorship for Measurement and Sensor Technology can be used to diagnose four battery cells simultaneously. However, the hardware can, in principle, be scaled as required to larger systems.
Moreover, the solution meets further requirements of the target applications: It is not only small but also robust and cost effective due to an embedded microcontroller.
The achieved measuring results allow the full utilization of batteries right up to their actual end of life. This gives manufacturers the opportunity to increase the range of their electric cars, extend the warranty period for their batteries, and to design smaller and thus less expensive battery systems – in line with their business model.
Why does a battery age?
The state of health (SoH) of a battery deteriorates over time. This is caused by:
- Calendric aging: The battery ages without being used, simply due to time. This process is influenced, in particular, by the ambient temperature.
Cyclic aging: It depends on the type of use, but above all on the operating cycles, the (dis)charging stroke, the end-of-charge voltage, and the strength of the charging and discharging currents. The possible number of cycles is determined by the type and quality of the rechargeable batteries and the temperature.
Dipl.-Ing.(FH) Andreas Mangler, Director Strategic Marketing & Communications, Rutronik Elektronische Bauelemente GmbH
Prof. Dr.-Ing. Olfa Kanoun, Chair for Measurement and Sensor Technology, Chemnitz University of Technology
Dipl.-Ing. Thomas Günther, Research Associate, Professorship for Measurement and Sensor Technology, Chemnitz University of Technology