Author: Dipl.-Ing. Norbert Sailer, Infineon Technologies Austria AG
Key Findings for Traction Applications
Accurate battery temperature monitoring is critical for electric vehicle safety and performance. Alongside individual cell voltage and battery pack current measurements, temperature monitoring is a fundamental task of battery management systems (BMS). While surface-mounted temperature sensors are standard in today’s BMS, they cannot directly measure internal cell temperature. Instead, they rely on electrothermal models to estimate it—a limitation that becomes critical during fast charging, high discharge rates, or abnormal operating conditions. Temperature gradients between the cell surface and interior increase with higher currents and lower ambient temperatures, making surface measurements potentially misleading precisely when accurate data matters most.
Why Internal Temperature Matters
Internal cell temperature provides a more reliable indicator of thermal stress and potential thermal runaway risk compared to surface measurements. Accurate internal temperature estimation can:
- Optimize fast charging protocols
- Extend battery cycle life
- Enable earlier detection of safety-critical operating conditions before they escalate
EIS as a Solution
Electrochemical Impedance Spectroscopy (EIS) offers a promising non-invasive method for estimating internal battery temperature. By measuring impedance characteristics at specific frequencies, the mean internal cell temperature can be estimated without requiring sensors inside the cell.
Key Research Findings
Through a comprehensive meta-analysis of 68 publications covering 83 different cells, combined with our own experimental data, our research identifies impedance phase measured between 100 Hz and 1 kHz as the optimal universal estimator for internal cell temperature across various cell formats and chemistries.
Critical findings include:
- Temperature sensitivity: -0.35°/K for temperatures below 23°C and -0.17°/K above 23°C
- Universal applicability: These sensitivities remain remarkably consistent regardless of cell capacity or chemistry
- Measurement precision requirements: To achieve sub-1 K temperature uncertainty in large-format cells above 23°C, EIS systems must maintain measurement uncertainty below 4 μΩ for both real and imaginary impedance components
Practical Challenges in Traction Applications
Real-world implementation faces significant challenges from electromagnetic noise generated by traction inverters:
- Below 3 kHz: Motor control signals create high-power, speed-dependent noise
- Above 5 kHz: Pulse–Width Modulation (PWM) frequencies generate continuous interference
Figure 1 illustrates these challenges using battery pack current data from a Citroën C-Zero during a dynamic test drive. Positive currents indicate battery discharge (acceleration), while negative currents represent charging (recuperation). Discharge currents reached up to 3.8 C, and recuperation currents peaked at 1.6 C.
The Short-Time Fourier Transform (STFT) in the lower panel reveals:
- PWM switching noise at approximately 7 kHz, present throughout the entire drive
- Motor control signals below 3 kHz, with both frequency and magnitude increasing during acceleration as motor speed rises, and decreasing during deceleration
Both noise sources can disrupt EIS measurements, requiring careful frequency selection and outlier detection strategies. Simply measuring at higher frequencies to avoid noise isn’t practical, as temperature sensitivity decreases with increasing frequency.
Implications for BMS Design
These findings establish quantitative design criteria for battery management systems:
- Target frequency range: 100 Hz – 1 kHz for optimal temperature sensitivity
- Required precision: Sub-4 μΩ measurement uncertainty for both real and imaginary impedance
By accounting for both electrochemical cell properties and application-specific system characteristics, engineers can implement robust online internal temperature estimation that enhances battery safety, optimizes performance, and extends lifetime.
Read the full paper: https://doi.org/10.1016/j.jpowsour.2025.239111
Infineon Technologies Austria AG is a subsidiary of Infineon Technologies AG, a global semiconductor leader in power systems and IoT. Semiconductors are essential for mastering the energy-related challenges of our time and helping to shape the digital transformation. Infineon’s microelectronics drive decarbonization and digitalization and enable groundbreaking solutions for green and efficient energy, clean and safe mobility as well as a smart and secure IoT.
Infineon Austria pools competencies for research and development, production as well as global business responsibility. It is the strongest research company in the country. The head office is in Villach, with further branches in Graz, Klagenfurt, Linz, Innsbruck and Vienna.
Infineon Technologies AG is a global semiconductor leader in power systems and IoT. Infineon drives decarbonization and digitalization with its products and solutions. The Company had around 57,000 employees worldwide (end of September 2025) and generated revenue of about €14.7 billion in the 2025 fiscal year (ending 30 September). Infineon is listed on the Frankfurt Stock Exchange and in the USA on the OTCQX International over-the-counter market.
The EU funded project NEMO aims to deliver next-generation BMS to extend battery life and enhance system safety for automotive and stationary applications over the long term. https://nemoproject.eu
References
- N. Sailer, C. Steffan and J. P. Schmidt, “Practical considerations and limitations of online EIS-based battery internal temperature estimation in traction applications,” Journal of Power Sources, vol. 667, p. 239111, 2026