Practical considerations and limitations of online EIS-based battery internal temperature estimation in traction applications

Based on a meta–analysis of 68 publications comprising 83 cells, extended by our own measurements, we propose the impedance phase, evaluated at frequencies between 100 Hz and 1 kHz, as an optimal estimator for internal cell temperature. Our findings indicate mean temperature sensitivities of -0.35 ° K -1 for temperatures below 23 °C and -0.17 ° K -1 for temperatures above 23 °C. Notably, these sensitivities remain constant regardless of the cell’s capacity and chemistry. To achieve a temperature uncertainty of less than 1 K for high-energy cells above 23 °C, an Electrochemical Impedance Spectroscopy (EIS) measurement system must maintain an uncertainty smaller than 4 µΩ for both the real and imaginary components of the impedance, based on the mean sensitivities. In addition to the electrochemical properties of the cells, system characteristics of the application must also be considered for online temperature estimation. Our proposed noise model for traction applications in general, and measurement results for an Electric Vehicle (EV) specifically, demonstrate that disturbances with significant power may disturb online EIS measurements up to 3 kHz. • Meta–analysis of 68 publications comprising 83 cells. • Impedance phase effectively estimates internal cell temperature. • Optimal frequency range: 100 Hz to 1 kHz. • Mean sensitivity of -0.17 ° K -1 for T > 23 ∘ C . • EV traction inverter noise might impact EIS up to 3 kHz.