Temperature dependency of diagnostic methods in lithium-ion batteries

Abstract

Accurate estimation of battery performance and capacity is essential for assessing the capabilities of different batteries and determining state of health (SOH). Typically, diagnostic methods such as constant current charge/discharge, current interrupt, and electrochemical impedance spectroscopy (EIS) are demonstrated in laboratory tests with controlled environmental conditions. In real-world applications, such as electric vehicles, batteries are subjected to wide temperature ranges and temperature control for performance and SOH estimation is often not possible. This study investigates how battery diagnostic techniques are influenced by small temperature variations in order to determine suitable temperature windows over which different diagnostic methods can be performed without negatively influencing results. A lithium-ion 18650 cell is subjected to constant current charge, pulse discharge and EIS between 0 °C and 40 °C in 2 °C increments. The influence of variations in operating temperature on the diagnostic abilities of incremental capacity analysis (ICA), differential voltage analysis (DVA) and equivalent circuit fitting is explored. The findings of this study indicate that small temperature variations between diagnostic tests can lead to significant errors in the ability to diagnose degradation mechanisms, particularly at low temperatures (<10 °C). If temperature variations of ±4 °C or greater are expected between characterisation tests, then temperature regulation will be required to ensure accurate diagnostics. EIS based diagnostics was shown to be less temperature sensitive than voltage-based analysis techniques such as ICA.