Experimental investigation of the heat generation rate of a large format pouch-type lithium-ion battery

Abstract

When battery cells are charged and discharged, heat is generated because of chemical reactions and the associated charge transport. The heat should be appropriately rejected to ensure the performance, safety, and lifespan of the lithium-ion battery. In fact, the performance of the cell varies dramatically with temperature, and cycle life can be extended if the operating temperature window is well maintained by a thermal management system (TMS) that properly rejects the heat and controls the temperature of the cell. For optimal design of such a TMS, understanding of the heat source terms, generating, and releasing mechanisms are required. The heat generation during battery operation is largely dominated by irreversible and reversible heat. The irreversible heat is governed by the cell’s internal resistance, while the reversible heat is governed by the entropy coefficient. Hence, to accurately predict the total heat generation rate (HGR), the resistance and entropy coefficient must be measured as a function of State of Charge (SOC) and temperature. An isothermal calorimeter is developed using thermoelectric assemblies (TEAs) to measure the thermal parameters of large format pouch-type lithium-ion batteries at different operating conditions. Entropy coefficient (EC) was experimentally measured with a 109 Ah capacity large format pouch cell with an NMCA cathode and graphite-SiOx composite anode. The hysteresis phenomenon was observed between charging and discharging EC. The operating temperature has a negligible effect on EC over the entire SOC and temperature range considered (0-45oC). Furthermore, the effect of C rate and temperature on HGR is considered. As the C rate increases, the total HGR increases proportionally, and irreversible heat becomes the dominant part. Additionally, as the temperature decreases, the HGR rapidly increases due to the internal resistance increases. The experimentally measured data are compared with the HGR calculated by the thermal model, and the result has a good agreement.