A lithium-iron-phosphate (LFP) battery is one of the ideal energy storage devices owing to its outstanding characteristics such as high power output, no memory effect, and slow degradation. While electrochemical models have been widely used to simulate the LFP battery behaviors, we found the accuracy of simulated LFP battery characteristics, including the cell voltage, lithium concentration distribution, and cell capacity, are highly related to the initial concentrations, namely, the state of charge (SOC) of each individual electrodes.

The amount of active lithium in an LFP battery, which determines the cell capacity, is a result of the SOCs of the anode and the cathode at the beginning of charge or the end of discharge processes. The SOC also dictates the equilibrium potential of individual electrodes, which in turn contributes to the full cell potential. However, it is difficult to estimate the amount of lithium in each individual electrode after the activation cycles because a certain amount of lithium is consumed to form the solid electrolyte interface (SEI) layer. Incorporation of inaccurate SOCs in models could lead to erroneous cell potential and capacity estimation. In this study, the initial concentrations of the anode and the cathode are estimated from the experimentally measured equilibrium potentials. By evaluating the characteristic points of individual electrodes and the corresponding reference points of a full cell, on the basis of low-rate discharge potential curves, the SOC of each individual electrode can be accurately calculated.

In order to evaluate the discharge performance of the LFP battery and gain insight into the effects of electrode SOCs, an electrochemical model based on coupled mass balance, charge balance, and reaction kinetics constitution laws is developed. To determine the initial SOCs of individual electrodes of the LFP battery, equilibrium potentials are experimentally measured by low-rate discharge while the characteristic points are calculated by identifying critical points of the discharge curves. The LFP battery model with such formally determined SOCs is validated by comparing the simulated discharge curves with experimental measurements under various loading demands, up to 2C. The simulated voltage plateau is closely matching the corresponding experimental results for a range of SOCs as shown in Figure 1. Overestimated cell voltage near the earlier stage of discharge and underestimated cell capacity at all C rates are observed in our simulation results, which could be improved by precise measurements of equilibrium potentials of the electrodes and the full cell. Further understanding of the effects of lithium loss during activation process with experiments and model simulation can assist the LFP battery design and optimize fabrication.