Diagnosis of Commercial GIC || LiFePO4 Cells with High Power and High Energy Designs

Abstract

Battery diagnosis can significantly increase the rich of information from the assessment and evaluation of cell performance, providing temporal quantification on cell degradation. Proper diagnostics can generate significant improvements on reliability and safety of a battery system and its functionality. To understand battery degradation behavior of different battery chemistries in a clear context of cause and consequence clearly identified remains very challenging today.Postmortem analyses might reveal the cause of degradation and failure of a battery, but they are not useful to provide temporal resolution for practical battery monitoring, protection, and diagnosis during operation. Without useful diagnostic information, prognosis could lose fidelity in representing the actual situation in cell degradation. In reality, an effective battery diagnosis tool needs two major elements: a reliable method to determine the state of the battery and a high fidelity model to simulate and analyze the complicated but crucial voltage variations as a function of cell operating parameters.A reliable model and simulation method has been developed and presented by us recently [1] that allows emulating a variety of “what if” scenarios to simulate the corresponding charge discharge behavior under various operating and aging conditions. The model employs three known degradation modes: loss of lithium inventory (LLI), loss of active materials (LAM) and reaction kinetics degradation (RKD), which is often detected by impedance changes; to simulation possible degradation mechanisms that might comprise various contributions from each modes.This work presents a case study of the aging behavior of two types of commercial GCI || LiFePO4 (LFP) cells with high energy (HE) and higher power (HP) designs. The capacity variations as a function of cycle number at four rates, C/25, C/5, C/2 and 1C, at 25°C in a cycle aging experiments with the HE cell were used in the analysis to compare with data obtained from the HP cell. Some peculiar aging behavior was observed, which displays a sensitive rate-dependent capacity variation in the process of aging in the HE cell but not in the HP cell.The variations in performance between the two cell designs are explained by the differences in the electrode architecture, in terms of loading ratio and loading matching of the two electrodes; and the extent of reaction in the electrodes under the test protocols. These performance characteristics were simulated and predicted by a model, the ‘Alawa toolbox, as reported in our prior work. The degradation in capacity fade in these two cell designs are further analyzed and simulated using the IC (dQ/dV) techniques and the ‘Alawa toolbox. Five degradation modes were considered in the analyses and simulations to derive useful information for deciphering the underlying mechanism for capacity fade in each cell design. Besides analyzing and comparing the rate-dependent capacity fading and IC (dQ/dV) variations in HE cell degradation, IC peak area was also used as a basis for key information to determine the underpinning modes in capacity fading.It is useful to point out that the use of ‘Alawa toolbox creates unprecedented benefits of this mechanistic approach to derive fading mechanism with quantitative comparisons for validation. The accuracy in the quantification, viability in comparing a large number of scenarios within a short time frame, the flexibility of comprising half-cell behavior into a full cell analysis, the universality of combining multiple degradation modes in the model simulations (for both capacity and voltage variations as fading symptoms) all explicitly illustrate the benefits and merits with this approach to resolve complicated fading behavior in LIBs. This comparison of the HP and HE designs in a typical GIC || LFP cell exemplify this aspect perfectly