(63j) Identification of Dominant Mechanisms for Capacity Fade of Lithium-Ion Batteries

The life-cycle cost of a battery depends on lifetime. The literature reports numerous possible mechanisms for capacity fade in lithium-ion batteries. Mathematical models including these phenomena are few and do not include all postulated mechanisms. Developing such a mathematical model is challenging due to (1) unknown model parameters, (2) cumulative non-separable effects of individual mechanisms occurring simultaneously, and (3) lack of computationally efficient solver schemes. Electrochemical engineering models will be reviewed with a step-by-step description of what each possible mechanism implies in terms of model formulation and numerical simulation. Due to the combination of multiple mechanisms possible for capacity fade, the data available (discharge curves at different cycles) may not be sufficiently informative to identify the dominant mechanisms in a particular operating scenario just by comparing fits of data to simulation results. To address this, Bayesian estimation, uncertainty quantification, and model discrimination techniques will be used to analyze mechanisms for capacity fade. The use of reformulated models for discharge curves facilitates the application of these systems techniques, which employ Markov Chain Monte Carlo simulation and polynomial chaos expansions. The computational cost and the identifiability of the model parameters will be compared for the inclusion of each mechanism.