Implementation of a Physics-Based Model for Half-Cell Open-Circuit Potential and Full-Cell Open-Circuit Voltage Estimates: Part II. Processing Full-Cell Data

Abstract

Physics-based models of lithium-ion battery cells require models of electrode open-circuit-potential (OCP) functions, as well as the stoichiometric values that bound the normal operating windows within the electrodes. In Part I of this series, we presented five approaches to estimate electrode OCP relationships, in which the proposed methods require no prior knowledge of physical parameters and can be applied to general cell chemistries. These methods rely in part on a physics-based multiple-species multiple-reaction (MSMR) thermodynamic model to overcome the “missing-data problem” and the “inaccessible-lithium problem”. This paper applies these same five methodologies to estimate the full-cell open-circuit voltage (OCV) function. The results for OCP and OCV are then combined to determine the operating boundaries of the electrodes via a nonlinear optimization algorithm. To assess the utility and generality of the five methods, their performance is validated in simulations and also applied to a commercial cell. It is often preferred to avoid destructive procedures when acquiring cell properties; therefore, we also analyze the possibility of determining models for OCP relations and regressing operating boundaries without actually tearing down a cell and collecting half-cell data. We find that there are limitations to how well we can do this and show that these are due to two distinct “observability problems.”