Origins and accommodation of cell variations in Li-ion battery pack modeling

Abstract

Rechargeable battery industry will see significant growth in the use of battery systems for portable devices and power electronics, renewable energy storage, power systems for transportation, and telecom backup power applications. Despite such promising market sentiment, the battery system management remains as a challenging issue to be resolved in order to provide a safe and reliable power and energy storage system. Here we report advancement in the battery management approach by providing a solution to analyze battery performance variations in a lot of batteries produced from the same manufacturing process. A lot of 100 Li-ion cells were analyzed in order to quantify the inherent cell variations associated with cell manufacturing process and test protocol. Both statistical and electrochemical analyses were used to characterize and quantify the capacity variations among the cells along with other parameters that can be readily derived from the test results. Information extracted from a minimal testing of the cells in the lot and more intensive characterizations on a few cells including one as the nominal sample cell allows the establishment of a single cell model (SCM), based on a generic equivalent circuit, with high accuracy in predicting cell performance. The analyses also permit a carefully crafted logic development of how to separate the origins that cause the cell variations in performance. Such separation of the attributes enable a proper tuning of the cell parameters in the model, which allows the accommodation of cell variations in a battery pack model to handle most of the imbalance issues. A careful validation of the SCM to predict performance of any arbitrary cell in the lot with high accuracy was demonstrated. Copyright © 2009 John Wiley & Sons, Ltd.