Battery Life Prediction Method for Hybrid Power Applications

Batteries in hybrid power applications that include intermittent generators, such as wind turbines, experience very irregular charge and discharge cycles. Because battery life depends on both the depth and rate of discharge (and other factors such as temperature and charging strategy), estimating battery life and optimally sizing batteries for hybrid systems is difficult. Typically, manufacturersgive battery life data, if at all, as cycles to failure versus depth of discharge, where all discharge cycles are assumed to be under conditions of constant temperature, current, and depth of discharge. Use of such information directly can lead to gross errors in battery lifetime estimation under actual operating conditions, which may result in either a higher system cost than necessary or anundersized battery bank prone to early failure. Even so, most current battery life estimation algorithms consider only the effect of depth of discharge on cycle life. This paper discusses a battery life prediction method, based on the work of Symons, which was developed to investigate the effects of two primary determinants of battery life in hybrid power applications: varying depths ofdischarge and varying rates of discharge. A significant feature of the model is that it bases its analysis on battery performance and cycle life data provided by the manufacturer, supplemented by a limited amount of empirical test data, thur eliminating the need for an electrochemical model of the battery. It performs the analysis for the user-prescribed discharge profile consisting of a seriesof discharge events of specified average current and duration. Sample analyses are presented to show how the method can be used to select the most economical battery type and size for a given hybrid power system application. The proposed method represents a work in progress, in the sense that the algorithm has not been validated against battery life test data.