Calendar ageing analysis of a LiFePO4/graphite cell with dynamic model validations: Towards. realistic lifetime predictions

The present study aims at establishing a methodology for a comprehensive calendar ageing predictive model development, focusing specially on validation procedures. A LFP-based Li-ion cell performance degradation was analysed under different temperature and SOC storage conditions. Five static calendar ageing conditions were used for understanding the ageing trends and modelling the dominant ageing phenomena (SEI growth and the resulting loss of active lithium). The validation process included an additional test under other constant operating conditions (static validation) and other four tests under non constant impact factors operating schemes within the same experiment (dynamic validation), in response to battery stress conditions in real applications. Model predictions are in good agreement with experimental results as the residuals are always below 1% for experiments run for 300-650 days. The model is able to predict dynamic behaviour close to real operating conditions and the level of accuracy corresponds to a root-mean-square error of 0.93%. (C) 2014 Elsevier B.V. All rights reserved.