Life prediction of batteries for selecting the technically most suitable and cost effective battery

It is necessary to be able to predict the lifetime of a battery in target applications in order to make sound technical and commercial decisions at the system design stage. In general, accurate lifetime prediction requires more than knowledge of ageing processes and the availability of battery models. A concise procedure linking user requirements, operating regimes and operating conditions of batteries to ageing processes and loss of performance has to be used. Quantified end-of-life criteria have to be defined with the details of the application requirements in mind. To verify lifetime prediction models it is necessary to have data of the battery when new and immediately before replacement, results of post mortem analysis and detailed data of the operation. This paper describes a procedure that can be used for lifetime prediction, outlines some of the requirements for a prediction and discusses the principles of battery models and their potential use for lifetime prediction. (c) 2004 Elsevier B.V. All rights reserved.