Lifetime prediction of electrochemical systems obviously requires a detailed understanding of aging processes and their causes. However, lifetime prediction on the basis of such detailed understanding is only possible in the few applications where one aging process dominates and where test procedures and methods are available, which allow investigation of this dominant aging process without the influence of other aging processes. | Lifetime Prediction DU Sauer RWTH Aachen University Aachen Germany H Wenzl Clausthal University of Technology and Beratung für Batterien und Energietechnik Osterode Germany 2009 Elsevier . All rights reserved. Introduction Lifetime prediction of electrochemical systems obviously requires a detailed understanding of aging processes and their causes. However lifetime prediction on the basis of such detailed understanding is only possible in the few applications where one aging process dominates and where test procedures and methods are available which allow investigation of this dominant aging process without the influence of other aging processes. A typical application where this is true is uninterruptible power supply systems which are operated at float charge conditions. For lead-acid batteries corrosion and dry-out are the major aging effects in this traditional application. However where a number of aging processes take place in parallel due to complex operation conditions for batteries . a combination of cycling partial state-of-charge PSoC cycling rest periods at different states of charge SoC incomplete or rare full charging and for fuel cells . start-stop operation high mass utilization and low-power operation at high voltage or operation in a wide range of temperatures the complex interaction between the various aging processes and the operating conditions must be analyzed. For mobile applications additional requirements and combinations of stress factors such as vibration may occur. Again it is necessary to understand if different aging processes can be analyzed in experiments one by one followed by a superpositioning of the effects or if aging effects interact with each other in a nonlinear way. This would . require vibration tests and electrical cycling of thermal stress test in parallel. This increases the number of parameter combinations dramatically. In any case it is very difficult to develop appropriate laboratory aging tests to analyze
