In the recent years large storage systems enter the limelight due to the growing storage demand in the field of regenerative energies and the electrification in the vehicle sector. This results in the need for considerably longer lifetimes as well as high reliability and availability of lithium-ion batteries in comparison to consumer applications. The main objective of this thesis is to address aging on system level taking the cell-to-cell variation as well as the influence of boundary conditions into account. A vast experiment on module level is carried out, applying varying topologies and temperature gradients. Moreover, the impact of failures in a parallel connection is investigated. Based on the measurement results, a system model is developed to estimate the lifetime expectancy of different battery system architectures. The boolean model is based on distribution functions for the individual components, using a mixed Weibull distribution for the lithium-ion cells. Strategies to optimize the battery system for different electric vehicle concepts are given by the simulation results.