This thesis introduces a modular variant of dependency-aware fault trees (DaFTs) and shows how dependent events can be handled by tracing their impact throughout all intermediate gates, up to the top-level event. Depending on the set of traced events at a gate, different additional probabilities are calculated to measure the influence of included active events. Logical modeling errors, like the presence of mutually exclusive event pairs, are recognized and the occurrences of dependent and negated events are taken into account during probability calculation. The safety engineer can additionally select any basic event of interest for tracing, making it possible to analyze and visualize the impact of any given event throughout the DaFT. Used in early design stages, the comparison of different modeling variants can be used as design decision base in the presence of specific requirements, like the maximal system impact of single events or a combination of events. For demonstrating the application of DaFTs, several examples of different size and complexity are presented in the evaluation chapter and appendix.weiterlesen