Within the framework of the initiative “Energy System 2050” and “KASTEL”, model-based methods for detecting cyber attacks on energy system components are being developed. The newly developed methods make use of physical inconsistencies of all system data, which result from the manipulation of individual components of the energy system (e.g. by manipulated sensors or communication paths in the process), to detect and localize cyber attacks. For this purpose, existing data reconciliation methods are used and further developed for the detection of cyber attacks.