Fault location in electric power distribution networks is essential to improve the continuity and quality of the power supply. Among the fault location methods, those based on artificial intelligence are less sensitive to noise in the input data and considerably more accurate compared to other methods. However, these methods require a substantial amount of training data. Thus, this study proposes a method to automatically generate a fault database for faults in electric power distribution networks using ATPDraw/ATP software and the Python programming language. The IEEE34 bus system was used to validate the proposed method, resulting in a fault database, made available to the scientific community, with 6700 files containing different types of faults, incidence angles, and fault resistances.