In this paper, applications of Soft-Computing paradigms i.e. Expert System, Neural Network, Fuzzy Logic and Genetic Algorithm (GA) Fault Diagnosis in power substations are discussed. Advantages of each paradigm are studied in a case study. Fault Diagnosis in a substation with “Double bus, one and half breakers” structure in Fars Regional Electric Company (FREC) has been chosen as the case study. A fuzzy logic expert system is proposed for fault diagnosis in this substation which can be used to aid operators in detecting the main failed section of the substation. Simulation results show the effectiveness of the proposed method.