The Fast Fourier Transform (FFT) was used to tackle the challenge of identifying faults in a distribution system in this study. By lowering the time it takes to detect and clear short circuit faults in the network, our aim is to enhance the reliability and quality of the power supply. To handle the problem of fault detection and type, the approach used FFT. Practical data from the Yirgalem distribution system in Ethiopia is used and modeled in MATLAB/Simulink for this purpose. MATLAB was used to simulate several sorts of fault circumstances on the model network in order to create fault condition data, which was then utilized to train FFT. Six inputs, three phase voltages and their corresponding phase angles are used in this fault detector. When a fault occurs on single phase to ground, the voltage recorded at the main terminals of substation transformers drops by 50% and the phase angle for the afflicted phase shifts by 180 degrees. The afflicted phases of a two-phase to ground fault had the same voltage and phase angle as the healthy phase, but no voltage was recorded during a three-phase to ground fault. Faults were properly recognized, and the kind of fault was determined.