Abstarct: Differential protection is one of the most widely used protections in power transformers. One of the challenges associated with this scheme is distinguishing the inrush currents and overexcitation, and external faults from other internal disturbances. So far, various methods have been proposed to prevent the differential relay from mal-operation in the mentioned conditions, which are not necessarily simple to implement or do not guarantee the proper performance in all scenarios. This thesis proposes a new method baxsed on the discrete energy separation algorithm to detect the inrush currents, overexcitation, and external faults from other internal and winding faults. The proposed method consists of two parts. The first part detects the external faults from internal disturbances, and the second part distinguishes the inrush currents and overexcitation from internal and winding faults. One of the advantages of this method is the parallel employment of these two parts in the logic of the method, which reduces the detection time. Also, several advantages, such as the high detection speed (1.5 ms), simple implementation, and low computational burden, make it possible to implement the proposed method in real-time. In order to evaluate the proposed method, the types of disturbances are simulated in EMTP software and analyzed in MATLAB software by considering the saturation of current transformers and noise conditions. The obtained results indicate the efficiency of the proposed method in detecting inrush current, overexcitation, and external faults from internal and winding faults, even in noisy conditions and saturation of the current transformer. Also, according to the obtained results, the proposed method has better speed, and accuracy than the harmonic restrain method, which is a common and widely used method in the industry.