<p>Power systems are essential for sustaining modern economies and advancing technology. The reliability of these systems influences all aspects of daily life, including household operations, critical communications, and emergency services. Power supply involves processes, including transmission and distribution through multiple power transfer stations. Consequently, a power system can be modeled as a Stochastic Flow Power Network (SFPN), where each link operates with stochastic capacities. The transmission efficiency for the end user is significantly affected by the loss rate in each link. This paper explores network reliability for SFPN, which is defined as the probability that the power system can transmit sufficient demand to satisfy the end user under loss rates. To analyze the SFPN model under loss rates, the power system is segmented into two or more subnetworks around a series of power transfer stations. To reduce the complexity of flow generation under loss rates, a concept of necessary demand is proposed to obtain flow and capacity vectors for each subnetwork efficiently. An algorithm is developed to calculate subnetwork reliability, determining the overall network reliability for the SFPN. This analysis gives managers a clearer understanding of practical power transmission performance and enables them to prioritize improvements in less reliable subnetworks to enhance overall system performance.</p>