FIRAOL KASAHUN MENGESHA2026-02-022024-04-12https://etd.hu.edu.et/handle/123456789/489The distribution system connects high-voltage transmission networks with end-users. Most of the time, power plants are situated distant from the consumer's location, resulting in large power losses in both the distribution and transmission systems, However, distribution system losses are typically greater than transmission line side losses. The main objective of this study is to reduce power losses and enhance system reliability using Distributed Generation (DG) in the case of the Guder Substation. The Guder Substation has three feeder lines that provide energy for different customers. From these feeders, the Guder town feeder has been chosen since it is frequently interrupted. The chosen feeder has been modeled in ETAP software, and simulation results have been obtained with both ETAP and MATLAB software. The results show that the feeder has a power loss of 611.9843 KW and 323.8237 kVar active and reactive, respectively. Additionally, the study investigates the existing reliability indices of SAIFI, SAIDI, and EENS, which have values of 303.7458 f/cust.yr, 306.4240 hr/cust.yr, and 2368.307 MWhr/yr, respectively. Particle Swarm Optimization algorithm has been suggested to decide the best size and position of DG. After renewable Distributed Generation penetrated the network, the real and reactive power loss reduced from 611.9843 KW and 323.8237 kVar to 302.75 KW and 132.34 kVar, respectively. Additionally, the SAIFI, SAIDI, and EENS system reliability indices were enhanced from 303.7458 f/cust.yr, 306.4240 hr/cust.yr, and 2368.307 MWhr/yr to 27.4968 f/cust.yr, 13.650 hr/cust.yr, and 111.758 MWh/yr, respectively. Finally, reliability indices and line losses before and after Distributed Generations penetrated the network are compared. In general, the simulation results indicate that the suggested method is efficient in maintaining system reliability and minimizing power lossesenDistributed GenerationETAPMATLABPSOReliability enhancementThesis