Optimal Feeder Placement Using Crow Search Algorithm and Its Impact on Voltage Profile Improvement and Loss Reduction

Abstract

Power distribution systems often experience technical challenges such as power losses and voltage drops, particularly in radial networks with long feeder lengths and high load growth. These issues directly affect system efficiency, reliability, and compliance with operational standards. In the 20 kV distribution network of PT PLN (Persero) ULP Mantingan, the minimum voltage during peak load conditions reaches 16.1 kV (0.805 p.u.), while power losses reach 15.12%, which is significantly higher than the national target of 5.7% in 2024. Furthermore, system-wide analysis shows that technical losses can reach 17.51%, accompanied by a large number of buses experiencing undervoltage conditions, indicating a significant degradation in system performance. This study proposes an optimization-based approach using the Backward-Forward Sweep (BFS) method for load flow analysis and the Crow Search Algorithm (CSA) as the optimization technique. The objective is to determine the optimal feeder injection point and network configuration to minimize power losses and improve voltage profiles in radial distribution systems. In the 20 kV radial distribution system, the implementation of the proposed method reduces active power losses from 17.51% to 8.65% and decreases the number of buses with undervoltage conditions from 144 to 43 buses. In addition, the minimum voltage increases from 0.7468 p.u. to 0.8276 p.u. A comparison with other optimization methods shows that CSA and PSO achieve comparable performance. However, CSA exhibits faster convergence with simpler parameter tuning, while the GA method results in relatively higher power losses. These results indicate that the CSA-based approach is effective in improving voltage profiles and reducing power losses in radial distribution systems, and can be considered as an alternative optimization method for distribution network planning and operation.