Multi-scenario reconfiguration of an isolated 20 kV distribution feeder using the firefly algorithm for power loss reduction, voltage profile improvement, and reliability enhancement

ABSTRACT High power losses, voltage drops, and reliability limitations remain significant technical challenges in radial distribution systems, particularly in isolated power networks where operational flexibility is restricted and feeder performance directly affects service quality. This study presents a power system optimization framework based on the Firefly Algorithm (FA) for multi-scenario distribution feeder reconfiguration of the 20 kV Paya Seunara feeder in the isolated Sabang distribution system, Indonesia. The novelty of the proposed approach lies in the integration of FA-based feeder reconfiguration with a multi-scenario evaluation framework considering light, normal, and high loading conditions while simultaneously assessing power loss, voltage profile, and reliability performance. The feeder was modeled in ETAP 19.0.1, whereas the optimization process was implemented in MATLAB using a switch-encoding scheme that preserves radial topology and satisfies operational voltage constraints. The research workflow consists of feeder modeling, load-flow analysis, candidate switch identification, Firefly Algorithm-based optimization, and comparative performance evaluation under different loading scenarios. The results demonstrate that the proposed optimization framework consistently improves feeder performance across all operating conditions. Under normal loading, total active power losses decreased from 8.4343 kW to 6.5950 kW, corresponding to a reduction of 21.81%, while the minimum bus voltage increased from 0.9601 pu to 0.9668 pu. Under light loading, losses were reduced from 3.5175 kW to 2.7602 kW (21.53%), accompanied by an increase in the minimum voltage from 0.9686 pu to 0.9729 pu. Under high loading, losses decreased from 16.5393 kW to 12.9087 kW (21.95%), while the minimum bus voltage improved from 0.9498 pu to 0.9598 pu. The optimized topology also enhanced reliability performance, reducing SAIFI from 4.6426 to 3.6574 interruptions/customer/year and SAIDI from 21.7143 to 17.2344 hours/customer/year. These findings confirm that the proposed multi-scenario FA-based reconfiguration framework provides an effective and practical power system optimization approach for improving efficiency, voltage quality, and reliability in isolated radial distribution feeders under varying operating conditions.