Firefly Algorithm-Based Optimal Deployment of Unified Power Quality Conditioner for Loss Minimization and Voltage Stability Enhancement in Distribution Networks

The increasing penetration of nonlinear loads and distributed generation (DG) in distribution networks poses significant challenges to voltage regulation and power loss minimization. Although Unified Power Quality Conditioner (UPQC), has proven effective in enhancing power quality and system stability, determining its optimal location and size remains a complex optimization problem. This study examines the optimal siting and sizing of UPQC in radial distribution systems with the objectives of minimizing real and reactive power losses while improving voltage profiles. A multi-objective optimization model is formulated and solved using the Firefly Algorithm (FA), with validation carried out on the IEEE 33-bus test system and the 71-bus University of Maiduguri feeder of Yola Electricity Distribution Company. For the IEEE 33-bus system, optimal UPQC placement at bus 5, achieved loss reductions from 201.775 kW and 134.991 kVAr to 56.478 kW (72.01%) real and 45.127 kVAr (66.57%) reactive and eliminated voltage violations. For the 71-bus feeder, UPQC placement at bus 39, reduced real and reactive losses from 306.586 kW and 289.050 kVAr to 144.902 kW (52.74%) real and 193.035 kVAR (33.34%) reactive and significantly mitigated voltage violations. Comparative analysis with the Artificial Bee Colony (ABC) algorithm shows that FA achieved faster convergence and offered slightly better loss reduction compared to ABC. It also delivered slightly lower UPQC ratings of 2705 kW and 1580 kW compared to ABC with 2720 kW and 1584 kW for the IEEE 33-bus test system and 71-bus University feeder, resulting in improved capacity economic efficiency. These findings demonstrate the effectiveness of the proposed FA method in accurately determining the optimal location and size of UPQC and highlight its strong potential for practical power quality enhancement in radial distribution systems.