Optimal Placement and Sizing of Static Var Compensators on the Nigerian 330 kV Transmission Network: A Comparative Study of Genetic Algorithm, Particle Swarm Optimization, and a Hybrid Approach

Abstract

This study addresses voltage instability in developing-economy transmission grids by determining the optimal placement and sizing of Static Var Compensators (SVCs) on the Nigerian 330 kV, 50-bus network. Although metaheuristic techniques have been widely applied to FACTS device allocation, most studies rely on standard IEEE test systems, limited optimisation objectives or limited statistical validation on practical transmission networks. To address these limitations, this study applies actual operational data, develops a five-objective optimisation framework that simultaneously minimises the voltage stability index (Lₘₐₓ), active and reactive power losses, generation cost, voltage deviation and SVC installation cost, and compares the genetic algorithm (GA), particle swarm optimisation (PSO) and a hybrid PSO-GA approach. Newton–Raphson load-flow analysis was performed in PSAT/MATLAB, with SVCs modelled as variable shunt susceptances. The optimisation used GA, PSO and a hybrid PSO-GA algorithm comprising 15 PSO iterations followed by 15 GA generations. A normalised weighted-sum fitness function was employed over 20 independent runs, and results were validated using the Wilcoxon signed-rank test (p < 0.05). The base case identified Damaturu, Gombe, Maiduguri, Yola and Kano as northern buses with undervoltage conditions. All methods restored these voltages to acceptable levels, while the hybrid PSO-GA achieved the lowest Lₘₐₓ value (0.311) and the lowest total normalised objective (650). The findings indicate that PSO exploration combined with GA refinement provides balanced reactive-power compensation. The framework remains limited to steady-state analysis and requires contingency, dynamic stability and detailed economic assessment before practical deployment.