COMPARATIVE ANALYSIS OF METAHEURISTIC OPTIMIZERS FOR OPTIMAL THYRISTOR-CONTROLLED SERIES CAPACITOR SITING AND SIZING IN THE NIGERIAN 28-BUS TRANSMISSION NETWORK

Abstract

The Nigerian 330-kV transmission network exhibits poor voltage profiles and localized voltage instability associated with long radial feeders, limited reactive reserves, and generation–load mismatch, motivating coordinated Thyristor-Controlled Series Capacitor (TCSC) siting and sizing. However, the combined use of voltage-stability-based candidate screening, comparative metaheuristic optimization, and N−1 robustness assessment that distinguishes electrical from structural criticality remains insufficiently addressed for the Nigerian 28-bus network. A framework comprising Newton–Raphson load flow, Fast Voltage Stability Index (FVSI) screening, and a common weighted objective function for Political Optimizer (PO), Doctor and Patient Optimizer (DPO), and Ant Colony Optimization (ACO) was implemented on the 28-bus, 330-kV system. The five highest-FVSI lines were screened as candidate locations, while 20 independent runs with matched population and iteration budgets enabled repeatability assessment; N−1 analysis was subsequently applied to the eight most critical lines. All three algorithms identified Line 1–20 with a compensation degree of (k_c=-0.6771), corresponding to a 67.7% reduction in line reactance. DPO and ACO reproduced this solution in all 20 runs, whereas PO achieved it in 13 runs. The optimized TCSC reduced the sum-of-squared voltage deviation by 53.9% and RMS VDI by 32.1%, increasing the minimum voltage from 0.9379 to 0.9751 p.u.; however, active losses increased by 18.0%, from 925.78 to 1092.04 MW. Under survivable N−1 contingencies, VDI improved in all three cases, while losses increased in two. The results demonstrate that FVSI effectively screens stressed lines. The optimal compensation siting highlights the need for optimization-based FACTS planning and explicit structural-contingency assessment in sparsely meshed transmission networks.