Cost‐Based Optimal Allocation of Shunt Capacitors in Radial Distribution Networks Considering Load Types Using Crow Search Algorithm

Radial distribution networks (RDNs) often experience high power losses, voltage instability, and operational inefficiencies because of their low‐voltage, high‐current characteristics, and fluctuating load behavior. This study investigates the optimization of capacitor placement and sizing using the Crow Search Algorithm (CSA) to enhance voltage stability, minimize power losses, and reduce operational costs. Power flow analysis was conducted via the Backward/Forward Sweep (BFS) method on the IEEE 33‐bus system, incorporating four load models: Constant impedance (CZ), constant current (CI), constant power (CP), and composite ZIP. The optimization objective was to minimize the total operating annual cost (TOAC), which includes active power loss costs and annual capacitor installation costs. CSA’s performance was benchmarked against invasive weed optimization (IWO), teacher learner‐based optimization (TLBO), and artificial bee colony (ABC) algorithms. Simulation results demonstrated that CSA improved the voltage stability index (VSI) from 0.61 to 0.68 (CP), 0.69 (CI), 0.65 (CZ), and 0.66 (ZIP), and reduced active power losses by 30.41%, 26.01%, 36.45%, and 33.78% for CP, CI, CZ, and ZIP loads, respectively. Corresponding cost savings achieved using CSA were 30.28% (CP), 25.85% (CI), 36.35% (CZ), and 33.67% (ZIP), confirming superior performance over the compared methods. The highest TOAC was observed for CZ loads ($303,570.2), while the lowest was for CI loads ($195,386.4), reflecting the economic influence of load model variation. To ensure practical applicability and robustness, a sensitivity analysis of optimization algorithms under varying load models was performed. This included evaluating the impact of varying capacitor sizing ranges, system sizes, load compositions, and algorithm initialization conditions. Results confirmed that CSA maintained stability and effectiveness under diverse scenarios, validating its reliability and adaptability in real‐world RDN planning. Overall, the proposed CSA‐based approach provided a robust and cost‐effective solution for optimizing capacitor placement and sizing in RDNs, supporting both technical performance and economic sustainability.