Wireless Rechargeable Sensor Networks (WRSNs) can become a backbone of evolving field of Internet of Things (IoT). In WRSN, determining cost-effective charging schedules is a NP-hard problem and metaheuristics are often employed to find optimal solutions. This problem gets worsen in large-scale deployments and network partitioning using clustering provides a viable solution. The paper presents a comprehensive performance evaluation of integrated clustering-metaheuristic methods. The three clustering algorithms K-Means, DBSCAN, and affinity propagation are combined with four popular metaheuristics viz. particle swarm optimization, ant colony optimization, simulated annealing and genetic algorithm. Extensive simulations conducted on low and high workloads and reveal that ant colony optimization and simulated annealing with K-Means outperform other combinations. The affinity propagation demonstrates superior clustering quality showing higher Silhouette scores and better inter-cluster separation exhibited by low value of Davies–Bouldin index. The results thus provides critical insights for designing energy-efficient charging strategies in large-scale heterogeneous WRSNs.