A cost-effective integration and operation methodology for battery energy storage systems in active distribution networks via a master–slave optimization strategy

This document proposes a master–slave optimization approach for the integration and operation of energy storage technologies (ESTs) in active distribution networks (ADNs), combining the multiverse optimizer (for selecting the optimal location and type of EST) with the vortex search algorithm (for determining the hourly operation scheme). This method accounts for the variability of distributed generation (DG) and the fluctuating power consumption patterns of ADN users, aiming to minimize system costs—including energy purchasing, investment, maintenance, and replacement expenses—over a 20-year planning horizon. The approach was validated on 33-bus and 69-bus test systems, both adapted to the demand and generation conditions of Medellín, Colombia, and compared against five metaheuristics: particle swarm optimization, the Monte Carlo method, the Chu & Beasley genetic algorithm, the salp swarm optimization algorithm, and population-based incremental learning. As observed in MATLAB simulations for the 33-bus system, the proposed methodology achieved the greatest savings, reducing annual costs by up to 14,138 USD and outperforming all methods. It also obtained the best average cost (2,965,728.33 USD) with a notably low standard deviation of 0.020%, while maintaining moderate processing times (170 min). In the 69-bus network, it similarly yielded the best cost results and confirmed its scalability to larger, more complex ADNs. These findings demonstrate that the master–slave synergy of the multiverse optimizer and vortex search algorithm offers network operators a robust, repeatable solution to reduce the total cost of ADNs when integrating ESTs under varying renewable energy and demand conditions. • Integrates storage in grids to reduce costs considering investment and operation. • Uses MVO and VSA for optimal siting, type selection, and dispatch of storage. • Outperforms existing methods in minimizing costs in active distribution networks. • Validated under real demand and renewable variability conditions in Medellín.