Optimal allocation of phase-switching devices for dynamic load balancing

Customer rephasing is one of the most efficient techniques for balancing loads, reducing losses, and improving electricity quality. Traditionally, phase changes were performed by an operator, interrupting the supply during the process. Instead, a more recent approach proposes using phase switching devices that dynamically and imperceptibly transfer the customer load from one phase to another. Although this solution requires an extra investment related to these devices, it preserves the balance over time. This paper presents a new method to determine the minimum number of devices that should be installed and find their optimal allocation. The method is based on genetic algorithms and only uses data from the substation and a fraction of the customers, requiring less than 50% of penetration of smart meters to fix the imbalance problem. A new control algorithm for the switching devices is also proposed. It was designed to reduce the load imbalance while maintaining the number of changes per customer to a minimum, reducing device degradation. The method was validated through simulations using data from Tucuman province, Argentina. With at least 40% of the customers measured, it is possible to allocate the switches to less than 20% of customers and correctly reduce the load imbalance.