Abstract—Reassigning the phase connections of selected single-phase customers is an effective strategy for mitigating load imbalance in low-voltage distribution networks. Computational methods designed to optimize phase configurations inevitably depend on individual consumption data and accurate knowledge of customer-to-transformer connectivity. However, during the early deployment stages of smart grids— characterized by low smart meter penetration—such granular information is frequently incomplete or entirely missing, rendering most data-driven approaches inapplicable. To address these constraints, this paper introduces a novel, cost-effective load-balancing framework optimized for environments with limited metering penetration. The proposed strategy minimizes initial capital expenditures by strategically identifying the absolute minimum number of smart meters required for deployment. Using data from these newly installed devices, the methodology simultaneously infers customer phase connections and substation topology. Furthermore, it pinpoints the optimal candidates for physical reallocation, maximizing unbalance mitigation while strictly minimizing the total number of physical switching operations. The framework is validated through simulations utilizing real-world consumption profiles from diverse residential sectors in Tucumán, Argentina. The results demonstrate that a substantial reduction in load imbalance is achievable by instrumenting only 30% of the customer base and rephasing up to 15% of connected loads.

Keywords—data-driven methods, load balancing, phase identification, phase switching

Cite: Victor A. Jimenez, Adrian Will, "Minimal Investment Installation Planning of Smart Meters for Load Balancing Using a Data‐Driven Approach," International Journal of Smart Grid and Clean Energy, Vol. 15, No. 1, pp. 45-61, 2026. doi: 10.12720/sgce.15.1.45-61

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).