As electricity demand continues to grow and renewable energy sources become more widespread, power distribution systems face increasing operational challenges. Distributed generation—including rooftop solar systems, small wind turbines, and local power plants—has become an effective solution because it supplies electricity closer to consumers, reducing the burden on long distribution feeders.
However, simply installing distributed generators is not enough. Their location and capacity must be carefully optimized. Poor placement may overload certain feeder branches while leaving others underutilized, reducing the overall efficiency and flexibility of the distribution network.
Most previous studies have focused on reducing active power losses and improving voltage profiles. In contrast, little attention has been given to how evenly electrical loads are distributed across feeder branches. Uneven branch utilization can create localized thermal stress, accelerate equipment aging, and increase maintenance costs.
To address this issue, Trieu Ngoc Ton developed a new performance metric called the Branch Loading Balance Index (BLBI). The index measures how uniformly electrical loads are distributed across all branches in a radial distribution network. A lower BLBI value indicates a more balanced utilization of network assets, reducing the likelihood of excessive loading on individual feeder sections.
The proposed BLBI was incorporated into a multi-objective optimization framework alongside two traditional performance indicators: active power loss and voltage deviation. Instead of optimizing only one aspect of network performance, the framework simultaneously minimizes all three objectives.
The optimization process uses the Harris Hawks Optimization (HHO) algorithm, a nature-inspired metaheuristic based on the cooperative hunting behavior of Harris hawks. Network performance is evaluated using the backward/forward sweep load-flow method, a widely used technique for analyzing radial electrical distribution systems.
To validate the proposed approach, the researcher tested the model on two internationally recognized benchmark systems: the 33-bus and 69-bus radial distribution networks. The results were then compared with three widely used optimization techniques: Particle Swarm Optimization (PSO), Grey Wolf Optimizer (GWO), and the conventional HHO algorithm.
The findings demonstrate that incorporating branch loading balance into DG planning consistently improves overall network performance.
For the 33-bus distribution system, the proposed method achieved remarkable improvements:
- Active power loss decreased from 202.67 kW to 72.85 kW, representing a 64.05% reduction.
- The minimum bus voltage increased from 0.9131 per unit (p.u.) to 0.9704 p.u.
- The Branch Loading Balance Index declined from 0.1846 to 0.1192, indicating a much more balanced distribution of feeder loading.
The improvements were even more significant in the 69-bus system:
- Active power loss dropped from 224.98 kW to 71.83 kW, equivalent to a 68.07% reduction.
- The minimum bus voltage improved from 0.9092 p.u. to 0.9784 p.u.
- The BLBI decreased from 0.1963 to 0.1168, demonstrating superior branch utilization compared with existing optimization methods.
Across both benchmark systems, the proposed framework consistently outperformed PSO, GWO, and conventional HHO. It delivered the lowest power losses, the best voltage regulation, and the most balanced feeder loading conditions.
According to Trieu Ngoc Ton of Thu Duc College of Technology, integrating branch loading balance into distributed generation planning offers operational benefits beyond traditional optimization objectives. Instead of merely reducing energy losses, the proposed framework also distributes electrical current more evenly throughout the network, reducing thermal stress on heavily loaded branches and improving overall asset utilization.
The study has important implications for modern power systems, particularly as countries continue expanding renewable energy generation. Utilities and distribution network operators could use this optimization framework to determine the best locations and capacities for distributed energy resources, including solar photovoltaic systems, small wind generators, and other decentralized power sources.
More balanced feeder utilization may also extend the service life of transformers and distribution lines, lower maintenance costs, and improve the reliability of electricity supply. These advantages are increasingly valuable as smart grids become more complex and renewable energy penetration continues to rise.
Beyond distributed generation planning, the proposed Branch Loading Balance Index could support future applications involving network reconfiguration, battery energy storage systems, and advanced smart grid management. The framework also provides a foundation for integrating artificial intelligence into next-generation power system planning.
The author recommends that future research expand the model by incorporating reactive power support from distributed generators, time-varying load conditions, renewable energy uncertainty, and large-scale practical distribution networks. Such developments could further enhance the applicability of the proposed framework in real-world electricity systems.
Overall, the study demonstrates that considering branch loading balance alongside conventional technical objectives enables more efficient, reliable, and sustainable operation of modern electrical distribution networks.
Author Profile
Trieu Ngoc Ton is a researcher and lecturer at Thu Duc College of Technology, Thu Duc, Ho Chi Minh City, Vietnam. His research interests include electrical power systems, distribution network optimization, distributed generation planning, computational intelligence, multi-objective optimization, and renewable energy integration.
Research Source
Title: Incorporating Branch Utilization Uniformity into Distributed Generation Planning for Radial Distribution Networks
Author: Trieu Ngoc Ton
Journal: Indonesian Journal of Advanced Research (IJAR), Vol. 5, No. 6, 2026, pp. 973–986.
DOI/Official URL: https://doi.org/10.55927/ijar.v5i6.16701
0 Komentar