Distributed Generation Placement and Sizing via Ant Colony Optimization in Energy Network for Power Loss Reduction and Voltage Improvement

Abstract

A lot of studies have been conducted on the exact location of distributed generation (DG) in a given distribution network so as to lower power loss which is a crucial index for voltage drops and poor power quality but none has used Ant Colony Optimization (ACO) voltage sensitivity matrices and loss sensitivity based candidate ranking in Amadi-Ama 11kV energy network which reduce optimization search space and improve convergence of the metaheuristic algorithm. This research aims at correctly placing and sizing DG (photovoltaic plant) in the Amadi-Ama 11kV distribution system for active power loss reduction across lines and voltage profile improvement across nodes. The study employs Forward Backward Sweep (FBS) load flow technique to ascertain the power flow and voltage level in the network before and after the integration of DG. FBS is deployed as a result of the high R/X (resistance/inductance) ratio in a distribution line, its weakly meshed configuration and large number of nodes. In the forward sweep of the FBS, voltage drops are determined with power flow update while in the backward sweep, the currents are computed from the branches in the last layer. Sensitivity based candidate ranking is employed to reduce the search space of the metaheuristic algorithms for DG integration, contributing to the overall efficiency, high score gives the expected benefit from DG. Ant Colony Optimization (ACO) employs pheromone which is responsible for assigning resource to positions and is updated by evaporation and deposition from good solutions. The update guides probablistic selection of DG locations. The ACO installed a DG size of 150 kW at bus 12 and records a voltage deviation of 0.4022 V and a corresponding total active power loss of 225.87 kW which is 40% of active power loss reduction, as against the base case power loss of 374 kW. Linear regression model is applied to examine the relationship between  the length of line (number of lines/buses) and power losses/voltage profile before and after the inclusion of the photovoltaic plant in the system. The optimizer is found to fulfill the primary objective of this research and maintains the regulatory voltage limit of 0.95-1.05 p.u. It is found that the optimizer maintained the DG power constraint of 120-180 kW. The results of this research are validated by comparing it with other existing models and found superior as the ACO records a 40% active power loss reduction against 22%. Therefore, the study recommends the ACO based voltage sensitivity matrices and loss sensitivity candidate ranking as a veritable approach for improving the operational performance of the Amadi-Ama 11kV distribution network. The simulation analysis was carried out on MATLAB/Simulink Software 2021.