Teaching-Learning Based Optimization Approach for Determining Size and Location of Distributed Generation for Real Power Loss Reduction on Nigerian Grid

Abstract

The ever increasing sensitization on the need for clean energies that are not only environmental friendly but also have comparative cost advantages encourages the use of distributed generation. Using distributed generation at the load ends or close to the load centers has not only reduced carbon emission, but also improves power system performances. Presented in this paper is the adoption of Teaching-Learning Based Optimization technique for determining the most suitable site and size of distributed generation for real power loss reduction on Nigerian power system. Backward/Forward Sweep technique was employed for the power flow analysis, while the suitable locations of the distributed generations were pre-selected using Voltage Stability Index and Teaching-Learning Based Optimization technique was employed to establish the optimal location and the optimum size of the required distributed generation. This approach was demonstrated on the IEEE 34-bus test system, with the placement of 1 kW DG at bus 11 of the system. The aggregate real power loss diminished from 571 kW to 208.5954 kW (63.5726% reduction), while Voltage Stability Index and voltage profile of the system also improved remarkably. Also, by placing distributed generation on typical Nigerian 11 kV feeder, the real power loss reduced from 1.1 kW to 0.75 kW while the magnitude of bus voltage increased from 0.8295 to 0.8456 p.u. Based on the results of this analysis, Teaching-Learning Based Optimization has demonstrated excellent performance on the two test cases and therefore would be a tool to adopt on the Nigerian radial distribution system.