Optimisation of Location and Size for Distributed Generation in Unbalanced Grids

Abstract

This paper proposes a distributed generation installation procedure that can be applied to balanced and unbalanced distribution grids. This is achieved through metaheuristic optimisation and by modelling the grid using the three-phase grid model rather than the one-line diagram model. Both the power loss in the lines and the voltage deviation from the nominal voltage are minimised using the multi-objective genetic algorithm. This installation procedure facilitates grid planning for the distribution system operator. In particular, various possible Pareto optimal solutions can be chosen, which all lead to significant improvements in the performance of the grid. When applied to the unbalanced IEEE37 grid, this method achieves an installation capacity of 3 MW while significantly reducing power loss in the lines by 83.7%, and voltage deviation from the nominal by 90.6%.