Optimized network planning of mini-grids for the rural electrification of developing countries

1.2 billion people, predominantly living in remote rural regions in countries of the Global South, currently live without access to any modern source of energy. Options for electrification of these communities include extending existing national grid infrastructure, deploying mini-grids, and installing standalone home systems (SHS). Deriving the most cost effective means of delivering energy to these consumers is a complex, multidimensional problem that normally requires determination on a case-by-case basis. However, optimization of the network planning may help to maximize the socio-economic return of the installed energy system. This paper presents an optimization process that minimizes the installation cost of a mix of generation sources for a rural mini-grid using a multi-objective particle swarm optimization (MOPSO) technique. Minimizing the cost of distribution layout is first formulated as a capacitated minimum spanning tree (CMST) problem and solved using the Esau-Williams method. Multiple cable sizes and source locations are then added to create a multi-level capacitated minimum spanning tree (MLCMST) problem, solved via a Genetic Algorithm (GA) employing Prim-Pred encoding. The method is applied to a case study village in India.