An Innovative and Holistic Approach for an Optimal Sizing of Mini-Grids in Rural Areas Integrating the Demand, the Supply, and the Network

The SE4all initiative of the United Nations wants to achieve universal electricity access by 2030 and one of the most significant challenges in this process is the electrification of remote rural villages in developing countries. Connecting these areas to the existing national grid is a durable but often prohibitively expensive solution. Instead, investors and electrification agencies are looking for low-cost and sufficient-quality electricity supply solutions. At the low voltage scale, individual solutions such as Solar Home Systems (SHS, a solar PV panel associated with a battery kit), or solar lanterns can be economically affordable but are usually considered as temporary solutions to reach the target. Autonomous mini-grids, combined with SHS kits can constitute a much more durable and economically affordable option. In this paper, we develop a methodology that automatically designs and scales the optimal mini-grid at the least cost to be installed in a given village and in a few minutes assesses the cost of electrification. Our methodology only requires some GIS (Geographical Information System) information of the village with a delimitation of its roofs. In a first step, we use machine-learning algorithms to predict the demand of each house in the village at a 15-minute granularity. In a second step, we use a mathematical optimization approach to best design the mini-grid to be installed to meet this demand. All generation assets (solar PV panels, diesel generators, SHS), storage assets (batteries) and the distribution grid reticulation are jointly optimized while calculating their capacity as an optimal investment, their optimal geographical location in the village, as well as their dynamic operations. All constraints relative to the technical functioning of the grid and the generation and storage assets are taken into consideration. Our methodology has two main advantages: first, by automating the full process, the calculation time of the electrification cost is drastically reduced and the methodology is easily scalable to any village. Second, it provides the cheapest and best-tailored mini-grid, thereby helping any investor or electrification agency in the process toward universal electricity access.