GIS-facilitated procedure for optimal rural electrification planning: A case study in Naeder, Ethiopia

Although the global electrification rate has reached 91 %, roughly 730 million people still live without reliable and affordable access to electricity, experiencing the first regression since 2013, following the COVID-19 crisis. In this context, this paper aims to define a comprehensive electrification strategy through an innovative model based on open-source data and machine learning algorithms, able to reduce the time and resource-consuming on-field campaign that is generally needed for gathering data, and subsequently define the electrification strategy. Following the location of human settlements and their socio-economic characterizations carried out by a novel open-source tool proposed within this paper named VANIA (Village ANalytics in Africa), the energy demand and hourly demand profile of each community are estimated through the application of machine learning techniques based on MTF (Multi-Tier Framework) surveys and a stochastic bottom-up model for load profile generation. The approach is designed to manage the complex nonlinear relationship between the energy needs of a community and its socio-demographic parameters. Then, taking the communities' demand profile as input, a GIS-facilitated procedure is utilized to optimize the electrification strategy for the territory under investigation, proposing the least-cost electrification solution. The final electrification plan focuses on long-term solutions enabling growth over time in which each community can be either connected to the national grid or supplied by an off-grid system. Ultimately, to demonstrate the approach and showcase its operational capabilities, the methodology is utilized for the electrification planning of the Naeder province in Tigray, Ethiopia, characterized by a predominantly lacking electrification status and low energy demand. The suggested solution advocates for the cost-efficient electrification of approximately 11,560 households clustered in 50 communities. Considering consolidated economic parameters and a perceived cost of electricity of 110 €/MWh showed a preference toward grid extension, with 39 out of 50 communities connected to the national grid. Finally, sensitivity analysis on the cost of energy showed that regardless of the value, 3 communities should be electrified with a microgrid, whereas for values upward of 130 €/MWh the microgrid starts becoming the more lucrative option, and at 145 €/MWh an extension is not economically justified.