Improving electricity demand growth estimation in rural mini-grids through data-driven appliance diffusion modeling

Abstract

Accurate load modeling is crucial for designing reliable and cost-effective mini-grids in rural, under-served communities accessing electricity for the first time. Current models often fail to capture the evolving energy demands associated with changes in appliance ownership and socio-economic growth. The study introduces a bottom-up, adaptable model that forecasts a progressive appliance adoption and customer base in order to improve long-term electricity demand estimations. This study employs real-world appliance adoption trends from rural Kenya as a case study and uses logistic diffusion and optimization techniques to model appliance diffusion. The findings highlight significant variability in appliance adoption rates already during the first years of electricity access between different household groups, identified through clustering algorithms. This variability underscores the need for dynamic modeling over traditional static categorization of end users to more accurately reflect evolving consumer energy consumption profiles. The proposed model serves as a tool to enhance multiyear load profile generation and support microgrid design in similar settings.