Calibration and extrapolation of a stochastic model for electricity demand forecasting in rural communities: the case of Samionta, Benin

To achieve sustainable rural electrification in sub-Saharan Africa, innovative approaches are required to address technical inefficiencies, high electricity costs and socio-economic disparities. In this context, accurate electricity demand forecasting is a fundamental prerequisite for efficient design and operation of solar microgrids in rural areas where consumption profiles are highly variable, which highlights the need for robust predictive tools. Therefore, this paper aims to calibrate and extrapolate the Remote-Areas Multi-energy systems load profiles (RAMP) model to anticipate future electricity demand in order to plan the increase in the capacity of the solar microgrid of Samionta, a rural community in southern Benin. Using data from smart meters and socioeconomic surveys, the calibration process integrates key variables such as socio-economic conditions, appliance adoption and usage patterns, and external conditions such as climatic conditions. The resulting improvements allow the model to accurately forecast the community’s energy demand over time. With this calibrated model, the study extrapolates demand trajectories with different scenarios highly influenced by PUE, offering reliable forecasts for system upgrades over several years.
Furthermore, the calibrated model has potential for application in other rural communities with similar conditions. This paper underscores the importance of adaptable demand prediction tools in solar microgrid projects. It establishes a replicable with the potential to minimize costs, improve energy access, and foster socioeconomic development. Such advancements highlight the broader relevance of this approach in accelerating sustainable rural electrification in Sub-Saharan Africa