Electricity demand forecasting for rural communities in developing countries: Calibrating a stochastic model for the Bolivian case

The world crusade to close the electrification gap is coming to an end in most regions of the world. In recent years the research in the area has concentrated on the development of planning methods to minimise the cost of implementation. Although successful, the lack of focus on the complex dynamics that govern electricity demand lead to over/under-sizing of technical solutions resulting in waste of resources and missed developing opportunities. In this sense, this paper aims to propose an electricity demand model for rural communities
in Bolivia, based on an open-source bottom-up stochastic tool for load profile computation. The “energy sufficiency” concept is used to ensure that people’s basic needs for energy are met in all the analysed cases. Information from various sources, such as on-site surveys, databases and national reports were used to characterise the main geographical areas in Bolivia and the relative specific categories of users. Specific load curves generated with the model were used as inputs in a micro-grid sizing tool and the results were compared with an approach using a demand analysis in less detail. Main results show that the model obtained is capable of generating stochastic demand curves for single or multiple rural communities according to contextual particularities. Notably, the geographic location and the socioeconomic characteristics have a significant impact in the peak loads and the total demand. Considering small industries as an income generating activity can increase in the peak load by about 45%, consequently, there is an economic impact when investing in the
solution.