Integrating isolated communities into national energy planning models - a case study in Bolivia

Abstract

Universal electrification in Bolivia requires demand representations that are context specific and compatible with system-level planning. This thesis develops an open-source two-stage framework that generates high-resolution electricity demand for unelectrified rural communities with the RAMP stochastic bottom-up model and integrates those demands into a national optimisation with EnergyScope MultiCell. Municipalities are grouped into Highlands, Valleys, and Lowlands. Information on electrified and unelectrified households for the 2024 baseline is obtained from the Instituto Nacional de Estad´ıstica (INE). RAMP inputs cover households, community services, and income-generating activities under an energy-sufficiency premise. Minute-level series are aggregated hourly and mapped to EnergyScope end-uses. On the supply side the model is extended with new cells, technology options suited to remote contexts, a population-density criterion to select off-grid solutions, explicit electrical losses and a constraint that allows home-system batteries to charge only from PV Home Systems. Three snapshots for 2024, 2035, and 2050 quantify technology choices, costs, and emissions. Results show marked regional heterogeneity, for example higher cooling and food preservation needs in the Lowlands, and higher water heating and space heating needs in the colder Highlands. They also show a dominance of PV plus battery portfolios in remote supply, together with an important contribution from diesel systems that decreases over time, and non-negligible impacts on national costs when remote regions are included. The contribution is a reproducible pipeline that includes data, code, and modelling choices, links micro-scale demand formation with country-scale planning, and strengthens the realism of scenarios for universal, affordable, and low-carbon access in Bolivia.