Energy transition pathways for isolated regions in the Global South - A case study of Bolivia's Northern Amazon

Abstract

The Bolivian Northern Amazon covers 21 municipalities and 84,209 households. A quarter of them have no public electricity supply, and the region produces almost all of its electricity from diesel. Those households have never been metered, so their demand appears in no record, and no line reaches them, so serving them costs money that a generation model does not see.

This thesis asks what alternative energy transition pathways towards universal access cost and emit for the region. Universal access here means that no household is left without electricity, whether from a line or from its own solar system, and not that every household is connected. RAMP, a stochastic bottom-up model, builds the demand of the unmetered households from the appliances they own. EnergyScope MultiCell, a linear programming model of a whole energy system, finds the cheapest mix of generation, storage and imports that supplies it. A geospatial layer compares, community by community, a medium voltage line with standalone solar home systems, and prices the network that follows. Five scenarios are built for 2025, and seven further runs extend the analysis to 2035 and 2050, one of them with an announced electricity import from Brazil. The comparison separates three levers: how the electricity is produced, the access given to the households that have none, and the level of service assumed for the households that already have electricity.

Rebuilding the supply is the largest lever, and it asks nothing of the population. Replacing diesel generation with utility solar and battery storage cuts the annual cost of the system from 66.51 to 42.15 million euros and reduces its emissions by two thirds. However, waiting is expensive: a supply that cannot be rebuilt costs 45.39 million euros more per year by 2035 and 70.85 million more by 2050, because it keeps paying for fuel while the demand grows.

Serving the 20,714 households that are badly served or not served at all costs 6.50 million euros more per year, of which 3.99 million is medium voltage line and 2.51 million new generation. A study that prices only the power plants therefore reports less than half of what access costs.

The level of service assumed for households with electricity is more important than access itself. Bringing every household of the region to the same sufficiency bundle, a normative minimum of energy services, costs 35.27 million euros per year. This is approximately half of what the region currently spends, because the connected households consume more than that bundle. Emissions follow the supply and not the demand: the scenarios that replace the diesel plants emit 58 to 74 kilotonnes of CO$_2$ equivalent per year, and those that keep them about 217. Rebuilding the supply should therefore come first, and the access programme should be planned alongside it. Only the storage decision should wait for the Brazilian interconnection, since electricity available at every hour does what a battery does.