Development of a modeling framework to forecast power demands in developing regions: Proof of concept using Uganda

Abstract

Accurate and detailed energy demand estimates are crucial to achieving adequate energy infrastructure planning. These estimates are often non-existent or deficient in many developing countries, and consequently, electricity supply is unreliable. A novel approach for estimating electricity demand is presented. Our approach uses a global geographical population database with 1km2 spatial resolution as the foundational input. The use of spatial population data is based on the premise that electricity consumption is dependent on where people are located. These population counts are converted to electrical customers to create spatial power demand data which can be mapped. The resulting power demand maps could be valuable for energy infrastructure planning. In this study, Uganda is used as a pilot case-study. Analysis suggests that an additional 1.5 GW of power generation capacity needs to be availed to meet the lowest power demand scenario. The methodology developed can be extended to other regions of interest.