Integrating Medium-Voltage Lines into the OnSSET Model for Electrification Planning

Abstract

Extending electricity access to the 1 billion people worldwide who live without is a complex and costly undertaking, requiring data-driven methods to inform decision makers. Models for electricity access optimization fill this role by leveraging geospatial data to determine the least cost grid and off-grid electrification options for a region of study. The emergence of modern machine learning techniques coupled with abundant satellite imagery promises to provide a new source of transmission and distribution infrastructure data for electricity access modeling, yet questions persist about how to best integrate this new data into existing models. My work focused on the popular OnSSET model for electrification planning, and I explored updating the model to accommodate new data on the location of Medium-Voltage (MV) transmission infrastructure. In addition to adjusting the core model to incorporate this new class of input data, I built an add-on to optimize grid-extension from MV interconnections. Applying this updated model to a case study in Afghanistan, I found small parameter changes can have large impacts on the electricity planning pathways for a region. Similarly, the inclusion of MV infrastructure has great implications for the feasibility of grid extension.