Increasing spatial and temporal resolution in energy system optimization model – the case of Kenya

Abstract At the time of writing 759 million people (2019) still lack access to electricity globally. Energy planning is important to describe plausible pathways to achieve national goals, where energy systems models are tools to explore scenarios and provide insight. Until recently, modelling energy access in countries with a low electrification rate was conducted at low spatial (e.g. national) and/or temporal resolution (e.g. annual time slices or ‘overnight’ electrification). In this paper, we develop methods in an open-source computational workflow with high spatial and medium temporal resolution in OSeMOSYS. We use Kenya as our case application where still 16 million people lack access to electricity (2019). The spatial resolution of approximately 40x40km cells leads to 591 demand cells split between electrified and un-electrified. Results show that a key limitation of PV + battery systems is that they struggle to provide peak demand. This manifests itself in very high marginal costs at peak times. That high marginal cost drives a transition of the system to mini-grid and grid connections as soon as demand levels are high.