Impact of spatial resolution in modelling decarbonized district heating networks

This study investigates the impact of spatial resolution on the large-scale optimization of district heating networks expansion using the hot-SPOT model. While previous research has focused on optimization of district heating expansion at district level, the effect of spatial resolution on the large-scale modelling of transport network expansion remains largely unexplored. This work addresses this gap by analysing how different levels of data aggregation, and thus spatial resolutions, influence key system parameters, including network topology, total system costs and computational efficiency. Results show that increasing spatial resolution leads to a more detailed representation of the network, with the benchmark resolution producing a hybrid transport-distribution network that is approximately six times longer than aggregated cases. Despite this, total heat generation mix and system costs remain comparable across resolutions due to the significantly shorter average branch lengths in the disaggregated case. The present work also quantifies the relation between model accuracy and computational feasibility. While finer resolutions provide more granular insights, they also introduce significant computational costs: for the disaggregated resolution, the model runtime increases to approximately 7400 min (5 days), with the optimization phase being the primary bottleneck, whereas aggregated resolutions remain below 60 min. Moreover, the study highlights a structural limitation in the cost formulation of hot-SPOT, a linear optimization model that is focused on large-scale energy transport modelling, which underestimates network costs by not differentiating between transport and distribution infrastructure, the latter typically characterized by higher specific costs. These findings emphasize the need for strategic aggregation methodologies to balance computational efficiency with network accuracy.