From Building Archetypes to Data-Driven Definitions: The Impact of Window-to-Wall Granularity on Urban Building Energy Modelling

Abstract. This research investigates how window-to-wall ratio (WWR) granularity affects urban building energy modelling for Positive Energy District planning. We use surface-level WWR data extracted from oblique aerial imagery and compare the resulting simulation outputs against fixed district-average and generic WWR assumptions across two neighbourhoods: Feijenoord and Prinsenland, in Rotterdam, The Netherlands. The analysis uses harmonised building data, adapted to the specific requirements of CitySim and SimStadt simulation tools to evaluate each tool independently against a baseline. Results indicate that fixed WWR assumptions significantly distort simulated energy profiles. The highest effects occur for cooling demand: applying a fixed 20% WWR leads to an increase in the cooling demand by 106.2% in Feijenoord and 127.8% in Prinsenland using CitySim. Such behaviour happens as well in SimStadt, where cooling demand values increase by 67.2% and 68.5%, respectively. Building-level distributions reveal even stronger local deviations, particularly where baseline cooling demand is low. Finally, the findings show that while neighbourhood-average WWR values reduce aggregate error relative to generic assumptions, they still obscure building-specific peaks that are critical for accurate district energy planning.