When engineers model a building’s energy use, they almost always start with weather data pulled from the nearest airport. It’s an easy default, but it introduces a subtle, persistent error. An open runway has virtually nothing in common with a 1930s apartment block wedged into central Athens, surrounded by asphalt, shade, and traffic.
In a recent paper for Energies, part of the EU-funded PRELUDE project, our team wanted to see just how far off those standard airport assumptions really are.
- Article title: Impact of the Urban Environment on the Thermal Performance and Environmental Quality of Residential Buildings: A Case Study in Athens
- Published via: Energies (2025), 18(8), 2062
- Project: PRELUDE, funded by the European Union’s Horizon 2020, Grant Agreement N° 958345.
- Authors: Maria Kolokotroni, May Zune, Petra Gratton, Thet Paing Tun, Ilia Christantoni and Dimitra Tsakanika
Why a generic weather fails building energy model in urban area
We focused on a five-story hostel on Patission Street in Athens. Using a 3D model of the surrounding neighbourhood alongside MIT’s Urban Weather Generator, we built a microclimate profile tailored to that specific street, then ran the numbers against generic airport models.
Street-level weather naturally diverged from the runway: higher temperatures from the urban heat island, less direct sun behind adjacent buildings, and shifted wind paths. The real surprise came when we mapped those microclimates back to building energy performance:
- Summer cooling spikes completely wiped out winter heating savings. Warmer ambient air did reduce cold-weather heating demand, but summer cooling needs jumped by a much larger margin, driving overall annual energy requirements far past standard model predictions.
- Location inside the building was a massive factor. Top-floor units took a double hit from direct sun and canyon wind, pushing summer cooling loads well beyond the shaded lower units. In a 2050 warming scenario, energy demand for that top space jumped ~75% over today’s baseline.
- Shoulder months were essentially free. For nearly a third of the year – mostly November and March through May – the building naturally sits right near comfortable thresholds if you simply manage windows and blinds strategically.
A low-tech fix for urban building performance
The catch, of course, is human behaviour. Most apartment blocks don’t have a facility manager constantly watching weather forecasts to manage windows.
To test a low-tech workaround, we built a simple regression tool in Excel that takes incoming forecasts and projects next-day indoor temperatures and levels, no IoT sensors or complex building management hardware required. During an April 2024 trial, we sent daily window-opening advice to residents based on the spreadsheet’s output. Indoor monitors showed average
dropped from 1,626 ppm on unguided days to 853 ppm on guided ones, effectively doubling fresh air ventilation without drawing an extra watt of power.
Generic weather files can make energy models wrong in different ways at the same time, hiding top-floor cooling spikes while still over-predicting winter heating demand. Understanding those blind spots doesn’t require massive simulation models; it just requires accurate, street-level data with UWG applications.
Leave a Reply