A Comparative Analysis of Different Future Weather Data for Building Energy Performance Simulation

Abstract

The building energy performance pattern is predicted to be shifted in the future due to climate change. To analyze this phenomenon, there is an urgent need for reliable and robust future weather datasets. Several ways for estimating future climate projection and creating weather files exist. This paper attempts to comparatively analyze three tools for generating future weather datasets based on statistical downscaling (WeatherShift, Meteonorm, and CCWorldWeatherGen) with one based on dynamical downscaling (a future-typical meteorological year, created using a high-quality reginal climate model). Four weather datasets for the city of Rome are generated and applied to the energy simulation of a mono family house and an apartment block as representative building types of Italian residential building stock. The results show that morphed weather files have a relatively similar operation in predicting the future comfort and energy performance of the buildings. In addition, discrepancy between them and the dynamical downscaled weather file is revealed. The analysis shows that this comes not only from using different approaches for creating future weather datasets but also by the building type. Therefore, for finding climate resilient solutions for buildings, care should be taken in using different methods for developing future weather datasets, and regional and localized analysis becomes vital.