Search

Climate has the most profound impact on the buildings' energy performance, especially now due to the ongoing global weather changes. Therefore, selecting appropriate weather data for building energy simulation is crucial. This paper aims to advance the knowledge about the use of different weather datasets for building performance simulation by addressing the following research objectives: (i) understanding the statistical relevance of using a typical weather year (TWY) for running building energy simulations, if compared to a series of actual weather years (AWY), for different buildings' typologies and under different climate conditions; (ii) verifying the role of building features on the discrepancies between TWY-based and AWY-based simulations. Tackling these objectives implied simulating a complex university building and a typical single-family dwelling by using two TWYs and ten AWYs pertaining to data recorded from 2010 to 2019 in both cold (i.e., Winnipeg, Canada) and warm (i.e., Catania, Italy) climates. Results show that in Winnipeg, TWYs predicted from 2.7% to 11.3% lower heating demand and from 10.5% to 82.4% higher cooling demand than the average long-term from AWYs, while in Catania TWYs predicted from 1.8% to 8.7% lower cooling demand and from 2.8% to 82.4% higher heating demand, suggesting that buildings designed using TWYs might not perform as modelled under actual weather conditions. • Evaluation of different TWYs and AWYs on the dynamic energy simulation of buildings. • TWYs and AWYs are developed from recent (2010–2019) weather observations. • Buildings with different features and under different climates are analysed. • TWYs could understimate the dominant energy demand in the range of 2%–11%. • TWYs could overestimate the minor energy demand in the range of 3%–83%. [ABSTRACT FROM AUTHOR]