Abstract
Building energy simulations require a detailed characterization of the boundary conditions to solve the air heat balance problem. For dry and dew bulb temperatures and wind speed, measured hourly profiles can be easily included in weather files. On the contrary, beam and diffuse solar irradiance values incident on the building envelope are not directly available. This requires the adoption of solar irradiance models, which are often based on statistical correlations derived from empirical data. Since the samples of solar irradiance measurements used for the models development have been collected mostly in North America and European localities, no model can provide an adequate worldwide representativeness or be precisely defined as the best one. In this research we investigate the impact of the choice of solar irradiation models on simulated hourly energy needs in five European climates (Berlin, Vienna, Trento, Rome and Messina). The full combination of 22 horizontal diffuse irradiance models and 12 irradiance models for tilted surfaces has been considered for the development of hourly solar irradiation profiles, used as input in building energy simulation (i.e., TRNSYS) for a set of 72 simplified reference buildings. The results show that the variability of the estimation of solar irradiation leads to different levels of uncertainty in hourly energy model predictions, also depending on the building characteristics.