Abstract
Although energy simulation can provide valuable information about building energy behavior, the inaccuracy of input data can undermine the reliability of the results. Despite the vast literature about uncertainty analysis, little is known about the influence of building characteristics on the propagation of uncertainty through the energy simulation models. This study investigates the extent to which uncertain thermal conductivity and specific heat of structural layers affect the annual heating and cooling needs for a set of 144 simplified reference building configurations in three European climates. The analysis is carried out by means of a Monte Carlo technique coupled with TRNSYS hourly simulations. This study points out that the uninsulated residential buildings with a high aspect ratio, a small transparent surface with south exposure and low SHGC are more sensitive in the cooling needs estimation to the propagation of uncertainty in material properties. Similarly, the heating needs precision is greatly reduced when the uncertainty affects the thermal conductivity in uninsulated buildings with a low aspect ratio, a high SHGC and small window size. On the contrary, the uncertainty in specific heat is emphasized in buildings with external insulation having large windows oriented to either east or south and with glazing characterized by high SHGC.