Influence of building thermal envelope modeling parameters on results of building energy simulation

Abstract

The paper investigates the influence of input values for building energy model parameters on simulation results. The methodology is based on dynamic simulations in TRNSYS software, including models for three different buildings, with variations in different parameters such as referent dimensions, infiltration, envelope thermophysical properties, and thermal capacitance of interior space. The method is applied to a case study of theoretical buildings located in Slovenia. Obtained results suggest that the accuracy of model parameters’ inputs related to thermal properties of glazing have the most significant impact on simulation results, where reduction in g-value from 0.62 to 0.22 reduces simulated qH,nd and qC,nd for 25% and 95% respectively. Referent dimensions for modeling floor area are the least influential parameter, where for Building III (BSF = 0.36) the change of referent dimensions results in variations of simulated qH,nd and qC,nd for less than ±1%, while for Building I (BSF = 0.62) qH,nd and qC,nd variations are up to ±20%. In general, for all model parameters, with a reduction of BSF, the influence of input variations is also reduced, meaning energy models for large buildings are more robust and less biased to deviation in simulation results from the inaccuracy of model inputs.