Comparison Of Quasi-Steady State And Dynamic Simulation Approaches For The Calculation Of Building Energy Needs: Thermal Losses

One of the aims of the European Directive 2010/31/EU (formerly 2002/91/EC) is to reduce the energy consumption of buildings introducing an energy labeling protocol which is expected to capture the attention and reorient the market, sustaining the diffusion of more efficient solutions. In order to evaluate the building energy performance, either analytical approaches or enhanced simulation tools are allowed. The coherence of the methods is important in order to avoid misleading results which can affect the evaluation by the market and eventually compromise the Directive effectiveness. The European Standard EN ISO 13790:2008 suggests to use the dynamic simulation in improving and tuning the quasi-steady state method proposed, and in particular to refine the correlation used to calculate the utilization factors (i.e., the dynamic parameters which reduce the thermal gains for heating need calculation and the thermal losses for cooling). Many efforts in calibrating the EN ISO 13790:2008 led to some changes on the correlations proposed in order to adapt the method to the climatic conditions, especially for the cooling season, and the building stock’s characteristics in different countries, but large discrepancies have been found. Differently from the previous works, the authors analyze the discrepancy sources focusing firstly on thermal losses, instead of considering directly the final result in term of energy needs. In this paper, the deviations between the thermal losses are evaluated, by means of an extensive use of simulation, analyzing a set of a 960 configurations obtained by the factorial combination of different values for the building shape, envelope insulation and composition, window type and size, ventilation rate and climatic conditions. Six different setpoint conditions were considered for the simulations. The analysis allowed the authors to identify the relevance of the deviations and to suggest ways to improve the correspondence between simulation and quasi-steady state methods in tuning processes