Uncertainty and sensitivity analysis applied to hygrothermal simulation of a brick building in a hot and humid climate

This paper presents a statistical approach for uncertainty and sensitivity analyses applied to 14 inputs whose 10 properties associated with brick material, using the four different EnergyPlus wall models. The variability of inputs has been extracted from several characterization works presented in IEA Annexes 14, 24 and 55, being coherent to the lack of knowledge in the early design stage. Besides the methodology, this paper presents the moisture effects on cooling energy demand and indoor air conditions, using a simple building geometry and the humid climate of Singapore. Results are presented in terms of uncertainty quantification, most uncertain parameters and sensitivity indices for all models, illustrating the impact of moisture and the importance of the need to well define moisture-dependent functions. The methodology is well adapted for use in complex interactive models with a low-cost simulation and can be used to reduce uncertainties in the design stage and promote reliability of retrofitting assessment.