Novel stochastic approach to predict the energy demand and thermal comfort in the office buildings considering materials and human-related Gaussian uncertainties

Abstract

In the decision-making process during the office buildings' design or management, a significant problem arises from the necessity of simultaneous satisfaction of the users’ thermal comfort and minimization of the energy demand. Since the features of the materials, the comfort perception and the number of people are uncertain quantities, the energy demand and parameters describing thermal comfort should be treated as random variables as well. The parameters, which influence mostly the energy demand and thermal comfort are gathered in two groups: the material properties of façade and the human-related factors. The thermal transmittance, solar heat gain coefficient of the transparent façade, number of people and internal temperature are Gaussian uncertainties, but meanwhile they should fulfil the specific requirements concerning health and safety. In the proposed framework, the perturbation theory and approaches based on the transformed random variables method, which enable to determine the probability density function of the energy demand and predicted thermal comfort, are employed and compared. The proposed approach is applied to investigate the uncertainty propagation considering the reference office building unit. The results show that the Gaussian randomness extends into the non-symmetrical uncertainty of energy demand and predicted thermal comfort. In comparison with the Monte Carlo method, the introduced methodology provides the explicit form of the probability density function of analysed random variables at a much lower computational cost.