Bayesian Inference Calibration of Building Energy Models for Arid Weather

Abstract

Abstract The building sector accounts for nearly 40% of global energy consumption and plays a critical role in societal energy security and sustainability. A building energy model (BEM) simulates complex building physics and provides insights into various energy-saving measures’ performance. The analysis based on BEMs has become an essential approach to slowing down increasing building energy consumption. The reliability and accuracy of BEMs have a high impact on decision-making. However, how to calibrate a building energy model has remained a challenge. In this study, Bayesian inference was applied to the calibration of an office building model under the arid weather conditions of Doha, Qatar. The coefficient of variation with a root-mean-square error of calibration and validation are 1.1% and 1.5%, respectively, which is highly satisfied with the monthly calibration tolerance of 15% required by ASHRAE Guideline 14. Additionally, the calibrated parameter results are with probabilities and degrees of confidence, so they are more reasonable and comprehensive than traditional deterministic calibration methods. This study conducted a sensitivity analysis to select the model’s dominant parameters under hot/arid weather conditions. This study will be among the first studies of stochastic calibration based on Bayesian inference for building energy performances in arid weather.