Computational model calibration of a multi-family housing located in the city of João Pessoa-PB

Abstract

This paper aims to present the calibration of a computational model of multifamily social housing located in João Pessoa City by comparing indoor air temperatures measured on-site with values simulated in the EnergyPlus program. The methodology used for the development of this research consists of three stages: monitoring a multifamily residential building by measuring temperature and relative humidity; computational modeling of the building; and finally, calibration of the simulation model through sensitivity analysis using NMBE, CV(RMSE), correlation coefficient (R), and determination coefficient (R²) as indicators of the linear regression between measured and simulated air temperature values. Variations in thermal resistance, absorptance, internal heat gains from equipment, occupancy patterns, and air infiltration were analyzed. The results showed that the initial model already had good results and could be considered calibrated according to the ASHRAE Guidelines 14 requirements. Some factors may have contributed to a good result of the initial modeling, such as considering the building surroundings and simulating natural ventilation considering pressure coefficients calculated by a CFD tool. In the sensitivity analysis calibration procedure, a more significant increase in R² was observed for the bedroom environment in winter.