Sky emissivity model calibration with data from Brazil and building simulation sensitivity analysis

Downward long-wave radiation is an important parameter for building simulation because it strongly influences the thermal balance of the external surface. Brazil has several distinct climatic regions, and little formal work has been done with observed downward long-wave radiation to build thermal energy balancing due to a lack of continuous data. In this regard, selecting a model and adjusting local parameters is fundamental for reducing the uncertainty of the computational simulation process. In the present study, we use data from four different latitude sites to adjust the parameters of four models integrated into the EnergyPlus simulation program. The results obtained in the analysis indicated that with an appropriate adjustment of local parameters, all models achieved good predictive performance and provided lower errors values than previous versions, when using versions incorporated into the EnergyPlus, and when using original versions. In addition, it is recommended to use the Berdahl and Martin's [10] model adjusted by this study in the building energy simulations carried out in Brazil, since it obtained the lowest rRMSE values, between 2.2% and 2.5%, while the standard model of EnergyPlus varies between 5.1% and 10.5%. After validating the adjustments and considering a single-family house of 43.25 m2, it was concluded that without proper adjustment, the models incorporated into the EnergyPlus, overestimate the total annual thermal load by 8% to 37% when compared to the model the proposed in the present work.

Automated summary

This study adjusts the parameters of four models used in building simulation with the data from different latitude sites, and finds that proper adjustment of local parameters can reduce the uncertainty of the computational simulation process.