Weather datasets generated using kriging techniques to calibrate building thermal simulations with TRNSYS. (September 2016)

• Record Type: Journal Article
• Authors: Eguía, P.; Granada, E.; Alonso, J.M.; Arce, E.; Saavedra, A.
• Is Part Of: Journal of building engineering. Volume 7(2016)
• Page Start: 78
• Page End: 91
• Publication Date: 2016-09
• Subjects: Building simulation -- Weather data -- Interpolation -- Kriging; Building -- Periodicals; 690.05
• DOI: 10.1016/j.jobe.2016.05.007
• Languages: English
• ISSNs: 2352-7102

Abstract

Abstract: Transient thermal simulations are a good method for predicting the energy consumption of buildings. Meteorological data are critical for proper simulation as they constitute the boundary conditions of the building. Meteorological data are not always available in the precise building location, and thus, data from the nearest weather station, are used widely. This paper studies the errors in the results obtained using three different techniques based on kriging interpolation (ordinary kriging, universal kriging with an altitude predictor and universal kriging with the altitude and distance from coast predictors) with data from the meteorological station located by the building under consideration. The data obtained from the weather station of Peinador (an airport located seven kilometers away from the building) are also analyzed. A transient thermal simulation is also performed (using TRNSYS software) to calculate and compare the thermal heating and cooling demands in a library building at the University of Vigo, as well as the errors in the heating and cooling demands obtained with the different weather files generated. The kriging technique results in lower errors in the heating and cooling demands of the building thermal simulation, with a 25% reduction in the RMSE compared with using the nearest meteorological station data of the Vigo airport. Highlights: Comparison of different kriging technics for weather data interpolation. Weather data kriging interpolation, applied to building transient simulations. Weather data gaps refilled with kriging techniques, in order to automatically create tmy files. Building transient simulation errors analysis, due to different weather data sources.