Automatic and rapid calibration of urban building energy models

Abstract

Urban building energy modeling (UBEM) is attracting increasing attention in the energy modeling filed. Unlike modeling a single building using detailed building systems information, UBEM generally uses existing limited building stock data consisting of high-level building information. Default assumptions are made for the building systems based on that information, such as building type, size, and vintage. This study introduced a novel method of automatic and rapid calibration of UBEM based on the annual site and source energy use by learning the correlations between key model input parameters and the building energy use from USDOE reference buildings. A case study was presented to calibrate 112 large office buildings built before 1978 in San Francisco. An EnergyPlus model was created for each building. Fifteen model parameters were selected to calibrate the models. The Monte Carlo sampling was used to create 1000 samples to represent the parameter combination space reasonably. Then 1000 simulations were performed for the DOE reference building for large office built before 1978 in San Francisco to create an energy performance database. The results showed that by learning from the energy performance database, it took 2.7 simulation runs on average to calibrate each building model rapidly. After the calibration, the distribution of each parameter was obtained to replace the single default value. For example, the lighting power densities after calibration were 9.87 W/m2 on average, very different from the default value of 21.39 W/m2. The case study successfully demonstrated the effectiveness of the novel calibration method for UBEM in the mild climate. For future work, the method needs to be tested in other climate zones and for other building types.