Analysis of the impact of using synthetic data correlated with measured data on the calibrated as-built simulation of a commercial building

Abstract

Measured data has shown to be useful in improving the accuracy and reliability of building energy simulations. However, a certain level of measured data is often not available due to time and cost limitation. This study analyzes the impact of using synthetic data that was correlated to measured data on calibrated as-built simulation and demonstrates its use for a case-study building, such as synthetic weather-normalized cooling energy use derived from measured motor control center (MCC) data and synthesized direct normal solar radiation from measured global solar radiation. The model recalibration of the case-study building is also discussed with respect to using the proposed calibration factors, including: a weather data file with the synthesized direct normal solar radiation, internal loads and schedules, maximum supply air temperature, hot deck and cold deck air temperature, and chiller operation and preheat temperature. As a result, the recalibrated simulation was determined to have an overall daily 23.82% CV(RMSE) and a daily 0.41% MBE. Consequently, the synthetic models that were strongly correlated to measured data and the calibration factors proposed in this study are found to be effective for calibrating whole-building energy simulation within an acceptable level of accuracy.