Operational Performance and Calibrated Model Analysis of a Large Low-Energy University Building

This paper demonstrates a reproducible methodology for calibrating detailed energy models using hourly measured data that has been applied to evaluation of a large naturally ventilated university building. The aim of the project is to develop a rigorous calibration method and use it to investigate Energy Conservation Measures (ECM) and retrofit renewable energy technologies to achieve carbon emissions reduction. The methodology is based on a standardized day-typing method using data visualization and statistical techniques to establish accurate schedules for non weather-dependent demands. The influential weather-dependent sources of output uncertainty have been studied using two forms of sensitivity analysis. It was found that applying the Modified Morris method resulted in effective calibration with a manageable quantity of annual simulations and allowed some insight into the physical behaviour of the building. We also show how data visualisation has allowed further refinement of the model.