SE-14-C038 -- Targeted Calibration of Energy Models for Existing Building PDF

Calibration is one of the important areas in building energy simulation. It compares the result of energy simulation with the metered energy consumption of the building. For new construction, the calibration process typically happens long after the design stage. So long that in many situations the as-designed energy model is never updated into an as-built calibrated energy model. In recent years, however, the focus of energy efficiency programs has shifted toward existing building. In the case of existing building retrofit, a calibrated energy model is not only essential but it becomes the basis of design. Furthermore, the calibrated energy model can also be used to tune up building operational strategies.

Many standards have been published to ensure the quality of the calibration result. ASHRAE Guideline 14 and IPMVP are the two most widely used standards. These standards however focus on how to express the error of the calibration result, and they do not provide any guideline on how to actually perform a calibration. A simulation engineer will have 12 months of metered energy data and 12 months of predicted energy consumption from the simulation. Based on the difference of the two sets of data, the engineer is expected to turn some dials (of simulation parameters) to minimize the error between the two sets. What is typically forgotten is that there are hundreds of dials that the engineer can turn. There is no systematic guideline on which dial to turn.

This paper proposes to use the ASHRAE's Inverse Modeling Technique (IMT) as the basis of the targeted calibration process. In this process, IMT calculates two energy signatures of the building based on the meter and the simulation. The difference between the two energy signatures will tell us not only the difference in thermal characteristics between the actual building and the energy model, but also the difference in the operational parameters. Based on that information, the calibration process becomes more systematic and targeted toward changing relevant building parameters. A number of case studies will be presented as a proof of concept on how the targeted simulation is performed in a real life.