Methodological approach for calibration of building energy performance simulation models applied to a common "measurement and verification" process

A big challenge for energy savings performance contracting projects is the transparent and reliable determination of the energy consumption before and after retrofitting. This process, the so called measurement and verification (M&V) process, can be performed by means of calibrating simulation models with measured data, e.g. the energy bill data. There are guidelines for reporting and conducting this procedure. According to reports from case studies, it seems that they are carried out using several different approaches. The main goal of this work is to develop a consistent, practical approach for the M&V process. The approach supports visual inspection methods, parametric studies, and optimization methods. The visual inspection method can help to understand characteristics of a specific building depending on an input parameter change via graphs. However, the visual inspection method requires an extensive effort for this trial and error process and it is strongly dependent upon the users' experience. The automated parametric study can be applied both for calibration and sensitivity analysis of uncertain parameters. The big challenge for the practical use is the need for automating the process due to an enormous number of simulation cases and the required skills in applied statistics. Optimization algorithms quickly provide the user with a quantitatively best solution. However, they do not analyze the actual problems and therefore do not contribute to an understanding of the system under consideration. Furthermore, results should be examined carefully because the optimization may produce mathematically correct but physically meaningless results. Due to the present advantages and disadvantages of these methods, a calibration tool including all three methods would be desirable in order to predict the impact of energy efficient retrofitting projects.