The use of yearly in situ measurements of a whole commercial building for sensitivity and uncertainty analysis of energy performance assessment

The purpose of this study is to assess the capabilities of a thermal model to represent actual building energy consumption when trying to best fit the input data of the model to the actual data of the building in operation. The approach has been applied to a building for which many factors affecting energy use have been monitored for a whole year. Beyond detailed inspection and advanced investigation have been carried on to assess some uncertain parameters. The objective is also accounting for difficulties related to the inspection of a building in use. The results show the importance of sensitivity and uncertainty analysis on the evaluation of the energy consumption of buildings; within the probalistic frame, energy consumption calculation no more provides a single point estimate but rather a prediction interval with a distribution of the probability density. The study shows the difficulty of collecting and estimating the necessary parameters for the calculation model inputs for existing buildings in operation despite a strong investigation of inspection. Estimation of uncertainty of these parameters is also a long process that needs a strong knowledge.  The overlap of distribution curve of the probability density of energy consumption for the case study between calculation and measurement is done on a small range which is outside the 90/90 tolerance interval of calculation. The main reason seems to be related to the difficulty to reproduce operating principles of HVAC systems into modeling (operating principles of air handling units for this case study) and difficulty to reproduce the real dynamic loads due to their complexity and diversity; many assumptions are used to estimate some of them, others were certainly not taken into account.  Adjusting parameters of building energy simulation model into a calibration process allows to fit the model results to real situation of the building. It’s helpful for energy efficiency services such as periodic verification of the energy performance of the building and continuous operation optimisation. In this field, sensitivity analysis method is necessary in the calibration process to orient the data collection work and to guide the parameters adjustment process [13].