Building energy data analysis by clustering measured daily profiles

Nowadays, new methods for the determination of energy efficiency in buildings can be applied since highly-resolved monitoring data of HVAC (heating, ventilation and air conditioning) systems are available. The presented method aims to aggregate similar daily profiles of HVAC-variables into behavior patterns. For the method's validation, one-year monitoring data from a renovated single-family house have been analyzed. The results show that the daily profiles of different HVAC variables are usually aggregated into 4 to 8 groups of typical and untypical behavior patterns. Obtained behavior patterns were examined more closely and pointed out a performance gap for several systems.