Nowadays, the building sector accounts for a significant share of the overall energy consumption, up to 40% in the EU. Within this context, academic and public buildings are particularly important. Unfortunately, analysing these types of buildings can be challenging due to multiple factors, such as the coexistence of multiple uses within the same structure (i.e., offices, labs, classrooms), and sometimes the dimensions of the building stock managed by the same public institution. This study highlights in which way a systematic analysis of the real consumption data can drive the efficient energy management of the building stock. In this paper, a methodology of data analysis tested with real data obtained from buildings of the University of Bologna is presented and discussed. A building complex was used as a demonstrator by installing a series of electrical and heat meters to assess the effective energy consumption within the building during the whole year. A Python script was used to automate the analysis of the sub-hourly energy consumption data available from the energy supply companies and the installed meters; it is demonstrated that the extrapolation of a series of key performance indicators useful for optimal energy management of the site is possible. The methodology was further applied to additional buildings of the University of Bologna to examine its applicability in identifying discrepancies between actual and expected consumption, highlighting all the singular behaviours which have to be corrected for the optimal energy management of each specific site of a building stock.

Keywords: Energy consumption, building stock, data analysis, energy efficiency, key performance indicators (KPIs).