Abstract
Energy audits of existing buildings are especially important for public buildings and, in particular, for schools, where a more efficient use of energy implies unquestionable benefits to public budgets. Schools audit can drive the public administrator to better address retrofit investments, facilitating the choices of energy efficiency measures in the renovation or operation phases. However, the energy audit of existing buildings can be onerous when the number of buildings is large and requires extensive monitoring campaigns, field surveys and energy performance calculation. A simplified method for building energy diagnosis is the “Energy Signature” (ES) method described in the annex B of standard EN 15603:2008. According to this approach, heating and cooling energy uses of a given building are correlated to climatic data over a suitable period. Plotting for several time periods the average heating or cooling power versus the average external temperature provides useful information on building energy performance and allows a fast detection of malfunctions or changes in the building operation/features, as well as the verification of the efficacy of any retrofit intervention. Although this method is preferably adopted in the case of constant internal temperature (e.g., fixed temperature setpoint) and when the external temperature is the most influential parameter (e.g., for buildings with stable and relatively low internal and passive solar gains), it can be applied also recording energy use for heating or cooling and accumulated temperature difference between indoor and outdoor, at average regular intervals (e.g., one hour or, for manual monitoring, a week). The ES is the best fitting linear regression between these two quantities and, consequently, can be characterized by means of intercept and slope.
In this paper, the building energy signature parameters have been used to analyze a large set of school buildings and to define the characteristics most influential on the energy needs. In particular, the weekly energy consumptions for heating of a set of 42 school buildings located in the province of Treviso, North East of Italy, have been considered. A cluster analysis based on multiple regression has then been used to identify the buildings’ subsets homogeneous as for the features affecting the signature parameters.