Hierarchical Clustering and CoClust Algorithm: A Nested Procedure to Analyse Sustainable Heating Data

In this work we combine two different clustering methods to investigate district heating data by taking into account both static and dynamic information concerning the buildings energy profile. The idea is to use the hierarchical clustering algorithm based on the Gower's index to find a first partition of buildings based on their static characteristics, such as age class, energy class, and heating surface, and, next, to investigate the within-cluster multivariate dependence of thermal energy demand among buildings. The two-step procedure we propose aims at assessing the usefulness of static information to support the management of energy demand in the urban area. We show the procedure on data concerning the district heating system of the Italian city Bozen-Bolzano.