Paper Conference

Proceedings of Building Simulation 2021: 17th Conference of IBPSA


Renovating Herentals: a building classification approach to assess large-scale renovation costs

Guillermo Borragán, Dorien Aerts, Glenn Reynders, Yixiao Ma, Lukas Engelen, Stijn Verbeke
VITO-Energy Ville, Belgium

Abstract: As part of the strategy to improve energy efficiency and decarbonize the building stock, the Flemish government has set the target to renovate the residential stock (Vlaamse Regering, 2020) by 2050. However, an old housing stock integrated by a large number of detached buildings suggest that finding a good cost-efficiency balance is neither an easy nor an inexpensive task. Having accurate figures about the costs and benefits of renovation is essential not only to anticipate public aids but also to boost private investment. Earlier studies trying to describe building typologies (e.g. IEE-Tabula - Ballarini, Corgnati, & Corrado, 2014) focused on energy use profiles rather than on renovation potential. The recent availability of new (big)(open) data (e.g. GIS, consumption data, heat maps...) enables the development of machine learning classification techniques to create more accurate building set representations. The purpose of the present study is to develop a massive classification approach to identify the type of renovation plan and the associated costs for the different building typologies in the Flemish region of Herentals. For this purpose, two different machine learning classifiers - supervised and unsupervised - are tested and developed.
Keywords: Classification, Clustering, Building renovation costs, Flanders
Pages: 334 - 341