Archetype selection process for the development of a building stock model

The building sector accounts for 40% of the EU’s energy consumption and 36% of greenhouse gas emissions, and is therefore a hotspot for mitigation efforts. Stock modelling is a powerful tool to support the assessment of the impact of the building sector and to test policy interventions to reduce it. A common approach to developing a stock model is through the identification of representative building archetypes. The goal of this paper is to propose a new selection process for archetypes in order to improve the modelling of the residential building stock, and to assess its robustness. The methodology is based on data obtained from the AmBIENCe project. Archetypes are selected for three EU regions, to represent existing buildings in two building categories (detached and multifamily houses). Member State-level archetypes are used as a basis for the identification of regional archetypes, which are selected in order to reduce the number of buildings on which the stock model is developed, while still ensuring its representativeness. A regional average building is defined using a statistical approach to identify its characteristics, considering two sets of parameters related to the building envelope: U-values and materialisation factors (factors indicating the main construction material of a building element). The total useful floor area is used to weigh these parameters among the various Member States. The building types with the smallest overall variation with respect to the regional average are selected as archetypes for the stock model. In a sensitivity analysis, the same process is re-applied, accounting for the relative importance of each envelope element in the selection of the best-representing archetype at regional level by using first their area and then their volume as additional weighting factors. As these methodological changes lead to a different archetype selection, the new and old archetypes are assessed against each other. The analysis provides an understanding of the importance of selection process of archetypes for stock modelling, and therefore contributes to correctly framing the interpretation of any results obtained through the use of a specific model.