Multivariate Data Analysis of the Thermal Performance of Portuguese Residential Building Stock


The purpose of this study is to explore the relationship between solar orientation, age (constructive characterization) and energy performance of Portuguese residential building stock and to assess the usefulness of exploring the Portuguese National System for Energy and Indoor Air Quality Certification of Buildings (SCE) database through multivariate analysis techniques. By using principal components technique, it was possible to condense the residential units’ features to only four principal components (PC): solar orientation; constructive characterization; geometry and energy performance, making information more workable. Grouping the entities into Clusters with favourable and unfavourable solar orientation and old buildings allowed to dilute the particularities of each entity, facilitating the interpretation of the data through generalization. A regression model was generated in order to explore/confirm which factors influence summer comfort the most. Using this approach, it was illustrated that the exploration of the SCE database through multivariate data analyses has an enormous potential to convert data into knowledge.

[1] J. F. Hair, W. C. Black, B. J. Babin, and R. E. Anderson, Multivariate Data Analysis: Pearson New International Edition., Seventh Ed. (Pearson, Harlow, Essex - UK, 2014).

[2] Concerted Action EPBD, Implementing the Energy Performance of Buildings Directive - Featuring Country Reports 2012, no. 1. (ADENE, Porto - PT, 2013).

[3] R. B. Cattell, ‘The Scree Test For The Number Of Factors’, Multivariate Behav. Res., vol. 1, no. 2, p. 2010,(1966).

[4] E. Reis, Estatística Multivariada Aplicada, 2a Edição. (Edições Sílabo, Lisboa – PT, 2001). (Portuguese)