Applying cluster analysis to identify target buildings for energy retrofit: An alternative to change-point model
Identifying energy-inefficient buildings is challenging, but critical for large-scale energy retrofit programs. Change-point model (CPM) is widely used to characterize energy performance of buildings; however, its application is limited to buildings with consistent energy use. This study proposes a...
Uložené v:
| Vydané v: | Energy and buildings Ročník 351; s. 116742 |
|---|---|
| Hlavní autori: | , , |
| Médium: | Journal Article |
| Jazyk: | English |
| Vydavateľské údaje: |
Elsevier B.V
15.01.2026
|
| Predmet: | |
| ISSN: | 0378-7788 |
| On-line prístup: | Získať plný text |
| Tagy: |
Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
|
| Shrnutí: | Identifying energy-inefficient buildings is challenging, but critical for large-scale energy retrofit programs. Change-point model (CPM) is widely used to characterize energy performance of buildings; however, its application is limited to buildings with consistent energy use. This study proposes a method for applying cluster analysis to determine energy-inefficient buildings, addressing CPM limitations. Considering the scarcity of high-frequency energy data for existing buildings, monthly gas consumption data from 344 apartment buildings in Ulsan, South Korea, were used. The study was conducted in two phases. In Phase I, CPM and k-means algorithm were used to identify energy inefficient buildings from 114 buildings meeting CPM requirements. Phase II expanded the method to all 344 buildings, and the identified energy retrofit target buildings were analyzed. Results from Phase I indicated consistency in the buildings identified as energy inefficient by the k-means and CPM methods. When applied to the entire dataset, including buildings excluded by CPM requirements, k-means algorithm effectively identified buildings in need of energy retrofit regardless of consistency in energy use. With a silhouette score of 0.72, the clustering algorithm identified 21 buildings with monthly gas consumption 2.6 kWh/m2 above average, indicating well-structured clusters and effective identification of energy-inefficient buildings. These findings establish a basis for the use of cluster analysis to overcome the limitation of CPM application for large-scale building energy assessment. The proposed method presents a potential application in facilitating the implementation of energy upgrade of existing building stock by prioritizing the best targets for energy retrofit. |
|---|---|
| ISSN: | 0378-7788 |
| DOI: | 10.1016/j.enbuild.2025.116742 |