The colours of comfort: From thermal sensation to person-centric thermal zones for adaptive building strategies.

Guide

Date of Issue

Jun.2020

Type

Article

Guide

Subject Keywords

Barcode

Class number

Abstract

Thermal comfort research has been traditionally based on cross-sectional studies and spatial aggregation of individual surveys at building level. This research design is susceptible to compositional effects and may lead to error in identifying predictors to thermal comfort indices, in particular in relation to adaptive mechanisms. A relationship between comfort and different predictors can be true at an individual level but not evident at the building level. In addition, cross-sectional studies overlook temporal changes in individual thermal perception due to contextual factors. To address these limitations, this study applied a longitudinal research design over 8 to 21 months in eight buildings located in six countries around the world. The dataset comprises of 5,567 individual thermal comfort surveys from 258 participants. The analysis aggregated survey responses at participant level and clustered participants according to their thermal sensation votes (TSV). Four TSV clusters were introduced, representing four different thermal sensation traits. Further analysis reviewed the probability of cluster membership in relation to demographic characteristics and behavioural adaptation. Finally, the analysis at individual level enabled the introduction of a new metric, the thermal zone (Zt), which in this study ranges from 21.5 °C to 26.6 °C. The thermal sensation traits and person-centric thermal zone (Zt) are a first step into the development of new metrics incorporating individual perceived comfort into dynamic building controls for adaptive buildings.

Identifier URL

Pages

Faculty

Faculty of Design

Citation

Source URL

Keywords

DOI

https://doi.org/10.1016/j.enbuild.2020.109936

ISBN

Place of Publication

Publication Year

Edition

Provider

ISSN Number

Volume Number

Issue Number

Publication

elsevier

Class Number

Collections