Personal differences in thermal comfort perception: Observations from a field study in Brisbane, Australia

Despite substantial evidence emphasising the diversity in thermal comfort perception, conventional models like the Predicted Mean Vote (PMV) index tend to overlook personal variability. In buildings, this oversight leads to occupant discomfort, dissatisfaction, and inefficient energy use for conditioning. In the past, technological and logistical constraints made personalised field studies an arduous task. However, today recent advancements in low-cost sensor technology allow one to pervasively capture the immediate and personal thermal environment experienced by building occupants. Between summer and winter 2022, we conducted a field study across 13 office buildings in Brisbane, Australia, investigating the personal variability in thermal comfort perception. Throughout the study, we recorded over 135,000 samples of continuous personal indoor environmental data at participants' workspaces and collected 813 occupant survey responses. Analysing the data, we found (1) significant personal variability between subjects, evident in a personal mean discrepancy from the PMV index ranging from −1.5 to +1.5 on the thermal comfort scale; (2) a limited accuracy of the PMV index ranging between 35 and 40%; when predicting the subject's thermal sensation vote, (3) offices with highly heterogeneous thermal indoor environments; (4) dissatisfaction with the workspace due to thermal discomfort and insufficient control of the personal thermal environment. Our findings suggest a high personal and spatial variability in thermal comfort, underscoring the need to consider these factors to improve the accuracy and occupant-focus of thermal comfort indices.