Thermal sensitivity of occupants in different building typologies: the Griffiths Constant is a Variable

Rupp, R. F.; Kim, J.; Ghisi, E.; de Dear, R.
Energy and Buildings

The Griffiths method is widely used in thermal comfort studies to derive building users’ comfort temperature, or thermal neutrality as it is sometimes known. A single value (so called the Griffiths Constant, typically 0.5/°C) is prescribed as a representation of thermal sensitivity of building occupants to indoor temperature variations, which in turn is used to estimate indoor thermal neutrality from a subject's actual thermal sensation vote at a measured room temperature. Despite the Griffiths Constant of 0.5/°C having been used widely across the thermal comfort research literature and in some generic standards, the constant was derived exclusively from office building data and its applicability across different typologies is yet to be rigorously validated. The objective of this study is to quantify how sensitive people are to temperature variations inside a building, and to investigate if thermal sensitivity differs between different contexts (including building typologies, modes of ventilation, outdoor climatic types, and genders). A collection of thermal comfort field studies in different building typologies containing around 11,500 datasets was used to statistically derive building users’ thermal sensitivity, i.e. the rate of change in thermal sensation per unit change in indoor temperature within a day. Our results suggest that occupant thermal sensitivity does vary depending on building typologies and building ventilation mode. In naturally ventilated spaces users are about half as sensitive to temperature variations as in air-conditioned spaces. Age, gender and climate are found to be factors that can also influence thermal sensitivity of building occupants. Our findings imply that reliance on a universal thermal sensitivity value, the Griffiths Constant, in comfort temperature (neutrality) calculations should be avoided because it is in fact a variable.