Comfort, preferences or design data?

Abstract

Buildings the lusually expected by their occupants to do more than simply provide ‘shelter’ and a significant proportion of energy use is consumed for heating and cooling buildings, in pursuit of some degree of comfort, or at least discomfort minimisation. There have been many efforts to define the conditions required for thermal ‘comfort’. Much of the thermal comfort research has been conducted under laboratory conditions and based on mechanistic, non-contextual paradigms. A prime example is the work of Fanger which forms the basis of the ISO 7730 method for predicting mean thermal sensation, a method which is extended to providing design data for heating and air-conditioning. Thermal comfort data is also an essential requirement in computer simulations used for the development of house energy rating schemes and energy efficiency standards. The use of incorrect assumptions regarding thermal preference can lead to inappropriate building solutions. Thermal comfort models which relate predictions to the real context of the building and its occupants are required. This paper presents the results of analyses of more than 48 000 thermal sensation votes colleced in recent Australian field studies. Thermal sensation votes are compared to the ISO 7730 PMV predictions and consistent differences are shown. The notion of neutrality seeking, a concept central to standard thermal comfort theories is questioned. Overall the empirical results indicate that the standard methods of thermal comfort which provide a basis for the advice to designers are not sensitive to important context relevant factors. The paper concludes with a brief discussion of the implications for the provision of thermal information for designers.