Calibrating a Clothing Insulation Model for Thermal Comfort Assessment in Educational Buildings

Thermal comfort assessment in buildings usually relies on the calculation of Predicted Mean Vote (PMV) which is determined by four environmental variables, such as air temperature, air humidity, air velocity and mean radiant temperature, and by two personal factors, namely the metabolic rate and the clothing level. The latter factor is fundamental in determining the thermal sensation since it can be changed and adapted in response to the indoor conditions, thus allowing an extension of the neutral temperature range. Moreover, the uncertainty of clothing level in simulated models can affect the reliability of results in terms of thermal comfort and overall IEQ assessment. This study aims to calibrate existing models on an extensive set of data collected in an Italian high school located near Rome and to build a new clothing model. The outdoor and indoor environmental conditions in 22 natural ventilated classrooms were monitored during the school years 2020-2022. Students’ thermal sensation votes and the corresponding clothing levels were surveyed during regular lessons. First, the physical variables used in the literature to predict clothing insulation were at first analyzed to highlight the significant ones based on the collected data. Second, the significant physical variable (i.e., operative temperature) was used as input to feed existing models and to predict clothing insulation; the predicted values were then compared with the observed mean clothing insulation of the students in each classroom. Third, a calibration of a clothing linear model based on operative temperature was carried out and a new linear model based on the indoor running mean temperature was set. Finally, to explore to which extent the linear clothing model based on Top can affect the thermal comfort simulation, the Predicted Mean Vote (PMV) was calculated.