Identification of the Minimum Number of Measurements Required for Thermal Comfort Assessment in Large Workplaces

Optimization of resources is the key to improve our ability to perform multiple tasks with limited time and money. In the context of thermal comfort assessment, optimization becomes important in large rooms where tens of individuals perform similar tasks. This work focuses on the identification of the minimum number of measurement points that allows an accurate description of the thermal environment. Accuracy of description is assumed if no significant loss of information is associated to the transition from the 'primary' thermal map based on all available measurement points to a 'secondary' thermal map based on a reduced set of measurement points. The concept of 'no significant loss' is quantified by requiring that the difference in PMV (Predicted Mean Vote) between the two maps is kept <0.1 in the vast majority (95%) of points. PMV is a standardized synthetic index that is used worldwide for quantifying thermal comfort (ISO 7730, 2005) taking into account both environmental (thermo-hygrometric) and personal (activity, clothing) quantities. We show that the uncertainty induced by the degraded resolution of the thermal map has a limited impact on the overall uncertainty on PMV. Application of the method to a few test environments shows that the room size perpendicular to the main window and the windows orientation play the largest role in determining spatial inhomogeneity in thermal maps. A minor concurring factor is the room area.