Comparison between PMV and modified PMV indices in the context of thermal comfort assessment and predicted productivity loss

Even though Predicted Mean Vote (PMV) index is widely used for thermal comfort assessment and it is implemented within the international standards, available literature suggest that there are limitations of its application. These limitations lead to the introduction of modified PMV indices such as the extended Predicted Mean Vote (ePMV) and the adaptive Predicted Mean Vote (aPMV), which can, under specific conditions, improve the accuracy of the original PMV. Taking that into consideration, this study aims at analyzing the differences between PMV and its modifications aPMV and ePMV, calculated for different sets of input parameters, while discussing the values of the productivity loss predicted based on them as well, which is the main novelty of this research. In accordance with appropriately defined dataset, the results are presented for PMV values higher than 0, and they suggest that for the input parameters coinciding with PMV values close to thermal neutrality, the differences between PMV and its modifications, as well as those between predicted productivity loss calculated based on them, are negligible from the thermal comfort standpoint. On the other hand, these differences can significantly increase for inputs which coincide with higher PMV values. This increase is heavily dependent on the adopted adaptive coefficient and expectancy factor values, and it can result with different occupants’ thermal sensation prediction. Furthermore, the difference between predicted productivity calculated based on the PMV and ePMV can reach over 20%, reaching over 15% in the case of comparison between PMV and aPMV. This implies that while predicting thermal comfort and, further, productivity loss by using these indices can be beneficial, it is crucial to appropriately select expectancy factor and adaptive coefficient values, especially if the original PMV predicts higher occupants’ thermal sensation.