Analyzing and modeling thermal complaints in a commercial building in France

Abstract Buildings are interactive environments in which their operations and occupants are linked. Although buildings are operated according to the standards, occupant complaints may arise when there is a mismatch between indoor environmental conditions and actual user needs. Therefore, the accuracy of thermal comfort prediction models suggested by the standards and alternative prediction models need to be investigated. This study aims at assessing the performance of the predicted mean vote (PMV) model suggested by the ISO 7730 Standard to detect occupant thermal dissatisfaction. In addition, a multivariate logistic regression model was developed to predict thermal complaints with respect to “too warm” and “too cold.” This case study was conducted in a commercial building located in Paris, France, between January 2017 and May 2018. Indoor environmental conditions were monitored via sensors and an online tool was used to collect occupant thermal complaints. A total of 53 thermal complaints were analyzed. The results showed that all the operative temperature measurements in both the heating and cooling seasons were within the thresholds suggested by the standards. The PMV method suggested that only 4% of the occupants were dissatisfied with the indoor environment whereas the actual dissatisfaction ratio was 100% under these indoor environmental conditions. In addition, the multivariate logistic regression model showed that operative temperature and season have a significant effect on thermal complaints. Furthermore, the accuracy of the developed model was 90.6%.