Robust evaluation method of thermal deviation of air distribution

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journalpeer-review

6 Scopus Citations
View graph of relations



Original languageEnglish
Pages (from-to)217-225
Journal / PublicationBuilding and Environment
Online published14 May 2019
Publication statusPublished - Jul 2019


Air distribution (e.g., mixing ventilation) is widely used to provide indoor thermal comfort. The thermal deviation is defined as the deviation between the thermal condition and thermal preference, and a thermal deviation closer to zero indicates a more comfortable thermal environment. The thermal deviation is significantly affected by uncertainties in the thermal preference and cooling load. However, the conventional design method of air distribution evaluates the thermal deviation in a deterministic manner, which could result in an improper design with poor thermal comfort. This study innovatively proposes a robust thermal deviation evaluation method and uses it for the design optimization of air distribution. Monte Carlo simulations are employed to treat uncertainties in the thermal preference and cooling load. From the Monte Carlo simulations, the thermal deviation D 0.95 is quantified with a confidence level of 0.95. The thermal deviation D 0.95 should be minimized to optimize the design of air distribution robustly. Tests based on experiments of an energy efficient air distribution, i.e., stratum ventilation, have been conducted to validate the evaluation method proposed. Results show that the robust design based on the evaluation method proposed can ensure thermal comfort, while the conventional design risks thermal discomfort due to the uncertainties. Compared with the robustly optimized variable-air-volume system of stratum ventilation, the robustly optimized constant-air-volume system of stratum ventilation improves thermal comfort by 8.5% with the thermal deviation D 0.95 reduced from 0.682 to 0.624. Therefore, the evaluation method proposed contributes to designing and comparing air distributions robustly for thermal comfort improvement.

Research Area(s)

  • Air distribution, Cooling load uncertainty, Design and comparison, Thermal comfort improvement, Thermal deviation evaluation, Thermal preference uncertainty