A Field-Based Methodology for Thermal Comfort Evaluation: A Case Study on University Classrooms

Ensuring thermally comfortable indoor environments is essential for occupant well-being, learning performance, and energy efficiency. This study proposes a field-based methodology for the assessment of thermal comfort, aligned with international standards and designed to be reproducible and applicable across diverse building types and climatic conditions. The objective methodology integrates both objective (measurements) and subjective (surveys) methods, combined with behavioral observations, to provide a detailed assessment of indoor thermal comfort conditions. The methodology integrates long-term and short-term measurements of air temperature, mean radiant temperature, relative humidity, air velocity, and along with the calculation of PMV/PPD. The data reliability is verified through calibration, inter-device agreement, and statistical evaluation using Mean Bias Error (MBE) and coefficient of determination (R²). Furthermore, occupant surveys and camera-based observations capture subjective perceptions and adaptive behaviors, linking measured values with real-world occupancy conditions. The proposed methodology was applied to three amphitheater-style university classrooms characterized by high window-to-wall ratios on the south-facing façade. Findings showed that thermal gradients near south windows, confirmed by surveys reporting higher dissatisfaction and reduced concentration in south façade-adjacent zones. Observations revealed limited adaptation, with windows seldom opened and curtains generally closed, and underscoring results indicated the need for passive or automated strategies. The results confirm the methodology’s ability to capture spatial and temporal variations, link predictive indices to occupant perception. Overall, the study offers a standards-based methodology suitable for assessing thermal comfort in different building type and climate.

Keywords: Thermal Comfort, Field-based Methodology, Case Study, PMV/PPD.