IAQ4EDU

2025

1. Indoor Air

   

   Assessing the fluctuation of indoor thermal conditions in naturally ventilated classrooms through K-means clustering

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Indoor Air, Mar 2025

   Air change rate Indoor thermal condition Natural ventilation Window and door operation Clustering analysis

   Abs DOI PDF

   For naturally ventilated classrooms, windows should be constantly open to maintain good indoor air quality during class times. However, it is unknown whether this would result in significant fluctuations in the indoor thermal environment, while the effects of ventilation rate and window operation are still unclear. Therefore, this study assesses the fluctuation of indoor thermal conditions under natural ventilation during class periods, based on the K-means clustering and statistical analysis of data collected in a large measurement campaign in Spanish schools. The results showed that regardless of the heating mode, the indoor thermal parameters remained stable in most cases. High fluctuations were observed only in a few cases: 15% for indoor temperature, 6% for relative humidity, and 3% for indoor air velocity. The variability of the indoor thermal parameters within an hour was less than 1°C for indoor temperature, 3% for relative humidity, and 0.1 m/s for indoor air velocity. Ventilation rate and window operation were found to be related to the fluctuation of the indoor temperature and air velocity, rather than the relative humidity. These findings shed light on the thermal conditions in naturally ventilated classrooms and have practical implications for the implementation of natural ventilation protocols in schools.

2. Build. Environ.

   

   Investigating students’ subjective comfort with window-airing during the cold season: Thermal sensation, humidity, air movement, and perceived air quality

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Building and Environment, Jun 2025

   Natural ventilation Thermal comfort Subjective sensation vote Perceived air quality Field experiment Window airing

   Abs DOI PDF

   Balancing the need for indoor air quality and thermal comfort during cold seasons is challenging for naturally ventilated schools. For students’ health concerns, opening windows to ventilate during class time could be inevitable in winter, especially during flu season. However, there is a lack of field studies investigating its potential impacts on students’ comfort. In this context, this study investigated students’ subjective comfort with window airing in winter through a field study in the Mediterranean climate, with a total of 34 field experiments with different window opening scenarios conducted in two university classrooms. The study analyzed the effect of thermal sensation, humidity, air movement, and perceived air quality on students’ comfort and identified correlated environmental parameters. The results showed that students’ comfort was mainly determined by thermal sensation and air movement, while they were not very sensitive to indoor humidity and air quality. Their thermal sensation was found to be directly determined by the indoor temperature and not correlated with the outdoor temperature. Achieving the required ventilation rate would not cause obvious discomfort in terms of air movement. Moreover, the study validated 4 main adaptive PMV models (nPMV, adPMV, arPMV, and epmPMV) and found them to be ineffective with forced window openings. The proposed comfort prediction model suggests that to meet the ventilation rate requirement, the indoor temperature should be maintained above 21 °C to avoid causing thermal discomfort problems. The specific practical recommendations are given in the conclusions.

3. J. Build. Eng.

   

   A novel approach to calculate the mean thermal sensation vote for primary and secondary schools using Bayesian inference

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Journal of Building Engineering, Apr 2025

   Thermal comfort Field survey Educational building Mean thermal sensation vote Bayesian inference

   Abs DOI PDF

   Existing thermal comfort models defined in relevant standards are often found to be less effective for primary and secondary school students in educational buildings. This is often thought to be primarily due to differences in thermal sensation between children and adults. However, one important factor that is often neglected is the uncertainty associated with thermal comfort survey data. The existing method for calculating the mean thermal sensation vote is oversimplified and does not properly address related uncertainties. As a result, it ultimately affects the performance of the developed thermal comfort models. Hence, this research proposes a novel approach to compute the mean thermal sensation vote data for primary and secondary schools using Bayesian inference. This approach addresses the error caused by the uncertainties associated with the collected thermal sensation vote data in order to improve the effectiveness of the developed thermal comfort models for students. The proposed method was validated through a holistic case study using five thermal comfort models. The results showed that the accuracy of the developed thermal comfort models improved by 10.1 %–30.9 %, and the R2 improved by 5.3 %–28.8 %. A benchmark for the Bayesian model parameter setting was proposed as the reference for relevant studies. Finally, an open, user-friendly software was developed and is available to relevant users to implement the proposed approach more efficiently. The results of this research have practical implications for the development and optimization of thermal comfort models for students.

4. Build. Environ.

   

   Validating single-sided natural ventilation models for educational buildings

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Building and Environment, Aug 2025

   Natural ventilation Single-sided ventilation model Educational building Field experiment Window opening

   Abs DOI PDF

   To maintain good indoor air quality in schools, it is often necessary to apply single-sided natural ventilation models using urban weather stations, since the local weather station is not available for most schools. However, the field study that validated the ventilation models with full-scale experiments in classrooms is very limited. Therefore, this study validated 6 existing single-sided ventilation models based on two field measurement campaigns conducted in naturally ventilated educational buildings, including the Warren & Parkins (1977 & 1985), De Gids & Phaff (1982), Larsen & Heiselberg (2008), Caciolo (2013), Tang (2016), and EN16798–7 (2017) models. The study also analyzed the cause of model prediction errors with respect to building characteristics such as weather station distance, building location, room floor, and incident wind direction. The results show that the distance between the school and the weather station has a significant impact on the model prediction accuracy. The prediction error generally exceeded 50 % with urban weather station data, while the actual ventilation rate was overestimated severely. The buoyancy and wind effects were overestimated by more than 2 times, due to the influence of the urban heat island effect and the complex built environment in the city. In addition, the prediction error related to the room floor was caused by the difference in environmental parameters, while the error related to the incident wind direction was mainly due to the limitations of the existing models. The results of this study provide valuable insights for the practical application and further improvement of ventilation models.

5. Build. Environ.

   

   NVAPF: An adaptive particle filter algorithm for CO2-based natural ventilation rate estimation with high temporal resolution and stability

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Building and Environment, Sep 2025

   Natural ventilation rate CO2 tracer gas Bayesian filter Adaptive particle filter Educational buildings

   Abs DOI PDF

   CO2-based ventilation rate estimation techniques have been widely used in relevant studies in naturally ventilated educational buildings. Such techniques are non-invasive, low-cost, simple, accurate, and do not interfere with the activities of indoor occupants. However, the estimation is significantly affected by the CO2 measurement noise, the uncertainties associated with CO2 generation rate, and complex natural ventilation dynamics. Existing techniques were found to have limited capabilities to deal with these aspects. Therefore, this research proposed an adaptive particle filter algorithm for CO2-based natural ventilation rate estimation and validated it through a case study in an educational building. Compared with the existing transient mass balance model and the extended Kalman filter technique, the estimation stability has been improved by nearly 6 times and 3 times, respectively. More importantly, the proposed algorithm is significantly more robust to abrupt changes in indoor CO2 and can effectively avoid large drifts in the estimated ventilation rate due to sudden window opening and sudden changes in room occupancy, demonstrating great practical applicability for real-time estimation with a high temporal resolution of 1 minute. To help relevant users with practical applications, the study also analyzed the algorithm parameter settings and the impact of simplification strategies commonly used in relevant studies, such as the use of a fixed outdoor CO2 concentration, an averaged CO2 generation rate, and an assumed constant room occupancy. Finally, considering that applying the proposed algorithm requires programming skills, an open, user-friendly software has been developed for relevant users for a convenient implementation.

6. J. Build. Eng.

   

   A long short-term memory physics-informed neural network model for CO2-based natural ventilation rate estimation

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Journal of Building Engineering, Nov 2025

   Natural ventilation rate CO2 tracer gas Kalman filter Long short-term memory Physics-informed neural network

   Abs DOI PDF

   The occupant-released CO2 tracer gas approach has been widely used for ventilation rate estimation. This approach is non-invasive, low-cost, and does not interfere with the occupants’ activities. However, the CO2 measurement noise and CO2 generation uncertainties can significantly affect the accuracy of the estimated ventilation rates, while the dynamics of the natural ventilation rates could challenge the stability of the estimators. As commonly applied techniques, the moving average filter and the extended Kalman filter have their own advantages and limitations in addressing these issues. To further address the challenges in the natural ventilation rate estimation, this paper proposes a novel model named “NVR-PINN”, based on the long short-term memory-physics-informed neural network and validates it with a case study. The proposed model combines the strengths of the moving average filter and the extended Kalman filter, demonstrating better practical values. It is capable of handling both CO2 measurement noise and CO2 generation uncertainty, effectively capturing the temporal dynamics of the natural ventilation rate, while processing the entire time series observation with a defined sequence window to yield more stable and consistent estimates. The analysis of the case study also revealed useful evidence for relevant research with regard to the applicability of existing ventilation rate estimation techniques.

2024

1. Build. Environ.

   

   Improving the thermal comfort model for students in naturally ventilated schools: Insights from a holistic study in the Mediterranean climate

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Building and Environment, Jun 2024

   Adaptive thermal comfort Predicted mean vote Field survey Educational building Thermal sensation correlation analysis Machine learning

   Abs DOI PDF

   The thermal comfort models defined in existing standards are usually ineffective in predicting students’ mean thermal sensation. Accordingly, overheating and overcooling are frequently reported, which affect students’ well-being and cause unnecessary energy waste. Existing studies have not further explored the model improvement strategies for educational buildings. Hence, this research aims to improve the mean thermal sensation prediction model for students in naturally ventilated schools by incorporating correlated factors, based on a holistic field survey in the Mediterranean climate. The results indicate that the modified extended predicted mean vote of thermal adaptations reinforced around thermal neutrality has the highest accuracy among all validated adaptive predicted mean vote models. However, the performance of the adaptive predicted mean vote models is limited by the deficiencies of the original predicted mean vote model. The overall performance of the improved models was doubled after introducing the identified correlated factors, but the intrinsic error of the predicted mean vote index can slightly affect the models. As a result, the improved model built directly with the identified correlated factors achieved the best performance. Moreover, the potential of the machine learning models is limited by the volume of the mean thermal sensation vote data that can often be obtained, therefore conventional models such as multiple linear regression can be considered an effective and practical tool for developing the mean thermal sensation vote models. The findings of this study benefit the well-being of students and the energy conservation of educational buildings.

2023

1. Indoor Air

   

   A Comprehensive Assessment of Indoor Air Quality and Thermal Comfort in Educational Buildings in the Mediterranean Climate

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Indoor Air, Nov 2023

   Indoor air quality Thermal comfort Natural ventilation On-site measurement campaign Clustering analysis

   Top cited article Abs DOI PDF

   Top Cited Article Award by Indoor Air in 2023

   Maintaining good indoor air quality and thermal comfort is a challenge for naturally ventilated educational buildings, as it can be difficult to achieve both aspects simultaneously. Nonetheless, most of the existing studies only focus on one aspect. To explore the potential of balancing indoor air quality and thermal comfort, both topics must be investigated concurrently. This study assessed indoor air quality and thermal comfort in 32 naturally ventilated classrooms of 16 primary and secondary schools in the Mediterranean climate, based on a large on-site measurement campaign lasting one year that gathered over 460 hours of data. The research investigated occupants’ adaptive behaviors, analyzed the actual thermal comfort of around 600 students, and characterized the representative scenarios leading to good and poor indoor air quality and thermal comfort by clustering analysis. The results showed that poor indoor air quality was mainly due to closing windows and doors in winter, while thermal discomfort mainly occurred in summer because of the high indoor temperature. The findings suggested that a proper ventilation protocol is the key to balancing indoor air quality and thermal comfort.

2. J. Build. Eng.

   

   Data-driven model for predicting indoor air quality and thermal comfort levels in naturally ventilated educational buildings using easily accessible data for schools

   Sen Miao^*, Marta Gangolells, and Blanca Tejedor

   Journal of Building Engineering, Dec 2023

   Machine learning Natural ventilation Windows and doors operation Indoor environment Educational buildings

   Research excellence award Abs DOI PDF

   Research Excellence Award by City Council of Terrassa, Barcelona in 2023

   Modeling indoor air quality and thermal conditions in educational buildings is significant for protecting students’ health, well-being, and productivity. The predictive models in existing studies were mainly built and applied in controlled environments with HVAC systems. These models did not involve occupant-related factors, had limited scope in a single building or space, and required indoor environmental monitoring data for the model input. This limits the applicability and generalization ability of the model in a large number of schools relying on natural ventilation, where indoor air quality and thermal conditions are significantly affected by occupants’ activities and ventilation practices. Hence, this paper proposes a methodology to develop a data-driven model for predicting the level of indoor air quality and thermal comfort in naturally ventilated educational buildings, and identifies the key influential factors. The model was developed using the class-weighted random forest algorithm with data collected from a large measurement campaign. The developed model demonstrated good accuracy, generalization ability, and robustness. The analysis concluded that occupancy, windows and doors operation, and outdoor environmental parameters are key factors must be involved, whereas building characteristics have no practical contribution to the prediction. The model inputs are easily accessible data for schools. Once the model is developed with data collected from an initial measurement campaign in representative educational buildings, it can be used in all local schools without requiring monitoring sensor networks, thereby rendering a “cost-effective” way of assessing indoor air quality and thermal comfort to help relevant stakeholders improve building management practices in schools.