Integration of the Adaptive Approach in HVAC System Operation: A Case Study

Academic Editor: Paulo Santos Received: 21 December 2024 Revised: 13 January 2025 Accepted: 22 January 2025 Published: 25 January 2025 Citation: Aparicio-Ruiz, P.; Ragel-Bonilla, J.C.; Barbadilla-Martín, E.; Guadix, J. Integration of the Adaptive Approach in HVAC System Operation: A Case Study. Appl. Sci. 2025,15, 1243. https://doi.org/ 10.3390/app15031243 Copyright: © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/ licenses/by/4.0/). Article Integration of the Adaptive Approach in HVAC System Operation: A Case Study Pablo Aparicio-Ruiz * , J. C. Ragel-Bonilla , Elena Barbadilla-Martín and José Guadix Grupo de Ingeniería de Organización, Escuela Técnica Superior de Ingeniería, Universidad de Sevilla, Camino de los Descubrimientos S/N, 41092 Seville, Spain; [email protected] (J.C.R.-B.); [email protected] (E.B.-M.); [email protected] (J.G.) *Correspondence: [email protected] Abstract: Although different investigations have been carried out on the analysis of adaptive thermal comfort in naturally ventilated buildings, fewer have focused on mixed mode operation. Moreover, there is limited research as for the implementation of adaptive comfort models into the control system of buildings. Therefore, this paper investigates how the application of a setpoint based on adaptive comfort control (ACC) would affect occupants’ comfort considering mixed mode operation and based on the results of a longitudinal field study in an academic office building of a tertiary educational institution in southern Spain. The manuscript analyses the Thermal Preference Vote over 12 months in a mixed mode room with an HVAC system whose setpoint is adjusted with a previously calculated adaptive algorithm for the building. For that, a thorough analysis was conducted in which users identified situations regarding thermal comfort and the operation of the conditioning system was collected. The results indicate that it is possible to develop adaptive comfort models that ensure the thermal well-being of occupants. Moreover, this study highlights the need for further research to assess the implications of ACC in terms of comfort and energy consumption as well as addressing the future improvements and the limitations of the work carried out. Keywords: field study; adaptive thermal comfort; mixed mode buildings; adaptive comfort control 1. Introduction An essential function of buildings is to provide comfortable indoor environments, which often entails high energy consumption. In particular, the interaction of occupants with building controls has been recognized as an important factor that influences commonly observed significant differences between actual and predicted energy consumption in buildings [ 1 ]. As for the field study of thermal comfort, research has traditionally adopted two main approaches [ 2 ]: the heat balance model [ 3 ] and the adaptive model [ 4 – 6 ]. In particular, with respect to the latter, the main international standards [ 7 ] establish admissible indoor temperature thresholds in buildings without mechanical ventilation based on comfort studies and over the past ten years, numerous field studies have been conducted along these lines [ 8 – 10 ]. Therefore, adaptive comfort has established itself as a major research field, defining standards that relate acceptable indoor temperatures to outdoor temperatures. These proposals reflect how climate influences our thermal preferences and how body regulation mechanisms drive thermoregulatory behaviors. As J.F. Nicol pointed out, it could be summarized in the adaptive principle: “If a change occurs such as to produce discomfort, people react in ways which tend to restore their comfort” [ 11 ]. J.F. Nicol also highlights that the Appl. Sci. 2025,15, 1243 https://doi.org/10.3390/app15031243 Appl. Sci. 2025,15, 1243 2 of 17 adaptive approach “to explain the range of temperatures people found comfortable in buildings with the variable indoor temperature’s characteristic of naturally ventilated buildings” offers a complementary solution to the challenge of heat balance. Since then, this approach has gained recognition, consolidating into standards such as ANSI/ASHRAE 55:2023 [ 2 ], the reference for thermal comfort in North America, and the European standard UNE-EN 16798-1:2020 [ 4 ], which covers both thermal comfort and other indoor environmental parameters. These adaptive approaches are mainly applied to the assessment of the thermal environment in naturally ventilated (NV) buildings where, e.g., the occupants have access to operable windows and have some freedom to adapt the insulation of their clothing. Countries such as the Netherlands [ 12 ] and Brazil [ 13 , 14 ] have also included adaptive models in their national regulations and standards. Utilizing natural ventilation in buildings can lead to considerable electrical energy savings for cooling and fans [ 15 ]. While adaptation is described as the gradual decrease in the reaction to environmental stimulation, and can be behavioral (such as clothing [ 16 ], windows, fans), physiological (acclimatization), and psychological (expectation) [ 17 ], in a practical setting, variations in the perception of the thermal environment were identified between occupants of naturally ventilated or free-running, fully air-conditioned and mixedmode or hybrid buildings [17]. Adaptive comfort approach have arisen as a sustainable and resilient solution that, according to Bienvenido-Huertas et al. [ 18 ], could reduce the incidence of fuel poverty, in contrast to the static model that relies exclusively on active cooling. Harold Wilhite [ 19 ] described fully conditioned buildings as “Buildings as comfort capsules” where the heating, ventilating and air-conditioning (HVAC) system is seen as one of the tools contributing to the “homogenization of people”, resulting in “thermal monotony”, although early HVAC engineers strived to avoid creating monotonous thermal environments [ 20 ]. The building sector is looking for sustainable building solutions, not only with passive heat dissipation capabilities [ 21 ] or from an aerothermal point of view [ 22 ], but also passive actions are expected to contribute to periods with natural ventilation, such as the adaptation of the clothing insulation level [23]. The application of adaptive comfort control (ACC) as an alternative to fixed setpoint systems (or the paradigm of a single temperature setpoint in each climatic season) would be justified by Fountain et al. [ 24 ] who explain that “after repeated exposure to variation in environmental conditions,a person’s expectations of those conditions may become more relaxed— even anticipatory of temporal changes” (p. 181). This idea provides insight into why adaptive comfort control is a valid alternative to fixed setpoint systems, as it suggests that people’s expectations of environmental conditions can be adjusted and become more flexible over time. This adaptive capacity comes from repeated exposure to such environmental variations, which fosters more flexible expectations and even positive anticipation of temporary changes in the environment. This reinforces the idea that an adaptive control system, by allowing controlled variations in indoor conditions rather than maintaining a fixed setpoint, can not only be more energy efficient, but also promote a natural thermal comfort experience aligned with human expectations. As highlighted in [ 25 ] (p. 16), the application of adaptive comfort systems has been shown to improve energy savings, enhance resilience to climate change and increase thermal satisfaction and well-being. The adaptive comfort control approach could be seen as a solution to the phenomenon described by De Dear [ 26 ] in which occupants of buildings with traditional HVAC systems develop an expectation of thermal constancy due to the strict maintenance of fixed temperatures. This expectation, as the author points out, causes even small fluctuations in temperature to be perceived negatively, generating a hypersensitivity that leads to frequent complaints. Automated controls that adjust HVAC setpoints according to the Appl. Sci. 2025,15, 1243 3 of 17 outdoor climate and seasonal variations can be an effective solution. In fact, Deuble and De Dear [ 27 ] suggest that increasing indoor climate variability is key to sustainable building design strategies. In this context, it becomes relevant to analyze mixed mode (MM) buildings, which combine natural ventilation and HVAC systems. Although there is no consensus in the literature on control strategies for mixed mode buildings [ 28 ], this study proposes the analysis of a real case of integration of the adaptive approach in the operation of HVAC systems to investigate the implementation of a real mixed mode situation. Given that an MM building does not have to operate as a fully conditioned building the whole time, and although some authors consider or analyze that they must operate in the HVAC mode above a certain temperature [ 27 ], the reality is that an MM building can be a free-behaving building. The hypothesis analyzed in this study is that operating a significant percentage of the time in passive mode, with the HVAC system serving as a temporary backup adaptive measure, would be a feasible solution. This hypothesis is supported by recent findings [ 29 ], although only 20% of case studies on adaptive setpoints in Building Automation Systems (BAS) have been implemented in real-world cases (in European countries, this accounts for 36% of the 20%), and despite the fact that the operational strategies in mixed mode buildings are suitable to such approaches (48% are categorized as changeover strategies). Therefore, the aim of this paper is to expand upon these studies, addressing the gap in literature by analyzing real-world feasibility. Moreover, in recent studies such as [ 30 ], it has been shown that in situations when extreme outdoor temperatures entail the use of HVAC systems, adjusting adaptive setpoints can lead to a reduction of approximately 35% in cooling demand, thereby contributing to both energy savings and greater thermal comfort. The user might rely on this measure (HVAC usage) without being aware of the building’s setpoint (due to its adaptive nature), thus, expressing its need to improve the temperature on an ad hoc basis. Consequently, the setpoint, as it is adaptive, would follow the temperature proposed by the adaptive comfort algorithm, with the consequent energy savings that this would entail, compared to a continuous conditioning operation based on a fixed setpoint. In relation to the number of field studies carried out in MM buildings, although a growing trend of research focused on this area can be detected in the literature, the number is lower than in naturally ventilated or fully conditioned buildings [ 29 ]. In terms of climatology, building type and HVAC systems, a higher volume of work can be identified in MM buildings in oceanic and humid sub-tropical climates, commercial or business buildings and change-over and concurrent operations strategies, respectively. There is also a wide variety of control algorithms developed for MM buildings. Although some studies propose rule-based control, considering variables such as outdoor temperature, indoor temperature, thermal comfort or adaptive behaviors, among others, other research focuses on complex computational controls based on fuzzy logic or machine learning algorithms, highlighting the diversity of approaches in this field. Additionally, there is a need for this type of case study, since, as indicated by [ 29 ], 80% of the case studies are based on simulations, which points to the need to analyze a larger number of real implementations. This article is structured as follows: Section 2describes the material and methods, describing the case study as well as including previous knowledge, relevant background information and milestones that have enabled the development of the current study, as well as the general contextual conditions and climatology of the case study. This section also discusses the data collection process and the preliminary analyses conducted and presents the proposed methodology. Section 3shows the results obtained after one year of practical Appl. Sci. 2025,15, 1243 4 of 17 implementation of the integration of the adaptive algorithm into the control system and discusses the findings. Finally, Section 4presents the conclusions of the work. 2. Materials and Methods 2.1. Case Study The study was carried out in a building located in Seville (37 ◦ N, 5 ◦ W), in the southern region of Spain. The climate is classified as Csa (hot summer Mediterranean climate) according to the Köppen–Geiger classification [ 31 ], and as B4 according to the Spanish Building Technical Code [ 32 ]. The main characteristics of the climate in Seville are hot and dry summers, and mild, humid and temporary rainy winters. It is also characterized by one of the highest average annual temperatures in Europe, frequently exceeding 40 ◦ C. During the sample period, spanning from July 2023 to June 2024, these temperatures, as shown in (Table 1), exceeded this threshold. Table 1. Monthly variation in outdoor temperature (T out) and average relative humidity (RH). July 2023 August 2023 September 2023 October 2023 November 2023 December 2023 January 2024 February 2024 March 2024 April 2024 May 2024 June 2024 Tout min (◦C ) 18.9 18.9 13.4 11.1 3.9 1.6 2.2 4.4 5.0 7.2 7.5 14.4 Tout max (◦C ) 41.1 42.8 36.1 36.7 26.6 20.0 22.8 22.8 27.2 31.1 37.8 37.8 Tout avg (◦C ) 31.7 32.6 25.7 23.3 16.8 12.0 13.3 14.7 16.0 20.2 23.9 26.1 RHavg (%) 35 30 53 59 70 79 81 73 68 53 39 47 Note: Tout min ( ◦ C ), Tout max ( ◦ C ) and Tout avg ( ◦ C ) represent the minimum, maximum, and average outdoor temperatures, respectively. RHavg denotes the average relative humidity. This field study was carried out in an office room of a tertiary, non-residential building. Following the proposal of Morgan and De Dear [ 33 ], among the conditions for the adaptive approach, “It is essential that building occupants are free to adapt themselves,primarily through clothing adjustment,to the variable indoor climate regime prevailing inside such buildings” ( p. 267 ), so the present study allows users to make adaptive adjustments, although it focuses on the analysis of the inclusion of an adaptive algorithm into the building management system. The adaptive comfort algorithm applied in this area of the building was presented by Barbadilla et al. [ 34 ], in which a single adaptive model was proposed for both the NV and HVAC operation modes. The algorithm was based on an analysis of more than five thousand surveys, leading to the following equation (Equation (1)). Tc=0.2427 ×Trm +19.284 (1) where Tcis the indoor comfort temperature and Trm is the outdoor running mean temperature. Figure 1shows the adaptive range corresponding to the behavior of this algorithm over a full year. Regarding Figure 1, in [ 34 ] a comparative analysis of the adaptive control algorithm considered for the development of the present case study was carried out with respect to other models, showing that the neutral temperatures obtained were lower than those proposed using the ASHRAE-55:2023 and UNE-EN 16798-1:2020 models for NV buildings. Moreover, according to [ 9 ] (pp. 6–7), where a summary of different adaptive regression models and their coefficients was carried out, most of the linear adaptive linear regression models are within the limits recommended in ASHRAE 55. As for the implementation of the adaptive control algorithm, in [ 35 ] a methodology is presented for implementing an adaptive control algorithm in a Building Management System (BMS). In the present study, the BMS was adapted to allow users to switch between natural ventilation and heating, ventilation, and air conditioning modes. During the study Appl. Sci. 2025,15, 1243 5 of 17 period, users had the option of not using HVAC, with a key feature of this adaptation being that, in NV mode, occupants could freely open or close windows at their discretion. Appl. Sci. 2025, 15, 1243 6 of 18 Figure 1. Daily average outdoor temperature, daily operating temperature and adaptive comfort range based on [34]. This previous approach is based on the “change-over” mixed mode, commonly applied in office buildings, where natural ventilation and HVAC systems are used in the same space but at different moments [10,36]. In the present study, HVAC mode operation required all windows to remain closed, and windows could only be opened if the HVAC system had been turned off or inactive for at least 5 min [37]. To make use of the HVAC system, a limited one-off measure in building retrofit, it was proposed to include a 2-h inactivity timer in the BMS; if no interaction occurred during this period, the system was automatically switched off. However, the system could be stopped at any time. This design not only improves energy use but also gives the user personal control over the climate control system. This perceived control has an important psychological effect, as it allows people to feel that they can directly influence their environment, thus, improving their comfort and satisfaction [38]. The decision to allow the use of the HVAC system, with automatic shut-off functionality after 2 h of inactivity or when reaching the temperature of the adaptive comfort setpoint, aligns with the findings of [38]. According to this study, the perceived ability of users to control their thermal environment improves their perception of thermal comfort due to the psychological influence of perceived control. Furthermore, when faced with extreme thermal conditions, occupants have a greater desire to adjust their environment, and this personal control is even more effective. The system design not only provides perceived control that reduces thermal discomfort even with slight improvements in thermal conditions, but also ensures that the thermal demands of the occupants are adequately addressed. This can ultimately minimize complaints about the thermal environment, optimizing thermal comfort as well as energy efficiency. The study group consisted of 10 people aged 24 to 50 years, with 7 men and 3 women, in a room with 5 workstations (Figure 2), where the desks allowed users to exchange throughout the work day (“hot desk”). The office hours considered were from 8:00 to 20:00 h, and the participation of the subjects was voluntary, as is the case in many thermal comfort field studies. The study area depicted in Figure 2 shows a room with sensors and Figure 1. Daily average outdoor temperature, daily operating temperature and adaptive comfort range based on [34]. This previous approach is based on the “change-over” mixed mode, commonly applied in office buildings, where natural ventilation and HVAC systems are used in the same space but at different moments [ 10 , 36 ]. In the present study, HVAC mode operation required all windows to remain closed, and windows could only be opened if the HVAC system had been turned off or inactive for at least 5 min [37]. To make use of the HVAC system, a limited one-off measure in building retrofit, it was proposed to include a 2-h inactivity timer in the BMS; if no interaction occurred during this period, the system was automatically switched off. However, the system could be stopped at any time. This design not only improves energy use but also gives the user personal control over the climate control system. This perceived control has an important psychological effect, as it allows people to feel that they can directly influence their environment, thus, improving their comfort and satisfaction [ 38 ]. The decision to allow the use of the HVAC system, with automatic shut-off functionality after 2 h of inactivity or when reaching the temperature of the adaptive comfort setpoint, aligns with the findings of [ 38 ]. According to this study, the perceived ability of users to control their thermal environment improves their perception of thermal comfort due to the psychological influence of perceived control. Furthermore, when faced with extreme thermal conditions, occupants have a greater desire to adjust their environment, and this personal control is even more effective. The system design not only provides perceived control that reduces thermal discomfort even with slight improvements in thermal conditions, but also ensures that the thermal demands of the occupants are adequately addressed. This can ultimately minimize complaints about the thermal environment, optimizing thermal comfort as well as energy efficiency. The study group consisted of 10 people aged 24 to 50 years, with 7 men and 3 women, in a room with 5 workstations (Figure 2), where the desks allowed users to exchange Appl. Sci. 2025,15, 1243 6 of 17 throughout the work day (“hot desk”). The office hours considered were from 8:00 to 20:00 h, and the participation of the subjects was voluntary, as is the case in many thermal comfort field studies. The study area depicted in Figure 2shows a room with sensors and tables, with an area of 14 m2with windows with double glass and manual and indoor blinds. Appl. Sci. 2025, 15, 1243 7 of 18 tables, with an area of 14 m2 with windows with double glass and manual and indoor blinds. Figure 2. Floor plan of the field study room and sensors’ locations. 2.2. Acquisition, Processing and Preliminary Analysis of Data Throughout the study, objective measurements (indoor and outdoor climate) and subjective measurements (thermal preference) were collected simultaneously. The study was carried out over 12 months (July 2023 to June 2024) in order to characterize all the seasons. An array of wireless technology sensors recorded the behavior of the room. The newly acquired sensors were verified prior to the study. A total of 685,152 indoor environmental records (located in Figure 2) and 35,136 outdoor records (Table 2) were collected. Measurements inside the room (temperature, humidity, etc.) were obtained at 10 min intervals. All sensors were new, calibrated, compared with similar sensors under the same conditions, and showed no significant differences in their measurements. Sensors were placed approximately 1 m from the occupants’ workstations in order to describe their immediate thermal environment under typical working conditions. Outdoor weather observations were collected from a weather station close to the building [39] and located about 100 m from it, of which outdoor temperature was of the greatest interest. Table 2. Data acquisition systems and characteristics. Sensor Number of Sensors Operative Range Accuracy Units Concentration of particle matter 1 0 to 500 ±10% µg/m 3 CO 2 Concentration 1 0 to 10.000 ±45 + 3% ppm Relative humidity 5 0 to 100 ±2% % Air temperature 5 −25 to 70 ±0.3 °C Globe temperature 6 −25 to 70 ±0.3 °C Temperature in the weather station 1 −50 to 60 ±2% °C Relative humidity in the weather station 1 0 to 100 ±2% % Figure 2. Floor plan of the field study room and sensors’ locations. 2.2. Acquisition, Processing and Preliminary Analysis of Data Throughout the study, objective measurements (indoor and outdoor climate) and subjective measurements (thermal preference) were collected simultaneously. The study was carried out over 12 months (July 2023 to June 2024) in order to characterize all the seasons. An array of wireless technology sensors recorded the behavior of the room. The newly acquired sensors were verified prior to the study. A total of 685,152 indoor environmental records (located in Figure 2) and 35,136 outdoor records (Table 2) were collected. Table 2. Data acquisition systems and characteristics. Sensor Number of Sensors Operative Range Accuracy Units Concentration of particle matter 1 0 to 500 ±10% µg/m3 CO2Concentration 1 0 to 10.000 ±45 + 3% ppm Relative humidity 5 0 to 100 ±2% % Air temperature 5 −25 to 70 ±0.3 ◦C Globe temperature 6 −25 to 70 ±0.3 ◦C Temperature in the weather station 1 −50 to 60 ±2% ◦C Relative humidity in the weather station 1 0 to 100 ±2% % Measurements inside the room (temperature, humidity, etc.) were obtained at 10 min intervals. All sensors were new, calibrated, compared with similar sensors under the same conditions, and showed no significant differences in their measurements. Sensors were placed approximately 1 m from the occupants’ workstations in order to describe their immediate thermal environment under typical working conditions. Outdoor weather observations were collected from a weather station close to the building [ 39 ] and located about 100 m from it, of which outdoor temperature was of the greatest interest. The 2859 working hours during which the users were working in the office were selected, after eliminating weekends, holidays, and intensive working days in July (when Appl. Sci. 2025,15, 1243 7 of 17 the room was not used in the afternoon). A total of 1320 valid surveys were collected and were grouped according to the time of day, resulting in the evaluation of 483 declared thermal preferences. At the end of the study, the BMS register was obtained with the HVAC system states. The base setpoint value of the HVAC system applied the adaptive comfort algorithm for the building, in which the running mean temperature was calculated with an alpha of 0.8. The range associated with the adaptive comfort law is represented in Figure 1by two green dashed lines, which indicate the total width of the comfort zone for the sample period. The same figure shows the behavior of the indoor and outdoor temperatures for the building, together with the comfort zone based on the implemented adaptive law. The range is indicated in its most restrictive form, with a scale of +2/ − 3 ◦ C, building category I according to UNE-EN 16798-1 [ 4 ]. This more restrictive range was adopted to prioritize perceived thermal comfort in this analysis, considering that even small variations can generate significant responses. Although in the UNE-EN 16798-1 [ 4 ] indoor comfort range for the design and evaluation of the energy performance of buildings would be +3/ − 4 ◦ C, in Carlucci et al. [ 40 ] a comparative analysis of different adaptive thermal comfort models can be found, in which the most restrictive acceptability percentage corresponds to 90%. These percentages are justified on the grounds that current models cannot accurately predict how an individual will feel on a specific day, so it is not reasonable to expect all users to experience thermal comfort in the same environment, even when it complies with current standards. Based on the BMS register, the HVAC system was required as a punctual measure in 379 h (13.25% of the total active hours of office use). Of these, 206 h was sufficient to achieve thermal comfort within the planned operation time, while in 192 h, the planned operation time was insufficient, requiring an extension to maintain comfort. A total of 166 h of punctual activity occurred in summer, from June to September, and 113 h in winter, from December to March. In the study, the HVAC system remained off with passive adaptation measures for 2480 h (92.7% of the total hours of use). 2.3. Methodology One of the main challenges of thermal comfort studies is to keep people motivated to participate over time. This is especially relevant in approaches that require active interaction over long periods of time. The present proposal was, therefore, aimed at analyzing HVAC as an adaptive measure in MM and, given that the field study was carried out over one year, a central question on thermal preference was used: “How would you prefer to be now?” [ 41 ]. The scale issue about the thermal preference arose “to overcome the ambiguity of thermal condition acceptability considering the 3-point McIntyre preference scale” [ 42 ], but based on ISO 10551:2019 [ 41 ], “A 7-degree scale should be applied in the case of environments judged to be close to neutrality; a 9-degree scale should be applied in the case of environments judged to be more intense”, although studies can be found with an 11-point scale [ 42 ]. In [ 43 ] it is indicated that the scale for the thermal preference vote (TPV) represents how an occupant would prefer to adjust their thermal environment and it is, therefore, an accurate measure for defining an optimal indoor thermal environment. According to the authors, unlike other measures, TPV directly suggests a change in current conditions, making it an effective tool to improve the prediction of the HVAC system and optimize energy efficiency. In order to conduct this analysis, it is necessary to have a meaningful data set. Figure 3, relating to average TPV and operative temperature, depicts how the average thermal perception of participants varies with operational temperature. It shows that 78.8% of Appl. Sci. 2025,15, 1243 8 of 17 the variation in average TPV can be explained by operational temperature, so, given the statistical significance obtained, the study set can be analyzed. Appl. Sci. 2025, 15, 1243 9 of 18 variation in average TPV can be explained by operational temperature, so, given the statistical significance obtained, the study set can be analyzed. Figure 3. Relationship between Average TPV and Operative Temperature. After applying the adaptive comfort equation (Equation (1)) and the associated temperature ranges for the one-year period, the indoor thermal comfort is analyzed hour by hour (Figures 4 to 6), considering the thermal preference of the users. For this analysis, non-working hours, holidays and weekends are excluded. Subsequently, evaluation metrics are calculated to examine the accuracy of the system by defining the following indicators from [42], but adapted to the case of the analysis with a comfort law applied and on the basis of users’ preferences: Figure 4. Representation of indicators. True Comfort (TC): This value is identified when the comfort law applied correctly predicts thermal comfort, that is, when the thermal preference of the users is neutral, Figure 3. Relationship between Average TPV and Operative Temperature. After applying the adaptive comfort equation (Equation (1)) and the associated temperature ranges for the one-year period, the indoor thermal comfort is analyzed hour by hour (Figures 4–6), considering the thermal preference of the users. For this analysis, nonworking hours, holidays and weekends are excluded. Subsequently, evaluation metrics are calculated to examine the accuracy of the system by defining the following indicators from [ 42 ], but adapted to the case of the analysis with a comfort law applied and on the basis of users’ preferences: Appl. Sci. 2025, 15, 1243 9 of 18 variation in average TPV can be explained by operational temperature, so, given the statistical significance obtained, the study set can be analyzed. Figure 3. Relationship between Average TPV and Operative Temperature. After applying the adaptive comfort equation (Equation (1)) and the associated temperature ranges for the one-year period, the indoor thermal comfort is analyzed hour by hour (Figures 4 to 6), considering the thermal preference of the users. For this analysis, non-working hours, holidays and weekends are excluded. Subsequently, evaluation metrics are calculated to examine the accuracy of the system by defining the following indicators from [42], but adapted to the case of the analysis with a comfort law applied and on the basis of users’ preferences: Figure 4. Representation of indicators. True Comfort (TC): This value is identified when the comfort law applied correctly predicts thermal comfort, that is, when the thermal preference of the users is neutral, Figure 4. Representation of indicators. Appl. Sci. 2025,15, 1243 9 of 17 Appl. Sci. 2025, 15, 1243 12 of 18 Figure 5. Representation of hourly thermal preference during a year of study within working hours. When analyzing the results in Figure 6, where the terms TC, FC, TD and FD are associated with the metrics shown in Equations 2 to 6, it can be observed in Table 4 that a PD of 6.5% during the work year and an accuracy of 92.96% have been obtained. This indicates that the model correctly identifies 92.96% of cases in which thermal comfort is predicted within the comfort range. This high value suggests that the proposed temperature range for thermal comfort is quite representative of the real conditions perceived as comfortable by the occupants. Figure 6. Representation of comfort preferences within the defined adaptive comfort ranges. It is important to note that the proportion of TC, FC, TD and FD varies throughout the year due to seasonal factors. In particular, from July to September, a higher level of thermal discomfort (TD and FD) is observed. This is mainly due to temperature characteristics during these months, which can exceed the limits of the proposed thermal comfort range. Figure 5. Representation of hourly thermal preference during a year of study within working hours. Appl. Sci. 2025, 15, 1243 12 of 18 Figure 5. Representation of hourly thermal preference during a year of study within working hours. When analyzing the results in Figure 6, where the terms TC, FC, TD and FD are associated with the metrics shown in Equations 2 to 6, it can be observed in Table 4 that a PD of 6.5% during the work year and an accuracy of 92.96% have been obtained. This indicates that the model correctly identifies 92.96% of cases in which thermal comfort is predicted within the comfort range. This high value suggests that the proposed temperature range for thermal comfort is quite representative of the real conditions perceived as comfortable by the occupants. Figure 6. Representation of comfort preferences within the defined adaptive comfort ranges. It is important to note that the proportion of TC, FC, TD and FD varies throughout the year due to seasonal factors. In particular, from July to September, a higher level of thermal discomfort (TD and FD) is observed. This is mainly due to temperature characteristics during these months, which can exceed the limits of the proposed thermal comfort range. Figure 6. Representation of comfort preferences within the defined adaptive comfort ranges. True Comfort (TC): This value is identified when the comfort law applied correctly predicts thermal comfort, that is, when the thermal preference of the users is neutral, slightly cooler or slightly warmer (TPV ∈ [ − 1,1]). In other words, the system forecasts comfort and it corresponds to the occupant’s thermal sensation. False Comfort (FC): This value is identified when the comfort law incorrectly predicts thermal comfort, that is, the thermal preference of users is significantly outside the comfort range (TPV /∈ [ − 1,1]). In this case, the system does not accurately reflect the occupant’s thermal experience. True Discomfort (TD): This value is identified when discomfort is predicted and the thermal preference vote of the occupant is outside the comfort range (TPV /∈ [ − 1,1]). In this case, the system aligns with the thermal state experienced by the occupant. False Discomfort (FD): This value is identified when discomfort is incorrectly predicted, since the occupant’s thermal preference is within the comfort range (TPV ∈ [ − 1,1]). This Appl. Sci. 2025,15, 1243 16 of 17 20. Healy, S. Air-Conditioning and the ‘Homogenization’ of People and Built Environments. Build. Res. Inf. 2008,36, 312–322. [CrossRef] 21. Carlosena, L.; Ruiz-Pardo, Á.; Rodríguez-Jara, E.Á.; Santamouris, M. Worldwide Potential of Emissive Materials Based Radiative Cooling Technologies to Mitigate Urban Overheating. Build. Environ. 2023,243, 110694. [CrossRef] 22. Rincón-Casado, A.; Rodríguez Jara, E.Á.; Ruiz Pardo, A.; Salmerón Lissén, J.M.; Sánchez de la Flor, F.J. Reducing the Cooling Energy Demand by Optimizing the Airflow Distribution in a Ventilated Roof: Numerical Study for an Existing Residential Building and Applicability Map. Appl. Sci. 2024,14, 6596. [CrossRef] 23. Fishman, D.S.; Pimbert, S.L. The Thermal Environment in Offices. Energy Build. 1982,5, 109–116. [CrossRef] 24. Fountain, M.; Brager, G.; de Dear, R. Expectations of Indoor Climate Control. Energy Build. 1996,24, 179–182. [CrossRef] 25. Hellwig, R.T.; Teli, D.; Schweiker, M.; Mora, R.; Choi, J.-H.; Rawal, R.; Lee, M.C.J.; Wang, Z.; Al-Atrash, F. Guidelines for Low Energy Building Design Based on the Adaptive Thermal Comfort Concept—Technical Report: IEA EBC Annex 69: Strategy and Practice of Adaptive Thermal Comfort in Low Energy Buildings. Aalborg University: Aalborg, Denmark, 2022. 26. De Dear, R. Adaptive Comfort Applications in Australia and Impacts on Building Energy Consumption. In Proceedings of the Sixth International Conference on Indoor Air Quality, Ventilation and Energy Conservation in Buildings: Sustainable Built Environment, Sendai, Japan, 28–31 October 2007; pp. 28–31. 27. Deuble, M.P.; de Dear, R.J. Green Occupants for Green Buildings: The Missing Link? Build. Environ. 2012,56, 21–27. [CrossRef] 28. Aguilera, J.J.; Bogatu, D.-I.; Kazanci, O.B.; Angelopoulos, C.; Coakley, D.; Olesen, B.W. Comfort-Based Control for Mixed-Mode Buildings. Energy Build. 2021,252, 111465. [CrossRef] 29. Peng, Y.; Lei, Y.; Tekler, Z.D.; Antanuri, N.; Lau, S.-K.; Chong, A. Hybrid System Controls of Natural Ventilation and HVAC in Mixed-Mode Buildings: A Comprehensive Review. Energy Build. 2022,276, 112509. [CrossRef] 30. Lamberti, G.; Contrada, F.; Kindinis, A. Exploring Adaptive Strategies to Cope with Climate Change: The Case Study of Le Corbusier’s Modern Architecture Retrofitting. Energy Build. 2024,302, 113756. [CrossRef] 31. Kottek, M.; Grieser, J.; Beck, C.; Rudolf, B.; Rubel, F. World Map of the Köppen-Geiger Climate Classification Updated. Meteorol. Zeitschrift 2006,15, 259–263. [CrossRef] 32. Salmerón, J.M.; Álvarez, S.; Molina, J.L.; Ruiz, A.; Sánchez, F.J. Tightening the Energy Consumptions of Buildings Depending on Their Typology and on Climate Severity Indexes. Energy Build. 2013,58, 372–377. [CrossRef] 33. Morgan, C.; de Dear, R. Weather, Clothing and Thermal Adaptation to Indoor Climate. Clim. Res. 2003,24, 267–284. [CrossRef] 34. Barbadilla-Martín, E.; Salmerón Lissén, J.M.; Guadix Martín, J.; Aparicio-Ruiz, P.; Brotas, L. Field Study on Adaptive Thermal Comfort in Mixed Mode Office Buildings in Southwestern Area of Spain. Build. Environ. 2017,123, 163–175. [CrossRef] 35. Aparicio-Ruiz, P.; Barbadilla-Martín, E.; Salmerón-Lissén, J.M.; Guadix-Martín, J. Building Automation System with Adaptive Comfort in Mixed Mode Buildings. Sustain. Cities Soc. 2018,43, 77–85. [CrossRef] 36. Kim, J.; de Dear, R. Is Mixed-Mode Ventilation a Comfortable Low-Energy Solution? A Literature Review. Build. Environ. 2021, 205, 108215. [CrossRef] 37. Deuble, M.P.; de Dear, R.J. Mixed-Mode Buildings: A Double Standard in Occupants’ Comfort Expectations. Build. Environ. 2012, 54, 53–60. [CrossRef] 38. Luo, M.; Cao, B.; Ji, W.; Ouyang, Q.; Lin, B.; Zhu, Y. The Underlying Linkage between Personal Control and Thermal Comfort: Psychological or Physical Effects? Energy Build. 2016,111, 56–63. [CrossRef] 39. Moreno-Tejera, S.; Silva-Pérez, M.A.; Lillo-Bravo, I.; Ramírez-Santigosa, L. Solar Resource Assessment in Seville, Spain. Statistical Characterisation of Solar Radiation at Different Time Resolutions. Sol. Energy 2016,132, 430–441. [CrossRef] 40. Carlucci, S.; Bai, L.; de Dear, R.; Yang, L. Review of Adaptive Thermal Comfort Models in Built Environmental Regulatory Documents. Build. Environ. 2018,137, 73–89. [CrossRef] 41. ISO-10551; Ergonomics of the Physical Environment - Subjective Judgement Scales for Assessing Physical Environments. ISO: Geneva, Switzerland, 2019. 42. Jazizadeh, F.; Marin, F.M.; Becerik-Gerber, B. A Thermal Preference Scale for Personalized Comfort Profile Identification via Participatory Sensing. Build. Environ. 2013,68, 140–149. [CrossRef] 43. Kim, J.; Schiavon, S.; Brager, G. Personal Comfort Models – A New Paradigm in Thermal Comfort for Occupant-Centric Environmental Control. Build. Environ. 2018,132, 114–124. [CrossRef] 44. Barbadilla-Martín, E.; Ragel-Bonilla, J.C.; Aparicio-Ruiz, P.; Guadix, J. Analysing the Indoor Temperature Ranges for an Efficient Control of HVAC Systems. Energy Build. 2023,301, 113681. [CrossRef] 45. Zhang, S.; Yao, R.; Toftum, J.; Essah, E.; Li, B. Machine Learning-Based Approach to Predict Thermal Comfort in Mixed-Mode Buildings: Incorporating Adaptive Behaviors. J. Build. Eng. 2024,87, 108877. [CrossRef] 46. Nicol, J.F.; Raja, I.A.; Allaudin, A.; Jamy, G.N. Climatic Variations in Comfortable Temperatures: The Pakistan Projects. Energy Build. 1999,30, 261–279. [CrossRef] 47. Rijal, H.B.; Yoshida, H.; Umemiya, N. Seasonal and Regional Differences in Neutral Temperatures in Nepalese Traditional Vernacular Houses. Build. Environ. 2010,45, 2743–2753. [CrossRef] Appl. Sci. 2025,15, 1243 17 of 17 48. Brager, G.; Paliaga, G.; De Dear, R. Operable Windows, Personal Control and Occupant Comfort. ASHRAE Trans. 2004,110. 49. Humphreys, M.A.; Rijal, H.B.; Nicol, J.F. Updating the Adaptive Relation between Climate and Comfort Indoors; New Insights and an Extended Database. Build. Environ. 2013,63, 40–55. [CrossRef] 50. Li, D.; Menassa, C.C.; Kamat, V.R. Personalized Human Comfort in Indoor Building Environments under Diverse Conditioning Modes. Build. Environ. 2017,126, 304–317. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.