**3. Results**

The results of the present study, part of a PhD dissertation [69], can be grouped into five subsections:


These values were analyzed according to seasons (winter, W, and midseasons, MS) and windows' and doors' operation (open windows, OW, closed windows, CW, open door, OD). In this way, 26 classrooms (55% of the case studies) had the windows closed during the measurement period, with 23 of these during the winter session, and 21 had the windows open (45% of the case studies), with 11 of them during the winter period. No intervention by the researchers was made to modify classroom-state, allowing us to gather operational actual conditions.

#### *3.1. Mean Values of Physical Parameters*

The measured interior air temperature (Ta) ranged between 17.8 and 22.7 ◦C during winter season (Table 5), with the lowest mean temperature values obtained for the case studies with closed windows, especially when inner doors were open, with values of 20 ◦C. It can be related with the outdoor conditions, given that the lowest outdoor temperature values (Ta, outdoor) were measured for classrooms with closed windows and open inner doors. In addition, 8 of the case studies had the windows open during winter, which can mean that there was a bad regulation of the heating system and the heat excess had to be dissipated, or that the students considered that they had to ventilate the classroom due to a poor environment perception. Indoor air temperature in midseasons was oscillating around 22.4 ◦C, without a direct relation with window operation. Although winter time temperature expectations range between near 20–22 ◦C to 20.4–22.6 ◦C if windows are open (central quartile lower and upper values), this band nearly doubles in middle season, when temperatures from 20.6 to 24 ◦C may be expected (21.1 to 24.5 ◦C if windows are open). A quartile distribution plot for indoor thermal parameters, air temperature, and operative temperature is proposed in Figure 4. It is noteworthy to highlight that there was a statistical significance between seasons in a windows-state with independent behavior aspect that was verified through test of comparison of samples, F-test for the variance and a K-S (Kolmogórov-Smirnov) for the distributions of probability with *p*-values under 0.05 in all the cases.


**Table 5.** Mean values of environmental parameters obtained during the field measurements related to seasons and windows' and doors' operation.

W are measurements during winter, MS are measurements during midseasons, OW-W are measurements during winter with open windows, OW-MS are measurements during midseasons with open windows, CW-W are measurements during winter with closed windows, CW-MS are measurements during midseasons with closed windows, CW OD-W are measurements during winter with closed windows and open doors, SD are standard deviation.

Although average values of mean radiant temperature (t*r*) were within the recommended operating temperature ranges for classrooms according to ISO 7730 standard [23] (22.0 ± 2.0 ◦C for category B), there was a high dispersion of figures with a standard deviation (SD) between 1.7 ◦C in winter with closed windows and 3.7 ◦C in midseasons with open windows, which was due to the operation of HW radiator system, especially when windows were open, with t*<sup>r</sup>* values of 27.2–28.0 ◦C with radiators on and values of 17.5–19.0 ◦C when radiators were turned off. This caused operative temperature to swing usually between 20 and 25 (central quartiles) during middle season, with typical values of 20.6 to 22.5 ◦C during winter, with a very similar band of 20.4 to 22.6 ◦C if windows were open, highlighting the effect of surface thermal control performed by the radiator heating system.

Relative humidity (RH) in winter was always over 40%, with a maximum value of 64% in the case of one of the classrooms with windows closed and inner doors opened. In midseasons, relative humidity was lower but with a higher oscillation, with a minimum value of 29%.

Air velocity (Va) values were oscillating under 0.05 m/s, both in winter and midseasons, only exceeding the recommended design limit for comfort category B established by the ISO 7730 standard [23] of 0.16 m/s in one of the case studies with open windows, with a value of 0.18 m/s. In the case of closed windows, air velocity was always under 0.09 m/s. This showed poor air movement and limited air displacement potential.

Measurements of the CO2 concentration usually show figures well above typical thresholds (Figure 5). The World Health Organization (WHO) recommends a limit for healthy indoor spaces of 1000 ppm [70]. In this way, the probability distribution derived from the measures showed that more than 92% of the distribution for closed windows was above this limit, while this only decreased to 88% of the time when windows were open. In addition, 47.5% of classrooms with windows closed exceeded the 2000 ppm threshold. The greatest relative effect of window operation was seen in the winter, when CO2 concentration can be decreased by 25%, comparing median values. However, figures were above desirable levels, indicating the lack of capacity of the window operation to solve a suitable ventilation. In general, during the intermediate season, the operation of the windows did not provide a significant improvement of indoor air quality, which may be related to the lack of thermal differential between indoor and outdoor air, limiting the air exchange due to the absence of a thermodynamic effect (Figure 6).

**Figure 4.** Quartile distribution for indoor air temperature and operative temperature. (**a**) Quartile distribution for indoor air temperature; (To) winter (top) and mid-season (down) with windows closed (0) and open (1). (**b**) Quartile distribution for operative temperature (To); winter (top) and mid-season (down) with windows closed (0) and open (1).

**Figure 5.** CO2 concentration distribution. (**a**) General CO2 concentration density trace for winter (blue) and mid-season (red). (**b**) Detailed CO2 concentration density trace for winter (blue) and mid-season (red) with closed windows (continuous line) and windows open (dot line).

The median room mean illuminance (E) in the case studies oscillated between 461 and 560 lx (both cases with a SD of 222) according to the season, with an average lighting uniformity (Uo) of 0.48. However, although there seemed to be a greater illumination associated with the half-season period, it was not possible to rule out, without further measurements, the fact of being biased by the activities in execution during the measurements. The high SD was due both to the use of the projector and the

solar protection devices (as low as 15 lx) and the lack of use of a solar protection device with direct solar radiation (figures as high as 1710 lx) (Figure 7). The homogeneity of the lighting solutions in almost all buildings generated visual fields with very similar characteristics, mainly dominated by the behavior of their electric lighting. The correlated color temperature was similar in all cases, varying from 3500 to 5500 K; hence, it can be considered that both the amount of light and hue did not affect the thermal perception of the participants, as exposed by Bellia et al. [71] and Acosta et al. [72].

**Figure 6.** Indoor and outdoor air temperature values (Ta), mean radiant temperature values (t*r*), relative humidity (HR) values and indoor and outdoor CO2 concentration values related to seasons and windows' and doors' operation.

**Figure 7.** Room illuminance. (**a**) Hourly distribution for mean-room illuminance by season: winter (blue) and middle-season (red). (**b**) Quartile distribution of mean-room-illuminance by season (top) winter (down) mid-season.

#### *3.2. Mean Values of Airtightness of the Samples*

The values of airtightness of the classrooms under study with a difference of pressure between indoor and outdoor of 50 Pa (n50 range) varied from 2.6 h–1 to 10 h–1, with an average value of n50 of 6.97 h–1 and a SD of 2.06 h–1.
