1. Introduction
The increase in outdoor temperatures and overheating due to global warming is starting to impact retrofitting strategies for existing buildings, especially educational buildings in the Mediterranean climate. This problem takes on great importance in classrooms because of the significant influence of indoor conditions derived from high internal gains from high occupancy. Retrofitting this building stock has mainly focused on energy improvements, with less consideration for indoor thermal and lighting comfort conditions [
1]. Nevertheless, the issue of indoor thermal comfort assessment in educational buildings is highly relevant in the scientific literature [
2,
3,
4], as discomfort conditions can even affect students’ activity and attention [
5], especially in the case of children, considered to be more sensitive targets [
6]. In addition, the effects of indoor environment quality (IEQ) in classrooms on comfort, health and learning performance can be clearly observed [
2,
7,
8,
9,
10], depending on different conditions such as acoustical, thermal, indoor air and visual quality.
Several studies reveal poor indoor air quality (IAQ) conditions in classrooms [
11,
12,
13,
14,
15] and evidence of links between poor environment quality and students’ reduced performance, learning ability, and attendance [
16,
17]. Shendell et al. [
18] even found a statistically significant association between CO
2 levels and absenteeism in schools. However, there is no scientific agreement on the relative importance of IEQ factors for overall satisfaction, as pointed out by Kim and de Dear [
19] using Kano’s satisfaction model. Through surveys conducted in British schools, Barret et al. [
20] conclude that the ambient factor, which most conditioned the design of a classroom, is natural lighting. Nonetheless, these authors also state that thermal comfort is one of the aspects that most influence IEQ and, along with air quality, is one of the key factors associated with energy efficiency and, in turn, energy and operational costs.
In the specific case of Southern Spain, most educational buildings do not usually incorporate cooling systems; as a result, comfort conditions are rather limited, with classes occasionally canceled due to high indoor temperatures in classrooms. Thus, thermal comfort analysis is a key aspect for the evaluation of the IEQ in these buildings.
At present, the suitable indoor temperature is commonly defined according to the thermal comfort theory formulated by Fanger [
21], based on heat balance model studies (1970), which led to the definition of a universally recognized comfort index: the predicted mean vote (PMV). The key concept of this model is that, when exposed to constant, moderate thermal conditions and subjected to continuous physical activity, the human body tends to produce a quantity of heat that balances dissipated heat without noticeable accumulation. Therefore, the internal production of heat, minus natural losses from evaporation and transpiration (regarded as “internal”), is equal to radiative and convective thermal dissipation.
Consequently, the specific quantities taken into account to describe all the components of the heat balance equation, chosen after a sequence of empirical tests, were two values connected to the occupant (i.e., activity level, which is the metabolic rate defined as “met,” 1 met refering to 58.2 W/m
2 [
22], and the thermal resistance of clothing, which measures the thermal insulation of clothes and is expressed as “clo,” where 1 clo corresponds to 0.155 m
2·°C/W [
22]) and four values connected to the environment (i.e., air temperature, mean radiant temperature, mean air velocity, and vapor pressure).
Assuming the equivalence between comfort and sensation of thermal neutrality, studies were undertaken to build an index of thermal sensation, a quantity that could predict the occupants’ thermal response, universally valid and based upon the previously mentioned variables. In order to standardize the results, Fanger [
21] used the widely deployed 7-point ASHRAE 55:2004 [
23] scale: from −3, very cold, to +3, very warm, with the average value of 0 as the perfect thermal neutrality.
Fanger [
21] also calculated the percentage of dissatisfied persons that corresponded statistically to a given PMV, which is important in normative terms, as it leads directly to the development of regulations related to the issues of comfort. The relation between PMV and dissatisfaction levels has been developed through the statistical evaluation of thermal votes obtained via an extensive sequence of climatic chamber tests. Considering all the votes outside the interval between −1 and +1 as unsatisfying, the PPD function (predicted percentage of dissatisfied) shows the correlation between the PMV and the percentage of dissatisfied people.
The main international standards in place regarding the determination of indoor environmental conditions—ISO 7730:2005 [
22], CEN EN 15251:2007 [
24] and its revision prEN 16798-1:2015 [
25], and ASHRAE 55:2017 [
26]—implement this approach. The Fanger [
21] equation is used to determine reference temperature ranges according to conventional seasonal clothing level and standard air humidity and velocity and relating to building use, corresponding to defined activity levels.
Fanger’s work [
21] is of utmost importance as it resulted in the first universally valid technique for the evaluation of thermal environments. However, its formulation was the object of some criticism, since this approach bases the definition of thermal comfort on mere physics, completely neglecting the social and psychological aspects of thermal perception. Moreover, as its formulation is completely steady-state, it determines a very narrow range of allowable temperatures throughout the year regardless of outdoor conditions.
Because of the above observations, field studies on environmental conditions and comfort determination inside actual buildings first began in the 1970s, aiming to verify the accuracy of the PMV/PPD indices in predicting the actual thermal sensation and preference expressed by the occupants. Given the important differences highlighted by the results [
27], the adaptive approach was developed as an alternative approach to the definition of comfortable environmental conditions. The main assumptions of this approach concern:
—the ability of human beings to adapt to environmental conditions (through conscious or unconscious changes in metabolic rate or clothing level) and to interact with the environment in order to adapt it to their needs (through available environmental controls);
—the influence of thermal experience on the occupants’ expectations regarding indoor conditions, which can be short-term, as a result of recent weather, or long-term, related to the general climate they are used to.
The extensive data collected in these studies [
28,
29,
30] allowed for a statistical analysis that revealed a direct correlation between the indoor comfort temperature and the outdoor one, leading to the formulation of Equation (1) as the base of the adaptive approach.
where
Tco = the indoor comfort operative temperature (°C);
Text,ref = the outdoor reference temperature (°C);
a = the slope of the function;
b = the y-intercept of the function (°C).
In recent decades, several specific formulations of the adaptive equation have been developed [
31] according to the results of different studies and the purpose of the equations. Formulations differ based on equation formulation (a and b values); outdoor reference temperature determination; acceptability range, indicating the allowable gap from the “ideal” comfort temperature; and equation applicability, which usually depends on the season and on the availability of air conditioning or cooling devices.
Even though adaptive comfort formulations are also implemented in the current international standards EN 15251:2007 and ASHRAE 55:2017, both adaptive models can only be applied with a given outdoor temperature range. In the case of ASHRAE 55, the outdoor reference temperature must be above 10 °C and below 33.5 °C. The equation of the EN 15251 model can be used for a specific outdoor reference temperature for both the upper and lower limits. For the upper limit, the temperature must be above 10 °C and below 30 °C, while the lower limit must be between 15 and 30 °C. For these reasons, the use of these adaptive models in hot climates presents various complications.
Notwithstanding the above, Becker [
32] found that people in warm climates prefer what they term a ‘‘slightly cool’’ environment, in contrast to the ‘‘slightly warm’’ preferences of people in cold climates. In South Portuguese secondary classrooms, Pereira et al. [
33] found that students are prepared to accept indoor temperatures higher than 25.2 °C, as well as high CO
2 levels, exceeding the limits set by the standards for both cases. Furthermore, even though some research considers adaptation actions by students to be limited (opening or closing windows, moving sun shading devices, adding or removing layers of clothing, etc. [
34]), most of these confirm the relevance of the impact of user patterns on the energy performance in buildings [
35,
36] and the relevance of the user capacity of adaptation to indoor environments.
In Southern Spain, climate change aggravates heatwave scenarios, especially in the summer season. As recorded from 1 April to 25 June 2017 in Seville, outdoor temperatures were above 30 °C for 67% of the days, exceeding 40 °C (considered the minimum value for heatwave temperatures; with at least three consecutive days in which 10% of the weather stations recorded maximum temperatures over the 95% percentile of the daily maximum temperatures in July and August of 1971–2000) [
37] for 16% of the days (
Figure 1). For this reason, the thermal comfort analysis based on the Fanger [
21] method established in ISO 7730:2005 was used in this study. Moreover, this method best reproduces the real metabolic rate and thermal resistance values in the case study, especially in winter and mid-season.
The main aim of this paper is to assess the impact of an egg-crate shading device, that is, one made of combined vertical and horizontal shading elements, on the IEQ for the retrofitting of an educational building, considering real user patterns. This type of device, among several other existing types [
38,
39], is particularly effective in regions with a hot climate as it produces highly efficient shading while allowing people to see outside [
40].
The results presented in this paper complement the conclusions of previous research as part of the same case study conducted by the authors [
41], where the influence of a low-cost egg-crate shading device made of methacrylate on the annual electricity consumption (kWh) of the classrooms was analyzed for three seasonal periods (winter, summer, and mid-season). Among the conclusions of the paper referenced, it should be highlighted that the use of these devices allows for significant savings in annual electricity consumption of the air-conditioning system (20%) and artificial lighting (50%), when compared to a classroom with no egg-crate shading devices.
2. Methodology
The methodology used in this paper is based on the in-situ monitoring of two classrooms in a school building in Seville (Southern Spain). An external egg-crate shading device was installed in one of the classrooms while the other was left in its original constructive configuration. The ambient and lighting variables recorded are air temperature, surface temperature of walls, aperture of windows, indoor and outdoor illuminance, rolling shutter aperture levels, and incident solar radiation on the window apertures. Monitoring was carried out during three distinct seasonal periods (winter, summer, and mid-season) under real user patterns, allowing for the simultaneous comparison of the performance of both classrooms under the same outdoor conditions.
A descriptive statistical analysis was conducted using the integrated development environment (IDE) of Matlab [
42]. Unlike the earlier research [
41], this paper focuses on the analysis of the influence of the egg-crate shading device on different ambient variables related to indoor comfort: incident solar radiation (W/m
2), indoor operative temperature (°C) and indoor natural illuminance (lux). To assess the real impact of the device, different occupied and unoccupied periods were selected. In the unoccupied period, rolling shutters were totally open and windows fully closed. Thermal comfort in the occupied classrooms (% comfort hours and operative temperatures in °C) and indoor natural illuminance (lux) were analyzed. The aim was to assess the real influence of user patterns on two variables freely controlled by users: window aperture and rolling shutter aperture levels.
In order to facilitate comprehension, this research analysis (
Section 4) is divided into several subsections. The first presents the impact of the egg-crate shading device on the solar radiation incident on the windows during the winter, summer, and mid-season periods selected. The second corresponds to the analysis carried out during the unoccupied period, where the real influence of the solar radiation on the indoor operative temperatures and indoor natural illuminance levels (rolling shutters totally open) was studied. The final subsection analyzes the egg-crate shading device from the IEQ perspective, considering occupied periods and taking the influence of users into account. Thermal comfort and natural illuminance results are reported.
2.1. Considerations for Thermal Environment Analysis
In order to determine the comfort band of the ISO 7730:2005 method, the parameters considered were an average relative humidity of 50%, an interior predicted percentage of dissatisfied (PPD) below 15%, a metabolic rate of 1.2 met, and a thermal resistance of 0.5 clo in summer and of 1.2 clo in winter and mid-season. Hence, the thermal comfort bands established were 16.1–24.1 °C for winter and mid-season and 22.8–26.8 °C for summer.
2.2. Considerations for the Daylight Environment Analysis
To guarantee adequate indoor lighting conditions, standard EN 12464-1:2012 [
43] was used for the natural daylight environment analysis. This standard establishes minimum average illuminance values for the working surface to design the artificial lighting system, depending on the type of building. However, in this paper this standard is used to establish reference values for qualifying indoor lighting conditions in both classrooms. EN 12464-1:2012 recommends a minimum average illuminance level on the work surface in a classroom of between 300 (secondary school) and 500 lux (adult education). In addition, the Spanish safety code in place [
44] establishes a minimum of 100 lux to facilitate circulation and the evacuation of indoor spaces.
3. Case Study
3.1. Case Study Description
In this research, a public educational building managed by the Andalusian government, Martínez Montañés Secondary School, was monitored. This building is located in Southern Spain, in a Mediterranean climate city (latitude 37.37° N and longitude −5.97° W). Two south-facing 6.95 × 6.85 m classrooms on the second floor of the building (
Figure 2) were monitored.
Both classrooms preserved their original external shading protection, consisting of rolling shutters made of PVC plastic. Additionally, an external egg-crate shading device was installed in Classroom A (
Figure 3). Classroom B maintained the conditions from the original construction, for use as reference in the analysis.
Each classroom has four window modules 2.40 m high by 1.40 m wide, facing south, with aluminum frames and a 4 mm single glazing (U = 5.70 W/m2·K, no thermal bridge). The window surface represents around 53% of the facade of each classroom. Each classroom has a capacity of 35 students and lessons take place both in the morning from 9:00 a.m. to 3:00 p.m. (secondary education) and in the evening from 5:00 p.m. to 8:00 p.m. (adult education).
In both classrooms, the air-conditioning consists of a direct expansion heat pump and two wall indoor units with an exterior inverter unit. Moreover, both classrooms include an artificial lighting system, with four luminaires with two 36 W fluorescent lamps each.
3.2. Shading Device Description
The solar protection system design is made of tongue-and-groove joint 3-mm-thick white methacrylate slats. The lattices are 1200 mm high, 800 mm wide, and 30 mm thick, forming small 43 × 43 mm lattices.
The egg-crate shading device was installed outside the window, 0.20 m away from the external facade line and 0.35 m from the glazing surface. An external auxiliary steel structure was assembled to create a 1640 × 2440 mm frame for the device.
This shading device was manufactured using a laser cutting machine in the Digital Fabrication Laboratory of the Higher Technical School of Architecture of the University of Seville (FabLab).
3.3. The Monitoring System
The monitoring system, installed according to EN ISO 7726:2011 [
45], consisted of nine data loggers recording information at 10 min intervals. This information was stored and uploaded by mobile card every 30 min to a file transfer protocol (FTP) server.
Table 1 shows the technical characteristics of the sensors used for monitoring the classrooms.
There were two monitoring panels in each classroom, with two data loggers each (four data loggers per classroom). The first stored information related to the electric consumption of the artificial lighting system. The second data logger was connected to a presence detector (indoors), four sensors to measure the rolling shutter aperture levels, four sensors to record the aperture of the windows, and four lux meters (three indoors and one outdoors). It is important to stress that, as the presence detector only registered information about classroom occupancy (occupied or yes vs. unoccupied or no), there was no information on the real number of students occupying the classroom at any time. Indoor surface temperature sensors were incorporated into the third data logger, while the fourth integrated four air temperature sensors, all inside the classroom. This data logger also recorded information from the pyrgeometer (outdoors) and pyranometer (outdoors). The pyranometer was placed vertically, between the egg-crate shading device and the rolling shutters of the windows (Classroom A). In contrast, for Classroom B, the pyranometer was placed vertically on the unobstructed facade in order to measure the incident solar radiation.
There was also a third monitoring panel outside the classrooms, with the ninth data logger. This last data logger controlled the electricity consumption of the air-conditioning system (AC), recording information on power consumption (W), voltage (V), and electric current (A).
A local weather station, located in some test cells at a distance of 450 m [
46], was also used to record outdoor air temperature. The characteristics of the sensors used in the weather station were units in °C, a range between −40 and +80, and an accuracy of ±0.15 ±0.1%. The thermometer was placed facing north to avoid direct solar radiation.
3.4. Monitoring Phases
Both classrooms were monitored for a whole year: from May 2017 to April 2018. This paper presents the results of the analysis of solar radiation levels (W/m
2), indoor operative temperatures (°C), and their relation with thermal comfort and indoor natural illuminance levels (lux). Data were recorded in the classrooms during three representative seasonal periods: winter, summer, and mid-season. The periods considered for the protocol selection were those with 10% or less data lost. Moreover, the selection of the summer period also took into consideration the fact that the academic year ends in mid-June. The characteristics of the periods analyzed are shown in
Table 2.
Additionally, an unoccupied period of 120 h (18–22 April) was selected for analysis to obtain objective results of the real performance of the egg-crate shading device without considering user influence. Specifically, the influence of solar radiation on the indoor operative temperatures (°C) and indoor natural illuminance (lux) was assessed, considering the rolling shutters to be fully open.