In 2007, the Federal Government of Brazil instituted the Proinfância Program as a national restructuring plan for public preschools for early childhood education to mend the deficit of vacancies for children in daycare centers and preschools. The Program aims to improve the Brazilian preschool infrastructure by constructing preschool buildings and acquiring equipment and furniture to guarantee children's access to public preschools. The Fundo Nacional de Desenvolvimento da Educação - FNDE (National Fund for Education Development) developed SPDs to facilitate the financial assistance process [5].

The Program's SPDs were primarily developed in 2007 as Types B and C, being replaced by Types 1 and 2 in 2014. Type 1 is larger and admits 376 children, while Type 2 admits 188 children. More than 5 billion Brazilian reais were invested from its creation until 2017, while more than 5,600 preschools of all types had already been built or were being built [6].

The thermal performance of buildings and users’ thermal comfort are different concepts. On the one hand, thermal performance only considers the building's overall thermal behavior and does not take into account the human thermal sensation. On the other hand, Standard 55 [7] defines thermal comfort as the mental condition that expresses satisfaction with the thermal environment and is assessed through subjective evaluation.

Some researchers developed studies about the Proinfância Program's SPDs' thermal comfort evaluation, pointing out the importance of reassessing these designs. Researches concerning indoor thermal comfort are important to developing standards aiming at performance and sustainable practices in buildings [8] since the development of thermal comfort regulations might be a challenge for politicians in some countries [9].

Peglow et al. [10] found low thermal comfort levels in long-standing rooms for the former Type C. These results showed a higher percentage of thermal discomfort by heat than cold in the Climatic Zone (CZ) 2.

Babick and Torres [11] evaluated the thermal comfort, lighting, and ventilation of the former Type C. They affirm the importance of choices' flexibility concerning the building envelope construction materials, shading elements, window frames, and glazing, besides developing a different standardized design for each climatic region.

Modler et al. [12] assessed the thermal comfort in the former Type B through a post-occupancy evaluation, finding discomfort from cold and highlighting the need to review some aspects of the standardized design.

Nascimento and Batista [13] investigated former Type B’s thermal comfort at CZ 8. Results showed that the indoor spaces with less propensity for natural ventilation, according to a qualitative experiment (water table), presented a higher percentage of thermal discomfort in the computer simulation. The authors state that the building's thermal performance might be affected by the building orientation and the heat exchange through the ground floor. They also pointed out the importance of adapting these aspects to the context of each location.

Rheingantz et al. [14] claimed that instead of prioritizing children, pedagogical practices, and relationships inside the school environment, the Program prioritizes public policies, responding to Proinfância's goals and objectives, with standardized designs that disregard regional differences and cultural values of communities.

In 2017, the FNDE made available a new version of the Manual of technical guidelines for elaborating school building designs, which included guidelines for the thermal performance evaluation of the Proinfância Program’s SPDs [15]. According to the evaluation criteria, the manual requires certain thermal performance levels by the NBR 15.575-1 thermal performance simulation method [16]. Although the Proinfância Program Manual indicates the NBR 15.575 simulation method to assess the preschool design's thermal performance, it is officially used for the thermal performance assessment of residential buildings. Among the reviewed authors, only Babick and Torres [11] used the NBR 15.575:2013 to compare the requirements of thermal transmittance and thermal capacity for walls and roofs of the preschool through the simplified method. However, none of the studies applied the NBR simulation method to evaluate the preschool's thermal performance. Besides, Proinfância Program’s Manual does not correlate thermal performance levels with suitable thermal comfort temperatures for school activities.

The Brazilian territory is divided into eight climatic zones [17] with specific and distinct climatic characteristics, thus requiring different construction strategies to obtain environmental comfort in each one [17]. It is essential to be aware of the climatic variables for proper building implantation since it is possible to take advantage of these variables to obtain a comfortable indoor environment for the occupants [18]. Proinfância SPDs might be built wherever any county is interested in implementing the preschool, as long as it complies with some prerequisites established by the Federal Government. However, a federal report [19] indicated inadequacies resulting from standardized designs. It pointed out that the SPDs did not contemplate the regional diversities, especially concerning the climate's physical and environmental conditions [19].

It is important to pay attention to the building orientation because it significantly affects the amount of received heat [20]. It is a valuable design decision for owners, builders, and designers as it plays a vital role in increasing the thermal comfort period and reducing annual energy consumption [21]. Besides, it is a decision that might be easily incorporated into the design phase [22]. Mascaró [20] points out that an inadequate building orientation in the main facades with a large thermal load might consume an average of 30% more operational energy per square meter than a well-oriented building. Hashemi and Khatami [23], assessing solar heat gain in homes in Uganda, showed that the east and west facades have an average solar heat gain 1.4 times greater than the north and south facades. It is recommended to position the building's main facades and openings in a north-south orientation for buildings located in low latitudes and tropical climates [23]. Short et al. [24] developed a study proposing viable low-carbon adaptation strategies and found that building orientation is an important aspect of influencing indoor thermal temperature and energy consumption and should be considered to improve thermal comfort levels, reduce energy consumption and also reduce CO2 emissions. Katafygiotou and Serghides [25] pointed out that schools' construction is a complex task since it is necessary to pay attention to heating, cooling, and ventilation costs while aiming to comply with adequate thermal comfort levels, as it means healthy spaces for educational activities. Thus, the choice of adequate building orientation might be a passive strategy to increase thermal comfort levels and reduce energy consumption in a building. Rheingantz [14] used qualitative and quantitative approaches to evaluate a former SPD and confirmed the importance of taking advantage of an adequate building orientation to arrange the openings. The design's configuration and implementation should control solar radiation while taking advantage of the wind flow patterns.

The most convenient building orientation might not be the same in all climatic zones. It is essential to foment the development of researches concerning guidelines for the design implantation and construction of standardized building designs. Building orientation directly affects the thermal comfort and thermal performance of buildings. When a municipal administration chooses one of the SPDs that fit their local demand, it is left to them the decision about the building orientation in the available site. However, there are no available studies about building orientation's impact on the SPDs' thermal comfort and thermal performance. Consequently, there is no official information regarding the SPDs' best building orientation in each Brazilian climatic zone.

In this way, the present study develops a method to define an adequate building orientation for Proinfância Program’s SPDs to be implemented in Brazilian climatic zones. It is based on thermal performance evaluation through the NBR 15.575 simulation method, mandatory by the Proinfância Manual, and on thermal comfort evaluation through the ASHRAE Standard 55 Adaptive Model. A decision-making framework combines these evaluations to generate specific building orientation’s recommendations for each Brazilian climatic zone.

The parameters used in the preschool building simulations for thermal comfort and thermal performance are detailed below.

2.2.3. Heat Exchange through Ground Floor

The NBR 15.575-1 [16] does not comment on heat exchange through the ground floor. Thus, this parameter was not considered in the Thermal Performance assessment, and the floor was considered adiabatic. On the other hand, thermal exchanges between the building and the ground were considered for thermal comfort assessments through the EnergyPlus Slab Tool's aid, which calculates the dynamic ground temperature throughout the year.

Fig. (5). Identification of the long-standing rooms in Type 2’s floor plan.

Table 4. Number of occupants and thermal loads for long-standing rooms.

Long-Standing Room	Nº of Children	Nº of Teachers/ Staff	Lighting (W)	Electric Equipment (W)
Administration	-	2	128	-
Nursery 1	10	1	256	150
Nursery 2	16	1	256	150
Nursery 3	20	1	256	150
Kindergarten 1	24	1	384	150
Kindergarten 2	24	1	384	150

A decision-support framework was developed to support the building orientation definition since there will be indications from Thermal performance and Thermal comfort assessments (Fig. 6). Some criteria were adopted to generate the recommendations, as follows:

(a) The preschool must reach the thermal performance level required by the Proinfância Manual for winter and summer in one building orientation, at least. Otherwise, SPD should be reviewed.

(b) When there is a unique coincident building orientation indication from both assessments, it confirms this building orientation indication.

(c) When thermal performance indication is different from thermal comfort indication, the thermal comfort result is prioritized due to its relationship with the occupants’ thermal sensation.

(d) When thermal performance and thermal comfort results indicate more than one building orientation, there are two possibilities:

1. If there is one or more than one coincident building orientation indication from both assessments, municipal managers choose any coincident building orientation indication according to local circumstances;

2. If there is no coincident building orientation indication from both assessments, municipal managers choose any thermal comfort indication that fits local circumstances.

Type 2 reached Intermediate and Superior thermal performance levels in all CZs and building orientations for the summer period, accomplishing the Proinfância Manual requirements, except for West in CZ 5. The building orientation variation generates different classifications in CZ 4, 5, and 8 only (Table 9). Again, it is noteworthy that some long-standing rooms’ temperatures as high as 33ºC, which achieved Superior thermal performance level, might not signify thermal comfort for the occupants.

Fig. (6). Decision-support framework for building orientation’s definition.

On the one hand, Type 2 did not achieve the Minimum thermal performance level for the winter period in CZ 1, 3, and 5 (Table 10), regardless of the building orientation, thus not complying with the Proinfância Manual requirement. On the other hand, Type 2 achieved Intermediate and Minimum levels in CZ 2 and 4, respectively, thus complying with the preschool building requirements. Again, the classification did not change regardless of the building orientation. It is important to highlight that Type 2 long-standing rooms’ resultant temperatures as low as 7ºC achieved an Intermediate thermal performance level, not corresponding to thermal comfort for the occupants.

In general, it can be said that Type 2 is much more suitable for the summer period than for the winter period, achieving better levels of thermal performance in warmer climates than in colder climates.

Furthermore, some CZs obtained the same thermal performance classification for all building orientations. Only CZ 8 presented an evident indication of better thermal performance in summer when oriented to the west. Also, CZ 4 e 5 presented some indications in the summer period. As a result, different building orientations have a low impact on the thermal performance assessment results, not providing evident indications about the most proper building orientation for all Brazil’s climatic zones.

The thermal comfort assessment’s results were more sensitive to building orientation changes, generating explicit recommendations for implementation.(Figs. 7-14) show the results of the Annual Occupancy Period within Thermal Comfort conditions (AOPC) for all CZs and building orientations for 23,328 accumulated hours per year.

Fig. (7). AOPC in CZ 1 (hours).

Fig. (8). AOPC in CZ 2 (hours).

Fig. (9). AOPC in CZ 3 (hours).

Fig. (10). AOPC in CZ 4 (hours).

Fig. (11). AOPC in CZ 5 (hours).

Fig. (12). AOPC in CZ 6 (hours).

Fig. (13). AOPC in CZ 7 (hours). Fig. 14 – AOPC in CZ 8 (hours).

Fig. (14). AOPC in CZ 8 (hours).

In CZs 1, 2, 3, 4, and 5, colder zones with higher annual thermal amplitude, Type 2 achieved the best AOPC results when north building orientation was adopted. In CZs 1, 2, 3, and 4, the worst AOPC results were obtained when the south building orientation was adopted, while in CZ 5, the worst was the east, although the south orientation achieved practically the same result. However, in the warmest CZs and with the lowest thermal amplitudes during the year - CZ 6, CZ 7, and CZ 8 - the results indicate the south orientation as the best, with close results for south and north in the CZ 8.

The east and west orientations did not present the highest AOPC in any CZ, even when the longest facades are east and west oriented.

Considering the Brazilian CZs in low latitudes and tropical climates, the results corroborate Hashemi and Khatami's [23] findings, who pointed out that the organization of main facades and openings of the building are recommended for north-south orientation while openings to east and west orientations should be avoided. Rheingantz [14] highlighted the importance of the right building orientation in Proinfância preschools and affirmed the vantage of defining the main facades in the North-South orientation, installing the openings to the North.

The CZs’ results showed a sensible variation in the AOPC when compared to each other. Thermal simulations revealed operative temperatures within the comfort zone more frequently for warm periods than for cold periods. Warmer CZs achieved higher AOPC than colder zones like CZs 1, 2, and 3. CZs 4, 5, 6, and 8 stood out for presenting the highest AOPC values, corroborating with the results of the thermal performance assessment, which also indicates a better adaptation of Type 2 to warmer climates. Type 2 achieved thermal comfort conditions for 46% of the total occupied hours for the best building orientation result in CZ 1, while CZ 8 obtained 87% of thermal comfort conditions for the total occupied period. So, it is expected to almost twice the comfortable temperatures’ period in warmer zones than in CZ 1, example is given in Fig. (15). This difference is significant and highlights the better suitability of Type 2 for warmer regions.

The Proinfância Program Manual contemplates the possibility of closing the perimeter of the internal roofed courtyard with glazed walls in cold regions. Although this internal yard was not simulated as a separated thermal zone in this study, it is essential to note that this open space is used as a lunch and recreation space for children during the whole year. When Type 2 is built in CZ 1, children have breakfast at under 10ºC typical temperature during winter, demonstrating the importance of closing the internal yard in colder CZs. However, like the HVAC systems, the Federal Government does not allocate funds for this building adaptation, making it more difficult to be adopted by local administrations.

In its turn, the CZ 7 stands out for presenting operative temperatures above the comfort zone since external temperatures are extremely high in this CZ. Even though Type 2 usually shows better results for warm periods, it also shows low AOPC values in regions where high temperatures are persistent. Fig. (16) shows an average AOPC of all building orientations compared to the total of 23,328 accumulated hours per year. It also shows how CZs 4, 5, 6, and 8 have a higher average than others.

Fig. (15). AOPC percentage of the best building orientation in each CZ.

Fig. (16). Average AOPC for all building orientations in each CZ (hours).

Fig. (17). AOPC variation between the best and the worst results of building orientation (hours).

When comparing the best and the worst building orientation in each CZ, Type 2 presented the most significant AOPC variation in CZ 2 - 227 hours - being north with 12,212 hours and south with 11,985 hours. Conversely, Type 2 showed the smallest variation in CZ 8 with only 48 AOPC from the North (highest value) to east (lowest value) orientation. CZs 4 and 7 also stood out for presenting a significant difference in AOPC (Fig. 17).

The results showed few variations in some CZs due to the layout of Type 2. Geometrically, the preschool design has well-protected facades by solariums and service areas, possibly reducing the building orientation impact on AOPC. Consequently, these facades do not suffer as much from the influence of direct solar radiation. On the opposite side of these facades, the rooms are also protected by internal corridors and by the roofed courtyard, making them well protected from direct external influence. Besides, the building design has a certain symmetry between the longitudinal and transversal axes, as shown in the Type 2 plan (Fig. 5), producing similar results for some building orientations.

Fig. (17) represents the sensitivity of the building to different building orientations for each CZ. Results are the difference between the best and the worst AOCP for all building orientations in each CZ. These results are caused by the interaction of the school’s geometric design, the building orientation, its construction systems, and the specific climate where the school is located. Even if the AOPC variation is small, it means an improvement in thermal comfort that can be obtained without additional costs. The higher the AOPC variation, the greater the influence of building orientation on Type 2 thermal comfort results, consequently, close attention must be paid to the implantation of the building. City managers must take special care when choosing the Type 2 building orientation in CZs 2, 4, and 7. On the other hand, changes in building orientation are less relevant in CZs 3 and 6. CZ 8 requires the least attention when implementing Type 2.

The rotation of Type 2 causes a significant AOPC variation in some CZs while small differences are observed in others, demonstrating the local climate's impact on the building orientation definition. The thermal comfort assessment’s results were more sensitive to building orientation changes than the thermal performance assessment’s results.

A list of building orientation recommendations for each CZ was elaborated and might be used by the representative cities’ local administrations Fig. (18). It also might be used by nearby cities within the same CZ and that do not have an available weather file for the necessary simulations. After the simulation results were collected, the decision-support framework (Fig. 6) was used to organize the data and guide the recommended building orientation. In CZ 3, for example, thermal performance simulation results indicated that the preschool did not reach the required NBR classification with any building orientation in winter. In this case, the decision-support framework recommends that the preschool design needs to be reviewed. In CZ 4, thermal performance simulation results presented that the best thermal classification for the preschool was North and West. In these cases, as the decision-support framework indicates, the thermal comfort simulation results must be taken into consideration, showing that the building orientation with the highest AOPC is North. As both thermal performance and thermal comfort assessment indicate the building orientation North, the coincident result should be chosen (Decision-support framework: option D.1).

An adequate building orientation towards thermal performance and thermal comfort might not be possible in some available sites. Eventually, some building orientation might make access to a preschool difficult, disturbing the local traffic, for example. In this case, municipalities could choose another site to place the preschool conveniently.

Fig. (18). Building orientations’ recommendations for the Type 2 implementation in each CZ. Obs: the building orientation refers to the main access facade.

However, it is essential to note that building orientation may not be the priority aspect for choosing the implantation site. The preschool's location's social and urban conveniences can override thermal comfort and energy efficiency issues.

Thus, in addition to presenting possibilities for suitable building orientations, the chosen site must also fulfill its social function in the urban network.

It is important to note that the thermal assessments were performed in naturally ventilated rooms since the Proinfância Program does not finance air conditioning systems’ acquisition. However, even if some municipalities afford to install an air conditioning system, the HVAC energy consumption costs could be reduced because of the possibly lower resultant thermal loads achieved through the convenient building orientation. In this way, this method could also result in public money savings.

Although the results presented in Fig. (18) are particular for the Type 2 orientation definition, the developed method might be applied to define other Proinfância Program SPD’s building orientation, like Type 1. Similarly, the method could be considered to be applied to define other standardized typologies’ orientation since they share the same broad characteristic of being occupant-controlled naturally ventilated buildings, even in different climates.