**1. Introduction**

Nowadays, global climate change is becoming one of the greatest environmental challenges [1]. It includes serious disruptions to the weather and climate patterns around the world, such as the impacts on rainfall, extreme weather events, and sea level rises, rather than just moderate temperature increases. One of the main reasons for the current environmental pollution problem is the excessive use of energy [2,3]. Across the entire world, buildings are major consumers of energy and major sources of greenhouse gas emissions. Around 30–40% of energy consumption and 30% of CO2 emissions come from buildings [4–7]. Over the last 50 years, there has been a growing demand for houses and the energy necessary to run them due to a rapidly increasing world population [8,9]. According to the International Energy Agency (IEA), between 2012 and 2030, the building sector's total energy consumption will increase by 4.74 quadrillion Btu (QBtu).

IEA reports indicate that space heating and cooling are responsible for 30% of all energy consumption [7]. Passive cooling is considered part of an overall environmental design strategy that attempts to provide comfortable conditions in the interior of a building and to minimize buildings' energy consumption [10]. Previous research has reviewed passive cooling techniques for buildings [10], examined the influence of environmental and building factors on the performance of passive cooling, and focused on improving

**Citation:** Liu, C.; Xie, H.; Ali, H.M.; Liu, J. Evaluation of Passive Cooling and Thermal Comfort in Historical Residential Buildings in Zanzibar. *Buildings* **2022**, *12*, 2149. https:// doi.org/10.3390/buildings12122149

Academic Editors: Yue Wu, Zheming Liu and Zhe Kong

Received: 9 October 2022 Accepted: 30 November 2022 Published: 6 December 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 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/).

passive cooling techniques [11]. When incorporating passive cooling measures, thermal comfort should always be considered another important factor when evaluating the indoor environmental quality of buildings [12–15].

Over the last decade, the energy efficiency and thermal comfort of historical buildings have attracted increasing attention globally [16–18]. In addition to the balance between energy and societal climate, the improvement of the energy efficiency of historical buildings should also take the building conversation and cultural heritage into consideration [16]. Numerous previous studies have focused on the development of the numerical modeling of passive cooling techniques for historical buildings, the evaluation of air quality when using passive cooling techniques in historical buildings, and energy efficiency evaluations of heating systems in historical buildings; however, little research has systematically evaluated the passive cooling effect on thermal comfort [19–21]. Ricciardi et al. and Diler et al. evaluated the thermal comfort of historical theaters and mosques [12,22]. However, historical residential buildings, which are more numerous than theaters or mosques and are more closely related to daily life, are not taken into consideration. For historical buildings, passive cooling strategies in terms of orientation, construction materials, openings, and shadings are limited to the local culture. Therefore, there is still a need to evaluate the effect of the passive cooling effect on the thermal comfort of historical buildings in hot–dry or hot–humid areas.

The purposes of this research are to explore the passive cooling features used in historic residential buildings, as well as to evaluate the effects of north-south orientation, natural ventilation, pitched roof, light color finishing, and window shading on the thermal comfort of these historic buildings. This study is expected to provide guidelines to assist architects in designing energy-efficient residential buildings, taking into account both the cultural heritage and thermal comfort of the buildings.

#### **2. Research Methods**

The developing world faces greater challenges than the developed world in terms of both the impact of climate change and the capacity to respond to it. These factors are even crucial for a country, such as Zanzibar, with high population growth and increasing energy demand. The Stone Town of Zanzibar forms a unique urban settlement due to a combination of geographical and historical circumstances. Due to its unique architecture and culture, it was declared a UNESCO World Heritage Site in 2000 and is one of the best manifestations of trade and maritime history on the East African Coast [23,24]. Therefore, historical residential buildings in Zanzibar were selected as the case study areas for the evaluation of the passive cooling technique and its effect on thermal comfort.

The study area was Stone Town, located in the urban historical site of Zanzibar; it has a humid tropical monsoon climate and is in a tropical region of East Africa with an equator line passing through it. Stone Town is a city of prominent historical and artistic importance in East Africa. Its unique architectural style mostly dates back to the 19th century, which reflects not only its once powerful role as the capital of an empire spanning two continents, but also the cultural fusion that it has encouraged and nurtured over the centuries. Several historical buildings in Stone Town can be found on the seafront, such as former palaces of the sultans, fortifications, and mosques.

In order to evaluate the thermal comfort of the historical buildings of Stone Town in Zanzibar, questionnaires and field surveys were applied regarding two buildings. The questionnaire design, selection of case study buildings, and physical measurement details in the field survey are presented in this section.

#### *2.1. Questionnaire Survey*

The questionnaire surveys were conducted over a series of four weeks in February, the summer season, in Stone Town, Zanzibar. The questionnaire survey was conducted three times a day, between 7 a.m. and 10 a.m., 12 p.m. and 3 p.m., and 5 p.m. and 9 p.m. The sample size of this study was 180, which accounts for 10% of the total population of Stone Town Historical City [25]. All participants lived in Stone Town, Zanzibar.

Subjective thermal comfort data were recorded using a questionnaire designed with considerations given to ASHRAE Standard 55 [26]. The questionnaire developed for this survey was divided into four main sections, as follows.

Part 1: Background and demography of the respondents. The first section consisted of general information about the participants, including their age, gender, and years of living in their current building.

Part 2: Building features. The second section covered several questions about the building's features, such as openings (door, window types, and materials), the height of the building, wall and roof materials, and the use of different thermal environment control means (windows and local fan/air condition/shading devices).

Part 3: Indoor and outdoor environment information. The indoor environment information included temperature, relative humidity, air velocity, and satisfaction evaluations regarding indoor thermal environment quality and personal environmental control. The outdoor environment information data during the field survey period were obtained from the Zanzibar Meteorological Department, including factors such as temperature, air velocity, and sun radiation.

Part 4: Thermal comfort evaluation. Thermal comfort was rated on an ASHRAE 7-point scale with the following aspects: thermal sensation vote (TSV, −3: cold, −2: cool, −1: slightly cool, 0: neutrality, 1: slightly warm, 2: warm and 3: hot), humidity predicted vote (HPV, −3 much too dry, −2 too dry, −1 slightly dry, 0 just right, +1 slightly humid, +2 too humid, +3 much too humid), and draft perception vote (DPV indicates that the air is perceived to be: −3 much too still, −2 too still, −1 slightly still, 0 just right, +1 slightly right, +2 too breezy, +3 much too breezy). The McIntyre scale was used to assess thermal preferences; it is a 3-point scale. The questions asked were, "Would you like to have it warmer (1), no change (0), cooler (−1)?" or "At this moment, would you prefer to feel warmer, cooler, or no change?" The thermal acceptability vote was used to determine the judgment regarding the perception of the thermal environment. The question was, "At this moment, do you consider the thermal environment acceptable (0) or unacceptable (1)? Detailed questionnaires can be found in Appendix A.

All the information collected by the questionnaire surveys was saved and analyzed in SPSS25.0.

#### *2.2. Field Survey*

To evaluate the thermal comfort level of the indoor environment in historical buildings, a field survey was carried out in two residential houses. The buildings were monitored for six consecutive days in February, as February is the hottest month and is the period in which the northeast monsoon wind blows relatively strongly. Each building was surveyed three times a day, between 7 a.m. and 10 a.m., 12 p.m. and 3 p.m., and 5 p.m. and 9 p.m., together with the questionnaire.

2.2.1. Selection of the Case Study Buildings

The criteria for selecting buildings for the field study were as follows:


Finally, two historical residential buildings, named HRB1 (Figure 1a) and HRB2 (Figure 1b), were chosen for the case study. The floor plans of HRB1 and HRB2 are presented in Figures 2 and 3.

HRB1 is a single-family residential house located in Shangani Ward and is 15 m from a main road. The building has a rectangular shape and a surface area of approximately 183.75 square meters. It is a three-story building; the ground floor is used as a shop, but it was not surveyed in this research, while the remaining upper two stories are used for residential purposes with a single family of four members. The first floor of this building is used as a living space, and the second floor is used for bedrooms (Figure 2).

**Figure 1.** Case study of historical residential buildings. (**a**) HRB1, (**b**) HRB2.

**Figure 2.** Floor plan for HRB1. The positions of the thermohygrometers are marked with red dots (Points A and B). (**a**) First floor, (**b**) Second floor.

**Figure 3.** Floor plan for HRB2. The positions of the thermohygrometers are marked with red dots (Points C and D). Spaces enclosed with dashed lines were studied in this research. (**a**) First floor, (**b**) Second floor.

HRB2 is a multi-story building located in Malindi Ward. The building has a square shape and a surface area of approximately 360 square meters, which is two times the surface area of HRB1, with a 4 m-by-4 m courtyard accessible only by ground floor users. There are two dependent flats/units on the ground floor and another two flats/units on the first and second floors, which together give a total of four flats/units and four households in HRB2. Due to time and financial constraints, only one representative double-floor apartment was selected for the study. The first floor of this apartment is used as a living space, and the second floor is used for bedrooms (Figure 3).

#### 2.2.2. Physical Measurements

The indoor environmental parameters, such as air temperature and relative humidity, were measured using a thermohygrometer device (TES1341) for a measurement period of approximately 10 min at an interval of 1 min. Three sampling points were selected, which resulted in 18 sampling points for this study in each surveyed building. For HRB1, the thermohygrometers were hung on the wall in the living room (Point A), bedroom 2 (Point B), and outside (Figure 2). For HRB2, the instruments were hung in the living room (Point C), bedroom 3 (Point D), and outside (Figure 3). The thermohygrometers were placed at a height of 1.1 m above the floor, which represents the height of the neck of a seated occupant, and 1.5 m away from the wall. Other environmental parameters, such as air velocity and direction, were assumed to be the same as those measured by meteorological stations.

#### **3. Results**

The results of the questionnaires and field studies from the study period are given in this section.

#### *3.1. Questionnaire Survey Results*

A total of 159 of 180 samples were valid for the questionnaire survey. The gender proportion of the samples was as follows: female, 96 (60.37%), and male, 63 (39.6%).

#### 3.1.1. Building Envelope Evaluation

A total of 100% of the respondents' results showed that their buildings were built using coral stone, as shown in Table 1. Stone coral was actually beneficial to the thermal conditions of the building. Their heavy structure cools indoor spaces by delaying heat transmission from the outside, but in the evening, they release heat to the interior space (as well as to the outside). More air movement via the opening of windows is helpful for decreasing indoor temperature at night and in the early morning.


**Table 1.** Results of the building envelope evaluation.

<sup>1</sup> CVB indicates curtains and verandah/balcony; VB indicates verandah/balcony; C indicates curtains.

Regarding openings, 26.41% of the occupants had between 1 and 7 openings, while 63.52% had between 7 and 14 openings, and the remaining 10.06% had >14 openings in their buildings. This shows the awareness of historical builders regarding natural ventilation. Nearly two-thirds (68.55%) of the respondents use louver/shutters; this kind of window can help to prevent sun rays from penetrating inside the building.

Regarding ceilings, the occupants' replies showed that 68.55% of their buildings have traditional ceilings, which are mangrove ceilings; this type of ceiling is made up of earth and mangrove poles. A total of 20.75% of the respondents had changed the ceiling to a hardboard ceiling, and 10.69% had reinforced concrete ceilings.

Regarding roof type, a total of 83.64% (hip and pitched roof) of the surveyed buildings had a roof structure that acts as a barrier against heat entering the living space and also creates a space under the roof, as it is the most common and economical heat insulator in every building.

A range of sun shading devices were listed in the survey, in which the respondents indicated which shading method they used and specified a reason for using the shading. A total of 15.72% of the survey participants indicated that they used curtains and a verandah/balcony (CVB) as their choice of shading, and 10.69% indicated that they used a verandah/balcony (VB). A total of 73.58% indicated that they used curtains (C).

#### 3.1.2. Indoor Environment Parameter Assessment

According to ASHRAE Standard 55, the factors required for the evaluation of thermal comfort include metabolic rate, clothing insulation, air temperature, radiant temperature, air speed, and humidity. The two personal parameters, such as clothing insulation value and metabolic rate, were estimated in accordance with ASHRAE Standard 55 and ISO 7730 [26,27]. In this study, we set the metabolic rate as 1.2 met, which represents sedentary activities (office, dwelling school, and laboratory), and the clothing insulation value as 0.5 Clo.

The overall indoor air temperature taken in each dwelling during the questionnaire survey ranged between 26 ◦C and 32 ◦C, and the indoor relative humidity ranged from 58.90% to 75.05%, with a mean value of 63.44%. Based on these data, the operative temperature *t*<sup>0</sup> was calculated according to ASHRAE Standard 55 as:

$$t\_0 = At\_a + (1 - A)\overline{t\_r} \tag{1}$$

where *ta* is the averaged air temperature obtained from questionnaire, *tr* is the mean radiant temperature, which is assumed to be equal to the air temperature, and *A* is a function of the average air speed. The air speeds were assumed to be the same as those measured by meteorological stations. The calculated operative temperature was 29.1 ◦C on average, which is outside the comfort zone limit of ASHRAE Standard 55.

The results of the evaluation of the indoor humidity perception vote (HPV) and the indoor draft perception vote (DPV) are presented in Figure 4. Only 63.51% of 159 respondents voted for HPV within the three central categories, showing that the occupants did not have acceptable thermal conditions. A total of 26.42% of the residents felt that the environment was much too humid, which is understandable in such a humid area, with a relative humidity between 58.90% and 75.05%.

**Figure 4.** Evaluation of the indoor humidity perception vote (HPV), indoor draft perception vote (DPV), and thermal sensation vote (TSV).

Regarding the DPV, a large portion (52.19%) of the subjects perceived the air to be too steady (−3 to −2), while 31.44% of the respondents felt that the air velocity was just right (−1, 0, +1), and only 16.35% felt that the velocity should be less. These results show that air movement and thermal conditions in their buildings were not acceptable.

#### 3.1.3. Thermal Sensation Vote (TSV)

The results of the TSV evaluation showed that the majority (73.57%) of the respondents voted between "−1 slightly cool" and "+1 slightly warm", while almost half (47.16%) of the respondents voted slightly warm. Linear regression was performed between thermal sensation and operative temperature to determine the strength of the relationship between them in residential buildings. The data are plotted in Figure 5. The fitted regression equation for the subjects' sensations versus the operative temperature was:

$$\text{TSV} = 0.650 \text{ T}\_0 - 17.737. \tag{2}$$

**Figure 5.** Linear regression of thermal sensation votes (TSV) versus operative temperature (OP).

The coefficient of determination (R2) between TSV and the operative temperature was 0.875, which means TSV correlated strongly with the operative temperature. The slope of the regression line, which indicates the subjects' thermal sensitivity concerning the operative temperature, was high (0.65 unit/◦C) compared with that in other warm places [28–30]. This demonstrates that the residents of Stone Town, Zanzibar, are less tolerant of a wider range of temperatures.

The neutrality condition derived by solving the regression equation for zero (neutral) yielded an estimated neutrality of 27.40 ◦C, as shown in Figure 5. A comfort band of 26.00–28.90 ◦C coincided with −1 and +1 sensation votes.

#### 3.1.4. Thermal Preferences

Thermal preference was evaluated using the McIntyre scale (−1 (cooler), 0 (no change), +1 (warmer)). A total of 68.55% of the respondents preferred to be cooler in their buildings, and only 31.44% of the subjects indicated that no changes in the thermal environment were required. These results suggest that people prefer to feel cooler than neutral, which aligns well with previous results from studies regarding warm places [28,31]. The thermal acceptability results align well with those of thermal preference. A total of 68.55% of the participants judged that their thermal sensation was not acceptable, and 31.44% accepted the overall temperature of their building.

#### *3.2. Field Survey Results*

#### 3.2.1. Passive Cooling Features

The passive cooling features of the two case study buildings (HRB1 and HRB2) regarding orientation, materials, and construction, and cooling and ventilation are presented in this subsection and are summarized in Table 2.


**Table 2.** Passive cooling features of case study buildings HRB1 and HRB2.

<sup>1</sup> N indicates north; S indicates south; E indicates east; and W indicates west.

#### Orientation

HRB1 is surrounded by four streets of 1.5 m to 2 m in width and is approximately 15 m from the Kenyatta main road. Being surrounded by higher buildings, HRB1 receives shade. The long axis of the building runs from west to east, which means the façade on the north–south axis is bigger than that on the west–east elevation axis. It has a good orientation toward the wind direction and deflects the dominant orientation of the sun in the east–west direction. None of the bedrooms are located on the west side, which is a disadvantaged side. Even though the living room is located with an eastern orientation, it does not receive much heating during the day because of the shading element (balcony on the first floor) along the whole length of the façade. Traditionally, balconies are used to protect walls from direct solar radiation and act as transitional spaces to inner living areas. No air conditioning (AC) was in operation during the investigation.

HRB2 is bounded by three streets about 1.5 m to 2 m wide on the northern, southern, and western sides of the plot. The eastern side has shared a wall with another building, but not the whole length of the building because in the middle there is a courtyard/air well, which prevents the building from being totally attached. The long axis of the building runs north to south, which is contrary to the recommendation of the best orientation for buildings in tropical climates. The living room, bedroom 1, and bedroom 2 of HRB2 are located on the west side; therefore, they heat up in the afternoon, although there is shade protection on the window to break up the sun rays and curtains for protection from solar heat gain in the living room and bedrooms. Bedroom 3 is well-oriented at the southern side of the building; this room also benefits from air movement and natural light through the courtyard/air shaft.

#### Materials and Construction

Natural, local building materials made from stone, wood, and soil were found to be used in this case study. Both HRB1 and HRB2 are constructed with coral rags laid in random gravel. The exterior walls of these buildings are very thick, with a 1 m thickness, and act as load bearing walls; the interior walls are 500 mm thick, with the exception of bathroom/kitchen walls, which are 200 mm thick; the thick stone walls act as thermal mass against heat penetration.

The exterior and interior walls of HRB1 are painted in white, which saves the cooling energy of the building [32–34]. Similarly, the outside wall finish of HRB2 is rough with pale yellow paint, while the interior walls are painted white.

There is no insulation on the walls, windows, or roofs of HRB1 and HRB2 because they are expensive and difficult to maintain. HRB1 has a pitched roof of 32◦, and the roof overhangs (600 mm), providing a sun shading effect to protect the walls. The existence of a ceiling also creates an attic space under the roof, as it is the most common and economical heat insulator that must be installed in every building. The roof overhangs can provide a sun shading effect to protect the walls. The floor slabs of this case study were found to be traditionally made of boriti poles (mangrove poles) with a lime–concrete cover laid over to form the floor above. The boritis or mangroves support the slab and help the walls remain upright. HRB2 has a flat roof; all rooms are prone to heat gain through the floor slab, even though the ceiling is painted with a whitewash.

### Cooling and Ventilation

The living room, dining room, and bedroom 1 in HRB1 have provisions for crossventilation via four pairs of windows for bedroom 1, three pairs of windows for the living room with a jalousie door that opens up to provide maximum airflow inside the spaces, and lastly, two pairs of windows for the dining room; thus, it seems that the spaces inside the room are affected by the air flow.

Regarding HRB2, the living room and bedroom 1 have provisions for cross-ventilation via three pairs of windows for bedroom 1 (two openings on opposite and adjacent walls) and three pairs of windows for the living room (two openings on opposite and adjacent walls); thus, it seems that the spaces inside the room are affected by the air flow. The windows in HRB2 are all double-winged shuttered windows. Although the windows in most of the rooms have been designed to enhance cross-ventilation, in this case, there are some windows that are not open every day, e.g., the living room window with a southern orientation is not opened during the whole year; this is because of the furniture (showcase) kept in front of the window. Lattice windows are used on the west façades of this building to enhance cross-ventilation and daylight. The shading devices and shuttering of the windows in the case study building are 400 mm by 2000 mm in dimension on all sides of the building. The orientation for wind flow is quite challenging only on the northeastern side (blockage of air flow) because, on its eastern side, the building shares a common wall with another building, but the historical builders demonstrated common sense by introducing a 4 m-by-4 m courtyard/air well that allows air flow and natural daylight to enter the building.

#### 3.2.2. Air Temperature and Humidity

In general, the indoor air temperature in the occupied zone for HRB1 varied between 28 ◦C and 31 ◦C, with an average value of 29.2 ◦C, and the indoor relative humidity was between 57.0% and 70.3%, with a mean value of 62.41% (Figure 6). For HRB2, the indoor air temperature as measured during the field survey varied between 26 ◦C and 29 ◦C, with a mean value of 28.16 ◦C (Figure 7). The indoor relative humidity in the building ranged between 49.4% and 76%, with a mean value of 63.89%. As presented in Section 3.2.1, the living room of HRB1 has six openings in three walls, which creates cross-ventilation. In bedroom 2, there is only one opening in the southern wall, and it is expected to have higher temperatures and humidity than in the living room. However, no significant differences between the temperature and relative humidity of the living room and bedroom 2 were found, as shown in Figure 6. A possible reason for this could be the diverse surface areas of the two rooms. For HRB2, both the living room and bedroom 3 have openings in different directions, which creates cross-ventilation. However, the temperature and humidity in the bedroom were found to be slightly lower than in the living room. This is because the openings in the living room are covered by furniture that cannot be opened frequently. Another reason is that the living room faces west and absorbs more sun radiation than

bedroom 3, which has a southern orientation. Therefore, orientation, surface area, and air flow are all important factors in indoor environments.

**Figure 6.** Indoor/outdoor temperatures and humidity averaged from 07:00–10:00, 12:00–15:00, and 17:00–21:00 (HRB1).

**Figure 7.** Indoor/outdoor temperatures and humidity averaged from 7.00–10:00, 12:00–15:00, and 17:00–21:00 (HRB2).

Figures 8 and 9 show the performance of both the living rooms and bedrooms in relation to the outdoor air temperature for HRB1 and HRB2. For HRB1, the outdoor air temperature value rose from 28.5 ◦C in the morning (07:00–10:00) to a peak of about 30.5 ◦C during the afternoon (12:00–15:00). The average temperature in the living room and bedroom was slightly different from that in the outdoors. This is because the indoor temperature rises as solar radiation increases. However, in the evening, the indoor temperature did not decrease as dramatically as the outdoor temperature, which means that, at night, the outdoor space is better than the indoor space. For HRB1, no significant change was found in the living room and bedroom temperatures, in contrast to the variation in the outdoor temperatures. However, for HRB2, the living room temperature was always

higher than the outdoor temperature, while the bedroom (south) temperature was between 27.0 and 27.5 ◦C, which is much lower than the outdoor temperature.

**Figure 8.** Temperature in the living room, bedroom, and outdoors as averaged over six days of measurements (HRB1).

**Figure 9.** Temperature in the living room, bedroom, and outdoors as averaged over six days of measurements (HRB2).

#### 3.2.3. Thermal Comfort Indices

The thermal-related input values for the calculation were determined via field measurements and are shown in Table 3 for HRB1 and Table 4 for HRB2. The calculations for the PMV and PPD were performed using a software comfort program that has been proposed by ISO 7730 [27].

The PMV of HRB1 was found to vary between 1 to 1.6, with a mean value of 1.23. A total of 80% satisfaction falls within the range of [−1.23 < PMV < +1.23]. The PPD varied from 26.1% to 56.3%, with a mean value of 37.35%. This indicates that 37.35% of the occupants in this survey building were expected to be dissatisfied (discomfort) with the thermal environment.


**Table 3.** Results of field measurements and thermal comfort indices (input variables for PMV and PPD) for HRB1.

**Table 4.** Results of field measurements and thermal comfort indices (input variables for PMV and PPD) for HRB2.


For HRB2, the PMV was found to range between 0.7 and 1, with a mean value of 0.85. A total of 80% satisfaction falls within the range of [−0.85 < PMV < +0.85]. The PPD varied from 15.3% to 26.1%, with a mean value of 20.56%. This indicates that 20.56% of the occupants in this survey building were expected to be slightly dissatisfied (discomfort); the required PPD value should be less than 10%. Additionally, in this case, we learned that the value of the PMV fell within the range of (−1, 0, +1), which represents 80% satisfaction, but, unfortunately, the values obtained are not in the ideal range of an acceptable thermal environment for general comfort, which is between [−0.5 < PMV < +0.5] (neutral).
