Influence of Roller Blinds Shading Strategy on West and South Facing Buildings

Abstract

This study investigates the impact of roller blinds as an energy-saving strategy in a subtropical climate. Lighting control experiments were performed on west-facing and south-facing windows in the Subtropical Performance-Testbed for Innovative eNergy Research in Buildings Laboratory (SPINLab). This study accepts sunlight below 8000 lux and requires the indoor lighting of at least a reading of 500 lux. For the west-facing window experiment, the curtain energy-saving strategy implemented in this study resulted in 10% less energy consumption for air-conditioning and saved 90% of the lighting power consumption of the west facing window building. For the south-facing window experiment, we found that sunlight can fully supplement the lighting without adjusting the roller blinds because the window opening is different from the sunlight direction. In addition, the significant heat flux difference between the west-facing and south-facing windows experiments is also reflected in the energy consumption of the air conditioners.

1. Introduction

Rapid economic growth in various countries prompted increased fossil fuels consumption; such growth in energy consumption may hasten the biofuel energy shortage situation. Thus, developing sustainable energy consumption technology is critical.

The U.S. Department of Energy (DOE) noted that, from 1980 to 2010, the proportion of building energy consumption in total energy consumption increased from 33.7% to 41.1% [1]. In order to become net-zero buildings, buildings should first achieve energy self-sufficiency. This means that the structure itself should fulfil its own energy needs. Therefore, the first task is to reduce the essential energy consumption inside the building as an effort to reduce its pressure of independent power generation.

Energy consumption in buildings can be reduced through passive energy-saving technologies, which focuses on building materials and shading equipment; another is through energy-saving service systems based on strategic adjustment of electrical equipment such as HVAC and lighting [2]. Passive energy-saving technologies can be further divided into building envelopes, passive shading equipment, and intelligent shading systems.

The building envelope is the contact structure between indoor and outdoor environments. The U-value of building materials directly affects the heat loss rate capacity inside the building. Both local climate and building materials significantly impact thermal comfort and energy-saving efficiency. Secondly, passive shading equipment refers to devices, either fixed or manually operated inside and outside the building, commonly used to block direct sunlight. Passive shading equipment affect indoor thermal comfort and can improve visual comfort [3]. Louver and roller blinds are the main passive indoor shading devices. Both have potential in adjusting the shading rate and updating the automation mechanism. However, passive energy-saving technology is also susceptible to local climates. In cold climates, passive energy-saving technology can improve visual comfort but may lead to increased heating equipment load. Conversely, in tropical climates, it reduces cooling apparatus loads and improves visual comfort [4].

Finally, the intelligent shading system provides adjustments to automation or energy-saving strategies under the premise of passive shading equipment. Recent research showed people are still committed to improving the thermal insulation performance of building shells [5]. However, results from other studies illustrated that climatic limitations of passive shading equipment [6,7,8] tend to increase the energy loading of indoor lighting devices [9,10]. Therefore, it is imperative to establish shading devices that can be passive, dynamic or even intelligent. Perino et al. [5] proposed that to achieve zero-energy building, energy-saving technology should evolve from a single device corresponding to a single to the concept of a single device corresponding to multiple functions, such as establishing an intelligent shading system that combines shading and solar power panels simultaneously.

According to the shading system integrated by Salwa et al., in 2019 [11], the dynamic complexity and diversity of shading systems have gradually increased; in addition, according to the recorded results, there are fewer research projects in subtropical climate conditions. Most experimental-based research considers the influence of indoor energy consumption and daylight, but few research cases discuss the architectural aspect.

Buildings with near-zero carbon emissions are currently uncommon in Taiwan. Although there is inefficient automated equipment layout, it can be easily found from previous literature that automated shading equipment is not only a trend, but also an excellent option that has been proven to provide better daylight performance and energy saving [12]. A review article published in 2018 by Niraj et al. [13] compiled relevant research on dynamic shading in buildings in recent years. A few important points worth noting are: (1) Few relevant studies evaluating the impact of dynamic shading on thermal comfort, (2) Many studies exploring the impact of multi-directional dynamic shading systems exist [14,15,16]. However, the implementation of full-scale experimental tests is very difficult, so there are few relevant experimental results to assist the refinement of building energy models. (3) Research focusing on the impact of dynamic shading systems and lighting controls on cooling or heating systems is mostly done by simulation [12,17,18]; thus, there is insufficient research on energy-saving benefit evaluation through full-scale experiments. According to the results of the current literature review, it was not until 2019 that Niraj et al. [19] conducted a full-scale multi-directional building dynamic shading experiment in Iowa, USA, but related research is still very rare. In this study, a full-scale experiment in a subtropical climate was carried out by SPINLab. Taiwan’s geographic location (23°0′ N) is closer to the equator than Iowa (40°23′ N–43°30′ N), and Taiwan, an island surrounded by sea, has a distinct climate difference from Iowa, which is located in the middle of the United States. In addition, SPINLab has a rotatable mechanism to make the external environmental conditions of the control experiment as symmetrical as possible, and completes multi-directional experiments from the same two rooms.

Our study aims to: (1) discuss the energy consumption impact of shading energy-saving strategies on subtropical climates is also the focus of this study; (2) understand the construction benefits and costs of investing in the renovation of net-zero carbon buildings; and (3) establish practical and efficient shading systems and energy-saving strategies for smart buildings in the future.

2. Materials and Methods

The experiment was conducted at SPINLab in Tainan, Taiwan. The indoor and outdoor environment diagrams of SPINLab are shown in Figure 1a,b, the configuration is the same as the previous study by Hu et al. [20], but we add the illuminance measure in this study. The experimental site within SPINLab contains a monitoring room and two labs with the same indoor configuration and single-sided window. In addition, the bottom of SPINLab is equipped with several sets of motors to rotate the experimental house and simulate the impact of different house orientations and different sun exposure angles on the building. This experiment used the rotatable feature of SPINLab to carry out the experimental conditions of the house facing west and south, respectively. This experiment installed temperature and humidity sensors on the roof to record outdoor weather conditions.

The dimensions of each laboratory are 6.5 m × 4.8 m × 3.8 m, and its interior layout is shown in the Figure 2. Each laboratory is equipped with a split-type air conditioner with a maximum cooling capacity of 14 kW, whose model is HITACHI RAD-140NX1 (average COP of 4.3). The air return and air outlets of the air conditioner are located at the laboratory’s window position and center position, respectively. Each laboratory has four sets of office desks and chairs, numbered 1–4 as shown in the Figure 2, two of which are arranged with dummies to simulate an office space that accommodates up to four people. At each numbered position in the room, a temperature and humidity sensor is installed 170 cm above the ground to record the indoor environmental information. The temperature sensing accuracy reaches 0.1 °C; the humidity sensing accuracy is 0.1% RH. In addition, this study placed the air-conditioning return air temperature sensor on the wall away from the glass, about 10 cm above the ground. The window had a window-to-wall ratio of 40% (for the single wall with window) and consisted of clear glass with a U-value of 5.02. We installed a heat flow meter at the center of the window and the two walls adjacent to the window to record the heat flux, and the accuracy of the heat flow meter is ±3% of reading (<10,000 lux) and ±4% of reading (>10,000 lux). In addition, the shading rate of roller blinds for shading is 93%. In indoor illuminance records, this experiment uses a portable illuminometer to record the hourly illuminance of each desk desktop. The illuminance meter used in this study has an accuracy of ±5%.

We referenced previous research data to gauge general human body heat dissipation [21,22,23]. As a result, to represent heat load, this experiment is equipped with a heater that simulates two workers and two computers, totaling 1 kW. A graphic configuration is shown in the Figure 2. In addition, the indoor lighting system consists of a group of dimmable lamps with a maximum power density of 11.2 W/m². When the roller blinds are completely closed, the lighting system provides a lighting requirement of at least 500 lux on the work plane with a power density of 4.4W/m².

In addition to two completely symmetrical laboratories, SPINLab also has a monitoring room adjacent to the laboratory at the rear. A data collection and monitoring system is installed in the monitoring room in order to record environmental data from all sensing devices every minute. Through computers and NAS systems, we can read and analyze historical data from sensors in the monitoring room. We can export the required data type, time interval, and other information in the form of an Excel file through the computer. The Excel file format will help us generate various data history charts and conduct complex data analysis work in the future.

We recorded continuously for 10 h from 0800 to 1800, this time range covers the normal working hours, and this setting makes the indoor environment more consistent with the office. Air temperature is set at 26 °C in both Room A and B for the duration of the experiment. The difference between room A and B is that room B has upper and lower illuminance limits, which are controlled by roller blinds. Another difference is that the indoor lighting of Room A is turned on for the duration. We set 500 lux as the minimum illuminance, and started the experiment with the indoor lighting of Room B turned off. Once the illuminance (four points: B1–B4) falls below 500 lux, additional indoor lighting is supplemented. The lower illuminance limit of 500 lux also complies with Taiwan’s national illuminance standard regulation CNS12112 [24]. T. Otsuka et al. [25] stated that illumination above 8000 lux will cause permanent damage to the retina, so the shading strategy of curtains will control the indoor illumination below 8000 lux. Furthermore, a two-stage shading strategy is applied. When the measured illuminance exceeds 8000 lux for the first time, the shading curtains are closed until the sunlight just does not reach the working plane directly, and the second shading is completely closed. The purpose of this setting is the same as that of Tzempelikos et al. [26] in SC-III, which is to apply more natural light into the room to reduce the energy consumption of artificial lighting. The difference is that the distance between the façade and the working plane in this study exceeds the 0.5 m set in the study by Tzempelikos et al. [26], so it is possible to get more natural light to enter, which also affects the energy consumption of air conditioners.

Figure 3 shows the monitoring images of Room A and B on experiment day of SPINLab with a west-facing window. In Room A, indoor lighting is noticeable at 08:00, and sunlight cannot directly illuminate the room. The desk near the window was illuminated by the direct sunshine at 16:00. Finally, the sunlight could no longer reach the room at 18:00. On the other hand, Room B implements our energy-saving strategy designed for this experiment. At 16:00, curtain in Room B is partially drawn to block sunlight from reaching the desktop. Finally, curtain in Room B is completely drawn at 18:00, and the illuminance requirements of 500 lux are made up through indoor lighting.

The PMV index can effectively indicate the human body’s own thermal comfort. The main factors affecting the PMV index are dry bulb temperature, relative humidity, average radiant temperature, wind speed, and the amount of clothing and activity of people. In this study, the dry bulb temperature, relative humidity, and wind speed data from the sensor can be obtained through the data collection system. In addition, since the office is used as the simulation scene, the amount of clothing (clo) = 0.5 and the amount of activity (met) = 1 are set. The value of the average radiation temperature is a simple approximation without measuring the black bulb temperature. Bean [27] provides an easy way to calculate the mean radiant temperature in a space:

T_mr = T₁A₁ + T₂A₂ + … + T_NA_N/(A₁ + A₂ + … + A_N)

(1)

where T_mr means the mean radiant temperature, T_N means the surface temperature of surface N (calculated or measured), and A_N means the area of surface.

Equation (1) demonstrates that it is necessary to extract the measured temperatures of six surfaces in the laboratory when the space conditions (6.5 m × 4.8 m × 3.8 m) are known. Furthermore, the temperature of the surface of the window and the adjacent walls on the left and right can be directly obtained by measuring the temperature of the sensor. Due to the limited number of sensors, this study substitutes the temperature of the wall above and below the window with the temperature of the wall adjacent to the left and right of the window. The floor and intermediate walls adjacent to the monitoring room are estimated by the closest sensor, the air-conditioning return air temperature sensor. A sensor was not installed to measure the temperature of the ceiling. Since thermal stratification is an inevitable phenomenon in air-conditioned rooms, we obtained the thermal stratification temperature difference through other air-conditioning experiments conducted in SPINLab. We estimate that the temperature is 2 °C higher than the air-conditioning return air temperature. It is worth noting that when the curtains were adjusted, the corresponding surface conditions of the room changed; the curtain surface temperature was not measured due to the limited number of sensors. Continuous calculation based on the sensor temperature on the window will overestimate the indoor mean radiant temperature. Therefore, after the curtains are adjusted, this study averages the temperature obtained from the sensor on the window and that obtained on the side of the window to approximate the surface temperature of the curtain.

After we have obtained all the temperature information, we can further calculate the mean radiant temperature. After completing the estimation of the mean radiant temperature, we obtained all the variables needed to calculate the PMV, and we up-loaded these data to the comfort index computer provided by the CBE Thermal Comfort Tool webpage [28] (https://comfort.cbe.berkeley.edu/upload accessed on 20 October 2022).

3. Results and Discussion

Figure 4 is a schematic diagram of the sensor layout in the office space. In Figure 4. The “window” orientation indicates the different building orientations facing south and west. In addition, we recorded the heat fluxes of the three adjacent surfaces of Room A and Room B near the window, and named them according to their orientations, such as AhE, AhW, AhS, AhN, etc. In the center of Room A marked A()1–A()4, we fill in i in () or leave it blank to distinguish lighting or temperature, the same for Room B. Finally, on the side away from the window, Aac and Bac represent the air conditioner temperature sensor of the two rooms respectively.

Figure 5 shows the indoor and outdoor climate conditions for rooms A and B on May 26, with the windows facing west. Information such as temperature, heat flux, and illuminance are also presented in Figure 5 from top to bottom.

In this experiment, the position of the indoor curtains was adjusted for the first time in Room B at 15:10, and the shading ratio reached 65%, as shown in the Figure 5. Illuminance first decreases due to shading, after the curtain position was adjusted for the first time. It then gradually increases due to the change of the sunlight angle. Finally, the indoor curtains were adjusted for the second time at 17:10, the curtains were completely drawn, and the shading ratio reached 100%. Indoor lighting supplements the minimum illuminance requirement of 500 lux, which remains the same as it is not affected by sunlight.

From the top of Figure 5, a 2 °C temperature difference between the window side and the indoor side (ex. A1 versus A4, B1 versus B4 et al.) is due to the window’s daylight and solar radiation effects. The air conditioner sensor (Aac) measured 27 °C at 15:00, which is higher than the set temperature of the air conditioner, 26 °C. Room B has lighting controls at 15:10; there is no corresponding increase in temperature. Then, temperature sensors on the surface of the windows registered the rising temperature, which continued to climb after 1 p.m. above the outside temperature. This setting shows that the thermal energy stored in the window is accelerated due to the larger temperature difference with the indoor environment. However, the measured temperature of over 50 °C in room B after 3 p.m. means that the roller blinds have a good thermal insulation effect, because the heat energy is blocked between the roller blinds and the window, so that the indoor temperature does not rise. Similarly, the cliff-type heat flux process occurring in Room B in Figure 5 can be explained, because the heat energy blocked by the roller blinds is trapped between the heat flow meter and the roller blinds. The temperature difference between the two sides of the heat flow meter becomes smaller, resulting in a smaller heat flux value. Our results indicate that the operation of roller blinds affects indoor illuminance. In addition, the lower heat flux results in a lower temperature rise in Room B during the afternoon. The significant differences in room temperature can be noted with the temperature measurement points near the window (positions A1, A2, B1, and B2). In the afternoon, the temperature of A1 and A2 was significantly higher than that of B1 and B2 because room B had two curtain adjustments. As the shading rate of the curtain increases, the heat conduction from the window is effectively blocked between the curtain and the window. Our experimental results demonstrated a maximum temperature difference of 4.3 °C at 17:30. Because the shading rate of the curtains reaches 100% after 17:00, direct sunlight to in-door space is completely blocked.

Our results also revealed Room B has a similar trough in the morning as Room A, while Room B has a sharp drop in heat flux due to the curtains being adjusted twice at 15:00 and 17:00. In contrast, in Room A, a history of heat flux unaffected by the curtains can be observed, the decrease in heat flux after 17:00 is due to less sunlight entering the room at sunset. Due to the thermal insulation of the roller blinds, the temperature of the heat flow meter on the side close to the roller blinds gradually rises, so that the temperature difference between the two sides of the heat flow meter becomes smaller, and finally the result of a cliff-like drop in heat flux is formed.

In the illuminance results, there are two cliffs in the illuminance graph for the two curtain adjustment responses in Room B. Namely, after the curtains are closed, the illuminance measurement point is not directly exposed to sunlight, thus showing illuminance values that meet the experimental conditions. The different illuminance measurements in Room A and Room B echo the light reflected from the tabletop at 16:00 in Figure 3.

According to historic climate data from the Central Weather Bureau (CWB), the average outdoor temperature in Tainan is 30.5 °C in May and 28.7 °C in June. Our experiment took place on 5/26 (which had a daily average temperature of 32 °C) and 6/10 (which has an average temperature of 30.3 °C). The average temperature on the day of the experiment was at least 3 °C higher than the setting of 26 °C for the air-conditioning, which makes it easier for us to observe the power consumption operation of the air conditioner by the curtain adjustment strategy.

Figure 6 shows the indoor and outdoor conditions for Room A and B marked on either side of 6/10. The building faces south, and the window facing south. Indoor and outdoor data are recorded in the same way as 5/26.

Notably, the south-facing windows are protected from direct sunlight. The indoor illuminance did not meet the conditions for starting the rolling blinds, so the recorded value of illuminance in room B was close to that in room A. In addition, since the solar insolation direction is still east-west, the buildings with windows facing south record lower heat flux values than in Figure 5.

Figure 5 and Figure 6 show that the curtain operation strategy only works when exposed to sunlight after 15:00 on May 26. This activation mechanism can be known from the il-luminance measurement results of Room A and Room B. The illuminance value of Room B should have increased as rapidly as that of Room A, but the illuminance measurement value after the curtain operation strategy was activated at 15:00 was effective to control. The sunshine intensity in June never exceeded the preset upper limit of illuminance. This result shows that the space with windows facing south is less likely to be affected by direct sunlight on the desk surface. On the other hand, windows facing west have the problem of direct sunlight affecting visual comfort. In addition, the temperature rises in the afternoon which affects the shading system on the west side and increases the air conditioning load.

Figure 7 shows the relationship between air-conditioning power consumption and indoor temperature for 5/26 and 6/10. Aac records the measurement data of the air-conditioning sensor, and the measurement accuracy and density of Aac are low, presenting a digital signal-like wave illustration.

A1 and B2 are near the window, which are easily exposed to direct sunlight, so the temperature is higher. A3 and B4 indicate the indoor temperature change near Aac.

The AC instantaneous power is converted between high and low-frequency operation modes according to the temperature difference between the Aac measured temperature and the set temperature. In comparing the AC instant power of the A, B two rooms of 5/26, the Aac measurement temperature reached 27 °C after 15:00 in Room A. The AC instant power gradually increased after this time interval and remained relative. On the other hand, the AC instant power of Room B around 15:00 is un-changed. Before 15:00, the Aac records of Room A and Room B are almost consistent with the air-conditioning power consumption. After 15:00, the AC instant power of Room A rises. Due to the operation strategy of the curtains, the cumulative power consumption of the air conditioner is greater than that of Room B.

In comparing air-conditioning power consumption on 6/10, the window was changed to the south. The curtain operation strategy does not need to be implemented throughout the day. Therefore, Room A and Room B are similar in the Aac process and have the same air-conditioning power consumption. This means that buildings with windows in the south are less dependent on curtain shading, which means that a shading strategy does not need to be arranged to make air-conditioning operate at low power and save energy. This result can reference cost considerations and adjustments to operating strategies.

Due to similar weather conditions, this study conducted a three-day experiment with west- and south-facing buildings, and one day was selected respectively for the above discussion. Among them, the test dates for west-facing buildings are from May 24 to May 26; the actual dates for south-facing buildings are from June 9 to June 11. In the following paragraphs, this study further discusses energy consumption records for all experimental dates to demonstrate the reproducibility and reference of lighting control experiments. Figure 8 shows the total power consumption of Room A and Room B when the building faces west during the period of 5/24–5/26, including air-conditioning and indoor lighting. The time history of Room B is marked with the time and proportion of adjusting the shade ratio of curtains and turning on the lighting (satisfying 500 lux) time point.

According to the preset control group setting, Room A turns on indoor lighting all day, so the proportion of lighting power consumption is significantly higher than that of Room B. In addition, since the lighting power consumption shows linear accumulation and is not affected by the indoor and outdoor environment, the maximum use of daylight without affecting visual comfort can effectively save lighting power. This conclusion can be drawn from the accumulation of Room B approaching zero. The lighting power consumption is proven, saving about 3.3 kWh of power, which is equivalent to 90% of the power consumption of lighting.

Next, the energy-saving benefits of this study’s western window experiment were compared with other studies which explored similar energy-saving strategies. Compared with the study of Niraj et al. [19], the geographical location and boundary conditions of the experimental field and the limit threshold of the illumination of the working surface are the biggest differences. In the experiment of Niraj et al. with windows facing east, window materials similar to those in this study (U = 3.12 W/m²K) and roller blinds with high shading rate (RS1) were used, and about 50% lighting energy saving was obtained, namely 3 kWh per day. The lighting energy saving result of this study is 90%, which is an average of 3.3 kWh per day. There are significant differences in the percentage portion but quite close values in terms of absolute energy savings. In the study by Niraj et al., their continuously dimmable lighting system cannot be completely turned off, so the minimum brightness still consumes about 40% of the continuous power consumption. So far, lighting savings percentages have had almost identical results. This result is not surprising, because based on the power consumption of lighting in this study, it is clear that under the condition of maximum application of sunlight (Room B), the power consumption of artificial lighting (0.4 kWh) is almost negligible.

Compared with the total power consumption data, Room A used 16.7, 16.8, 16.8 kWh respectively during the three days; Room B used 12.1, 12.1, 12 kWh respectively, and the average saving of these three days was 4.7 kWh, which is equivalent to 28% of the total power consumption. This data expresses the operation strategy power-saving benefits for traditional offices. On the other hand, comparing air-conditioning power consumption data, Room A used 13, 13.3, 13.1 kWh respectively during the three days; Room B used 11.8, 11.8, 11.7 kWh, and the average saving of 1.36 kWh in these three days is equivalent to saving 10% of the air-conditioning power consumption.

In terms of the use of HVAC, since Niraj et al. [19] arranged the experiment from March to September, the energy consumption of HVAC is mainly cooling energy consumption. In their experiment, the energy saving of the air conditioner is about 3 kWh, which is about 25% of the energy saving of the air conditioner. Although Niraj and associates were able to save more energy, they did not provide detailed experimental information, so it is difficult to pinpoint the main factors leading to the difference. Thus, we can only surmise the main possible factors are illumination threshold setting and climate difference. Because of the more conservative illuminance limits (500–2000 lux), the Niraj et al. study needed to start adjusting the height of the roller blinds earlier. Our study used low-e windows (VT = 0.65) in the west-facing window experiment and needed to start adjusting the curtain height at 14:00. This adjustment may have resulted in relatively high cooling energy savings. On the other hand, the climate difference between Iowa and Taiwan may account for additional variations.

Simulations by Wankanapon et al. [29] concluded that dynamic roller blinds can save about 12–36% of the total energy consumption, which is comparable to our results of 28%. We contribute this similarity to same climate zone (subtropical) and season (summer season). The biggest difference lies in the control condition, which is based on the solar thermal radiation or illuminance threshold. This study records thermal radiation values so that they can be compared with experimental conditions for similar solar thermal radiation. In the west-facing part of our experiment, the total energy saving, lighting energy saving, and cooling energy saving in this study are 28%, 90%, and 10%, respectively; in work of Wankanapon et al., 18.5%, 47%, and 13–16%, respectively. In comparison, the difference in lighting energy saving is mainly attributable to different control group benchmarks. The difference in cooling energy savings is relatively small, but the relatively high energy savings may be attributable from experimental conditions adjusted for solar heat radiation. Comparing the same solar thermal radiation threshold with the thermal radiation and illuminance history of this study, it can be found that the time for curtain adjustment needs to be advanced. From this condition it should be possible to estimate the difference in cooling energy savings. Finally, differences in total energy savings are difficult to assess because the total energy savings in this study do not include heating energy consumption and therefore cannot be compared.

The results at this stage echo the simulations of Tzempelikos et al. [26] Maximizing the use of daylight has advantages in terms of lighting energy savings, but is prone to overheating problems in summer climates, thereby increasing the cooling load.

The results of this study show that external vertical radiation has a large impact on cooling energy consumption. According to the heat flux results presented, it has a significant heat flux blocking effect and effectively reduces cooling energy consumption. This finding echoes that published by Niraj et al. [19].

Figure 9 shows the total power consumption of Room A and Room B when the building faces south during 6/9–6/11. Because curtains do not cover Room B throughout the day, the difference between the total power consumption of Room A and Room B is the amount of lighting power consumption. In this experiment, the air-conditioning power consumption of Room A and Room B are 10.4, 10.3, 10.4 kWh, and 10.4, 10.2, 10.4 kWh, respectively. Among them, the lighting conditions of 6/9 are delayed, but it is known from the results that it does not affect the air conditioning load.

The results of the south-facing part of the experiment demonstrate that the total energy saving, lighting saving, and cooling saving in this study are 26%, 100%, and 0.5% respectively; compared to the research findings of Wankanapon et al. [29], 16.2%, 40%, and 13–16%. The baseline conditions of the control groups are different, making it difficult to analyze and compare similar studies. Differences in cooling energy is likely because the illuminance threshold of this study is higher, and the results in the southside did not require curtain adjustment, so no energy saving occurred. Finally, the total energy savings cannot be compared because the study did not use heating.

Figure 10 shows the power consumption history of the air conditioners in Room A and Room B for all experiments. The upper and lower halves of Figure 10 show the experimental results facing west and south, respectively.

Further, superimposing the aforementioned air-conditioning power consumption results, it is evident that before the shading rate is further changed, the air-conditioning power consumption history of Room A and Room B almost wholly overlap. Only after the energy-saving strategy is started is there a clear difference, which can be proved in that the difference in air-conditioning power consumption is dominated by energy-saving strategies.

The results of this experiment show that the building faces west. In the afternoon, there will be strong sunlight directly into the room, increasing the indoor heat source, worsening indoor comfort, and increasing the air conditioning load and power consumption. If the high shading factor is used to shield the room from direct sunlight, it can maintain indoor comfort and reduce the air-conditioning load and power consumption. The building is facing south, and since there is no period when strong sunlight directly shines into the room, the shading with high shading coefficient is not activated.

Figure 11 and Figure 12 show the PMV values calculated by the CBE Thermal Comfort tool.

Figure 11 shows that before the curtain control strategy starts to adjust the curtains (15:00), the PMV values of the two rooms are quite close. The same indoor and outdoor conditions lead to similar indoor thermal comfort. This case illustrates that the conditions in the two compartments are similar enough that the differences in results are small. After the curtain control strategy was adjusted (15:00 and 17:10), the PMV values of the two rooms began to diverge significantly. After the curtains are adjusted in height, the solar radiation heat in room B is blocked by the curtains near the outdoor side, thereby improving the temperature on the indoor side, and thus improve the overall indoor thermal comfort. According to the classification of PMV, PMV = 0 is Neutral, PMV = 1 is slightly warm, and PMV = 2 is warm. From the results of PMV, it can be found that the result of room B has a value above 1 only when the air conditioner is just turned on at 8:00 in the morning, and the subsequent indoor thermal comfort is controlled by the air conditioner and curtains and is lower than 1. This is because the temperature has not dropped to the target temperature, since the air conditioner is just turned on in the morning. This phenomenon is the same as the initial temperature result in Figure 5. At the same time, the peak power consumption of the air conditioner is also increased due to the high indoor temperature to meet the demand, as shown in Figure 7.

On the other hand, the curtains in the control group were not adjusted, so there was still a period of PMV > 1 in the afternoon, showing a slightly warm result. According to CBE Thermal Comfort, the range of PMV that the human body feels comfortable is between −0.5 and 0.5. From Figure 11, it can be found that despite being affected by the curtain control, room B still causes thermal discomfort to the human body during the afternoon period (15:00–17:00). This may be due to the fact that the curtains have not been completely closed during the period from 15:00 to 17:00, and some heat radiation is still transmitted through the windows into the room. Therefore, it can be inferred that the thermal discomfort of the window-side seats is higher than the currently calculated PMV result. After 17:00, the curtains are completely closed, and the heat energy entering the room is blocked by the curtains to a greater extent on the window side, and the outdoor temperature gradually drops as the sun goes down.

Room A and Room B in the south-facing experiment (Figure 12) were operated under exactly the same indoor and outdoor conditions, so there was no significant difference in the results.

4. Conclusions

This study adapts an indoor shading strategy to reduce heat flux and illuminance from sunlight. In terms of heat flux and illuminance, measurements in room B showed a significant drop after two levels of shading. In a west-facing experiment, a roller blinds shading strategy enabled indoor lighting to only turn on after 17:00, saving 90% of lighting power consumption. On the other hand, in south-facing experiments, we obtained similar lighting savings. The results show that making full use of sunlight in summer can bring huge energy saving potential for artificial lighting. In terms of temperature, the temperature histories of Room A and Room B are significantly different, especially on the window side. The maximum temperature difference of 4.3 °C occurs at 17:30 due to the roll-blind shading strategy reducing the effect of heat introduced into the room.

In terms of air-conditioning power consumption, lower indoor temperature leads to lower air-conditioning load, and the roll-blind shading strategy saves 1.36 kWh of electricity for air-conditioning on average, which is equivalent to 10% of the energy consumption of air-conditioning. And this energy saving effect is concentrated in the peak period of electricity consumption (15:00–17:00). In the south-facing experiment, the results of Room A and Room B are mostly consistent. The contrast between west-facing experiments and south-facing experiments provides a reference for energy-saving strategies and device updates. It is important to note that while the energy-saving strategy of this study utilizes sunlight to a greater extent which may contribute to a minor air-conditioning energy saving, visual comfort may be neglected.

Finally, this study also provides the thermal comfort index of PMV to illustrate the actual impact of roller blind shading on indoor human comfort. In the experiment with west-facing windows, there was a significant difference in thermal comfort between Room A and Room B in the afternoon. Additionally, Room B benefited from the influence of the height adjustment of the roller blinds, as indicated by an PMV of below 0.8 throughout the whole process, keeping the indoor comfort between Slightly Warm and Between Neutral. This is the result of maximizing lighting energy saving. If there is a need to increase cooling energy saving and further improve thermal comfort, the illuminance threshold must also be optimized, because the benefits of cooling energy saving and lighting energy saving are trade-offs. Since the windows are south-facing, they are not affected by direct sunlight, and the demand for roller blinds is low. The PMV results can also be kept below 0.5, which is comfortable for the human body, unless the air-conditioning system is activated.

Multi-directional full-scale building sunshade energy-saving experiments are still very rare. This study provides a real-time rotating house experiment with fully symmetrical conditions. Our study provides comprehensive experimental results, as well as confirms previous simulation and experimental results in literature with new interpretation. Roller shading strategies have higher operational demands and actual energy savings in buildings with west-facing windows. Buildings with south-facing windows may have greater potential for daylight utilization. In addition, this study also discusses the thermal comfort index, PMV. At present, there are few related studies that discuss energy consumption history and thermal comfort concurrently. However, while emphasizing energy-saving benefits, it is also necessary to establish a long-term solution for sustainable development on the premise of maintaining the thermal comfort of the human body.