**1. Introduction**

It is well known that energy consumed by buildings accounts for more than 40% all global energy consumption, and 60% of that goes to ensuring indoor thermal comfort for occupants [1,2]. Air conditioning systems, which play a crucial role in creating a thermally

**Citation:** Zhu, X.; Liu, J.; Zhu, X.; Wang, X.; Du, Y.; Miao, J. Experimental Study on Operating Characteristic of a Combined Radiant Floor and Fan Coil Cooling System in a High Humidity Environment. *Buildings* **2022**, *12*, 499. https:// doi.org/10.3390/buildings12040499

Academic Editor: Rafik Belarbi

Received: 11 March 2022 Accepted: 15 April 2022 Published: 17 April 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/).

101

comfortable living environment, are used extensively since most people spend 80–90% of their time inside buildings. Because air conditioning systems make up such a large portion of total energy consumption, researchers are paying more attention to their optimization [3–6]. The goal is to reduce building energy consumption by optimizing air conditioning systems without reducing indoor thermal comfort. To achieve this, efficient operation strategies based on building envelope characteristics and loads and meteorological conditions are required. Over the years, many studies have investigated energy consumption by radiant cooling systems and found that, compared to conventional convection air conditioning systems, radiant cooling systems can reduce energy consumption by about 40% [7–11].

Radiant floor cooling (RFC) has attracted widespread attention because of its good thermal comfort, low energy consumption, building space saving, and low operating noise. Researchers have conducted many practical application studies and field measurement studies on radiant floor air conditioning systems [12–15]. Sourbron et al. compared the response times of RFC and convection air conditioning systems based on the thermal inertia of the envelope and proposed an operational strategy utilizing intermittent operation [14,16–18]. Romaní et al. proposed a water supply temperature operation strategy where the water supply temperature is continuously adjusted according to the outdoor temperature, but this strategy was more complicated than others. A water supply temperature operating strategy is well adapted to maintaining stable operation while accounting for uncertainties resulting from increased indoor heat loads or solar radiant heat gain throughout the day. However, RFCs are not able to immediately deal with changes in hot and cold loads due to the high thermal inertia of buildings, so other energy handling systems must be set up to buffer rapid change [19–23]. In light of this, Mikeska et al. demonstrated the advantages of using the RFC system at night or during off-peak hours, thus effectively utilizing low temperature air from natural ventilation and off-peak electricity prices. In this intermittent operation, when the system is stopped, heat accumulates in the air near the radiant floor surface, and when the system is started again, a strong temperature gradient is created near the radiant floor surface. This strategy increases the heat capacity of the radiant floor system and enables it to absorb heat for a short period of time while reducing energy consumption [19,24–26].

A summary of previous applications and studies on radiant cooling are listed in Table 1. Zarrella et al. studied radiant floor cooling with different ventilation systems using simulations, their results showed that RFC systems need to be combined with ventilation systems to achieve efficient and energy-saving operation [27,28]. Zhao et al. proposed a simple method for predicting the performance of RFC systems with solar radiation in steady states or large open spaces using a combination of experiments and simulations, which considered the role of solar radiation in RFC systems [29]. Jin et al. showed that the risk of condensation on a radiant surface is high during the start-up phase and that an air convector can effectively reduce the risk of condensation on the radiant surface [30,31]. Srivastava et al. studied a roof radiation system combined with different ventilation systems using both simulations and experiments, and their results showed that the radiant cooling system was able to maintain efficient operation in different climate zones [32,33]. Feng et al. showed that the cold transient rates of a radiant system are faster than those of an air system during heat gain by an experimental building [34]. Fernandez-H et al. studied a radiant floor with fan coil/displacement ventilation systems using experimental methods, and their results showed that radiant cooling systems not only improve energy efficiency, but also the system response time [35,36]. Joe et al. showed that different operating strategies can achieve different levels of energy savings while ensuring indoor thermal comfort [37,38].


**Table 1.** Summary of the existing research on radiant cooling.


**Table 1.** *Cont.*

Convection air conditioning systems have short response times and are capable of rapid temperature adjustment compared to RFC systems. Here, to utilize the advantages of both RFC and traditional convection air conditioning systems, a combined radiant floor and fan coil cooling (RFCAFC) system was developed. As with conventional air conditioning systems, the adoption of suitable operation strategies is central to the energy consumption reduction strategy of the RFCAFC system [39]. Indeed, by studying operation strategies using combinations of radiant floors with different ventilation systems, Atienza et al. found that integrating an air conditioning system not only ensured indoor thermal comfort, but also resulted in a greater potential for energy savings. In addition, since weather conditions and indoor heat gain are dynamic and have a large impact on the performance of radiant cooling systems, an effective operation strategy is especially important [40,41].

Currently, RFC systems are not widely used in residential buildings and small offices without central air conditioning because of the high risk of condensation, slow response time, difficulty in controlling multiple zones, and insufficient floor cooling capacity due to intermittent use within high humidity environments [41–47]. Moreover, the existing field experiments on the operation strategies of combined radiant floor and fan coil cooling systems have been based on manual adjustments and have not implemented automated operation strategies. This means that RFC systems have been difficult to integrate into the automated systems of buildings to achieve efficient smart building functions. Based on the above problems, we built a RFCAFC automatic operation system and analyzed the impacts of different operating strategies on building energy consumption and thermal comfort in a high humidity environment. The RFCAFC automatic operation system was installed in Jinan, China. During the experiment, relevant parameters of the combined system were monitored and the operating characteristics and cooling capacity of the combined system under different outdoor meteorological conditions were analyzed. This study provides a strong reference and basis for further study of the operating characteristics of combined systems under high humidity environments and for further promotion of the combined system in small scale residential applications.

#### **2. Experimental Description**

#### *2.1. Introduction of the Laboratory*

The experimental laboratory was located in Jinan, China, which is in a hot summer and cold winter climate zone. The enclosed structure adopted aerated concrete integrated composite wall panels, which have low thermal conductivity and good thermal stability. The laboratory was a single building with length, width, and height of 4.00 m, 2.80 m, and 3.00 m, respectively, and a total construction area of 11.2 m2. The exterior walls were all 250 mm aerated concrete integrated composite wall panels, and the heat transfer coefficient of the exterior walls was 0.536 W/(m2·K). The south facing exterior wall had an aluminum framed double-glazed exterior window with an area of 2.4 m<sup>2</sup> and a heat transfer coefficient of 2.40 W/(m2·K). The floor plan of the laboratory and building is shown in Figure 1.

The radiant floor coil adopted the double circuit dry buried pipe arrangement. The floor radiant coil pipe diameter was 12 mm, the pipe wall thickness was 2 mm, the coil spacing was 60 mm, and the pipe was PE-RT pipe. The radiant floor surface decoration material for the tile surface and radiant floor structure is shown in Figure 2. Its basic structure from the bottom to top was made of the following parts: floor layer, insulation layer, buried pipe layer, soil layer, screed layer and surface layer, where the insulation layer and buried pipe layer for integrated processing were combined into a radiation cooling module. The radiant floor thermal parameters are listed in Table 2. Indoor temperature sensors, humidity sensors, and anemometers were installed to measure indoor temperature, humidity, and vertical air temperature difference. The measurement points were arranged as shown in Figure 3. PT-100 thermal couplings were installed on the interior walls to measure the temperature of each wall's surface.

**Figure 2.** Radiant floor structure.

**Table 2.** Thermal parameters of the radiant floor.


**Figure 3.** Location of the indoor measuring points.

#### *2.2. Introduction of the Combined Cooling System*

The operating principle of the RFCAFC system is shown in Figure 4. The RFCAFC system consisted of a cold source, a constant temperature water tank, a primary distribution system, a secondary distribution system (mixer pump), a safety component, and an air conditioning terminal.

**Figure 4.** Schematic of the cooling system.

The combined cooling system used a single-phase inverter air source heat pump (KXD80/60EA-V) with a rated cooling capacity of 7.2 KW, an IPLV of 2.81, and a fan coil (FP-51) with a rated cooling capacity of 2.7 kW. To measure the cooling capacity of the combined cooling system, an electromagnetic flow meter (LDG-MIK DN25) and a water temperature sensor (MIKwrn-130) were installed on the supply and return pipes of the fan coil. To measure the energy consumption of the combined cooling system, an intelligent socket (KTBL03LM) was installed on the system power transmission and distribution system to record power consumption. The experimental equipment of the combined cooling system is shown in Figure 5 and the detailed parameters of the experimental equipment are shown in Table 3.

**Figure 5.** Illustrated schematic of the cooling system, (**a**) temperature controller of fan coil, (**b**) fan coil, (**c**) flow meter, (**d**) thermometer, (**e**) radiant floor coil, (**f**) temperature controller of radiant floor, (**g**) mixed water center, (**h**) constant temperature buffer tank, (**i**) air source heat pump, (**j**) controller of air source heat pump, (**k**) supply water pump.


**Table 3.** Summary of the experimental instrument operating parameters.

In summer, chilled water produced by the air source heat pump first flows through the constant temperature buffer tank where it is brought to the required pressure by the circulating pump to overcome the resistance of the pipeline. The chilled water is sent to the air conditioner, passing through the temperature measuring instrument on the way, through the circulating pipeline made of polyvinyl chloride (PVC) plastic pipes.

As a radiant floor is more prone to condensation in high a humidity environment, the combined cooling system adopted the operation logic that the floor radiation water supply temperature should be higher than the indoor air dew temperature to prevent condensation. The specific implementation was as follows: The radiant floor water supply temperature always followed the indoor air dew temperature. An indoor dew thermostat collects the indoor air temperature and humidity and transmits the data to the central controller, which calculates the dew temperature based on the indoor temperature and humidity and the built-in enthalpy and humidity graph algorithm, and then adjusts the mixing valve at the mixing pump station to adjust the water temperature so it is higher than the dew point temperature by 2 ◦C, thus realizing the control logic flow, as shown in Figure 6.

The operation of the combined cooling system was realized primarily by setting the desired indoor temperature. In this system, when the indoor temperature is lower than the set value, the air source heat pump will run at a low frequency or be stopped, and when the indoor temperature is higher than the set value, the air source heat pump starts to run again. The low temperature chilled water produced during this operation was not only supplied to the radiant floor and fan coil air conditioning, but also stored in a thermostatic buffer tank. When the system needs low temperature chilled water again, it can be taken directly from the constant temperature buffer tank. This avoids frequently starting and stopping the air source heat pump, which prolongs the heat pump's service life and saves energy.

During the initial operating stage of the combined cooling system, due to the high humidity of the indoor environment, the radiant floor and fan coil are used together to provide cooling. When the fan coil stops running, the radiant floor provides all the cooling for the room.

**Figure 6.** Control logic of the cooling system.

#### *2.3. Uncertainty Analysis*

In order to quantify the accuracy of measurements, an uncertainty analysis of the measured data was carried out based on the general law of uncertainty propagation [48]. The uncertainty was obtained using the following equations [49] and the results are listed in Table 4.

$$
\mu\_i = \left(\frac{w\_i^2}{3}\right)^{0.5} \tag{1}
$$

$$\mathcal{U}\_R = \left(\sum\_{i=1}^n \mu\_i^2\right)^{0.5} = \left(\frac{1}{3}\sum\_{i}^n w\_i^2\right)^{0.5} \tag{2}$$

where *ui* is the standard uncertainty of the measured parameter; *wi* is the accuracy of the device; and *UR* is the standard uncertainty of the parameter determined by other measured parameters.



#### **3. Experimental Cases and Evaluation Indices**

#### *3.1. Experimental Cases*

These experiments were conducted in August and September, which are representative of the summer cooling season in Jinan, China. The outdoor temperature varied from 26 ◦C to 40 ◦C and the outdoor humidity varied from 35% to 94% during the test period, as shown in Figure 7. The outdoor weather was classified as overcast, sunny, or cloudy, which could be directly related to changes in outdoor total horizontal solar radiation. These climate characteristics provide conditions in which intermittent operation of the combined cooling system was required. During the experiment, the room temperature was set at 25 ◦C and the fan coil was set at the lowest operating air speed. During the operating time, two heated dummies (123.6 W and 137.5 W) were used to simulate indoor heat sources that approximated human bodies.

According to the temperature, humidity, and outdoor meteorological conditions, the environmental conditions were classified into three groups, as shown in Table 5. When the outdoor temperature and humidity ranges were 21–29 ◦C and 74–94%, respectively, the conditions were referred to as low temperature and high humidity (LH). When the outdoor temperature and humidity ranges were 24–38 ◦C and 42–90%, respectively, the weather conditions were referred to as high temperature and low humidity (HL). When the outdoor temperature and humidity ranges were 22–33 ◦C and 50–88%, respectively, the weather conditions were referred to as medium temperature and medium humidity (MM).

**Table 5.** Outdoor weather conditions classification table.


**Figure 7.** *Cont*.

**Figure 7.** Meteorological conditions during the experiment, (**a**) outdoor air temperature, (**b**) outdoor air relative humidity, and (**c**) solar radiation.

Different outdoor weather conditions required different start-stop strategies in the RFCAFC system. The start-stop conditions of the combined cooling system during the experiment are shown in Table 6.

**Table 6.** Summary of the combined cooling system start/stop strategy under different outdoor weather conditions.

#### *3.2. Evaluation Indices*

#### 3.2.1. Radiant Floor Surface Uniformity Temperature Coefficient

The average temperature of the surface of the radiant floor directly affects the heat exchange at the surface of radiant floor. This study used the *S* value to express the cooling capacity of the radiant floor. The closer the average temperature of the radiant floor surface to the lowest temperature of the radiant floor surface, the closer *S* is to 1. The better the *S*, the greater the cooling capacity of the radiant floor surface supply. The *S* calculation method is shown in Equation (3).

$$S = \frac{T\_{\text{s,max}} - T\_{\text{s,min}}}{T\_{\text{g}} - T\_{\text{h}}} \tag{3}$$

where *Ts,max* is the maximum radiant floor surface temperature, ◦C; *Ts,min* is the minimum radiant floor surface temperature, ◦C; *Tg* is the radiant floor water supply temperature, ◦C; and *Th* is the radiant floor return water temperature, ◦C.

#### 3.2.2. Time Constant

The time constant (*τ*) is used as a time scale to measure the time required from the start of the combined cooling system for the radiant floor to reach a relatively stable state. *τ*<sup>63</sup> is defined as the time required for the temperature to change from the initial value *Ts* to *Te* (approximately 63.2% of the total desired change). *τ*<sup>63</sup> reflects the rapidity of the radiant floor surface temperature change. *τ*<sup>95</sup> is defined as the time required for the radiant floor surface temperature to reach approximately 95% of the desired change. *τ*<sup>95</sup> is a measure of the time required for the surface temperature of the radiant floor to reach a relatively stable state from the initial state. According to the physical meaning of *τ*, the *τ* calculation method is shown in Equations (4) and (5).

$$
\pi\_{63} = T\_{im} - (T\_{im} - T\_{rs}) \times 63.2\% \tag{4}
$$

$$
\pi\_{\mathfrak{F}5} = T\_{im} - (T\_{im} - T\_{rs}) \times \mathfrak{F}5.0\% \tag{5}
$$

where *τ*<sup>63</sup> is the time required for the radiant floor surface temperature to reach 63.2% of the total change, h; *τ*<sup>95</sup> is the time required for the radiant floor surface temperature to reach approximately 95% of the total desired change, h; *Tim* is the initial radiant floor surface temperature, ◦C; and *Trs* is the desired radiant floor surface temperature, ◦C.

#### 3.2.3. Room Temperature Fluctuation

The sample standard deviation (*S*) was introduced in this experiment to evaluate the stability of the indoor thermal environment and characterize the degree of dispersion of the indoor temperature. Larger sample standard deviations indicate a greater degree of fluctuation in indoor temperature during the test period, which would indicate poor stability in the indoor thermal environment. The *s* calculation method is shown in Equation (6).

$$S = \sqrt{\frac{\sum\_{i=1}^{11} \left(T\_i - T\_{\text{age}}\right)^2}{n}} \tag{6}$$

where *S* is the standard deviation of indoor temperature; *Ti* is the indoor temperature at the *i th* hour after the start of the combined system, ◦C; and *Tage* is the average indoor temperature during the test period, ◦C.

#### 3.2.4. Thermal Comfort

The international standard ISO7730 uses the predicted mean vote (PMV) and predicted percentage dissatisfaction (PPD) to describe and assess the comfort of an indoor thermal environment. PMV represents the sensations experience by the majority of people in the same environment and can be used to evaluate the comfort or discomfort of a thermal environment. However, as a mean value, the PMV metric ignores individual variability, which means that it does not represent the thermal sensations of all people within a thermal environment. Therefore, researchers have proposed the PPD index which uses probability analysis to indicate the percentage of the population dissatisfied with a thermal environment. The recommended PMV value range in the international standard ISO7730 is −0.5 to +0.5, and the recommended value of PPD is ≤10%. The PMV calculation method is shown in Equation (7) and the PMV quantitative index is shown in Table 7.

$$\begin{aligned} \text{PMV} &= \left(0.303 \text{e}^{-0.058M} + 0.0275\right) \times \left\{M - W - 3.05[5.733 - 0.007(M - W) - P\_d] - 0.42(M - W - 58.2) - 0.073(M - W) - 0.007(M - W)\right\} \\ &- 0.0173M(5.867 - P\_a) - 0.0014M(34 - t\_a) - 3.96 \times 10^{-8} f\_{cl} \left[ \left(t\_{cl} + 273\right)^4 - \left(t\_{\text{grr}} + 273\right)^4 \right] - f\_{cl}h\_c(t\_{cl} - t\_a) \end{aligned} \tag{7}$$

where *M* is the human energy metabolic rate, W/m2; *W* is the mechanical work done by the human body, W/m2; *Pa* is the partial pressure of water vapor around the human body, kPa; *f cl* is the garment area coefficient; *tcl* is the temperature of the outer surface of the garment, ◦C; *tage* is the average radiation temperature, ◦C; *hr* is the convective heat transfer coefficient, W/(m2·K); and *ta* is the air temperature around the human body, ◦C.

**Table 7.** Predicted mean vote index.


3.2.5. System Cooling Capacity

In principle, when a radiant floor is used for summer cooling, it needs to be combined with an air supply system. However, most existing studies on the cooling capacity of RFC systems were based on experimental benches without air supply systems, which does not consider the influence of different air supply forms on the cooling capacity of a radiant floor, and also ignores the influence of air supply on the convective heat exchange of the radiant floor. The introduction of the air supply system not only affects the convective heat exchange between the surface of the radiant floor and the room air, but also changes the thermal environment condition of the room, which directly influences the RFC capacity.

The cooling capacity and operation status of the combined cooling system was evaluated by calculating the heat load carried at different air conditioning terminals. The operation of the combined cooling system at each stage of the day was analyzed based on the heat load capacities of different system components. The cooling capacity of the RFCAFC system can be calculated according to Equations (8) and (9).

$$Q\_R = c\_p m \left( T\_\text{h} - T\_\text{g} \right) \tag{8}$$

$$Q\_F = G\_\pi (h\_2 - h\_1) \tag{9}$$

where *QR* is the RFC capacity, W; *QF* is the fan coil cooling capacity, W; c*p*is the specific heat at constant pressure, J/(kg· ◦C); *m* is the mass flow rate per unit area of circulating duct, kg/s; *Gn* is the fan coil air supply volume, kg/s; *h*<sup>2</sup> is the return air outlet enthalpy, kJ/(kg·K); and *h*<sup>1</sup> is the enthalpy at the supply air outlet, kJ/(kg·K).

#### 3.2.6. System Operation Status and Energy Consumption

The *EERh*, a measure of cooling capacity per unit power, was used to evaluate the energy efficiency of the combined cooling system during operation. The greater the *EERh*, the better the energy conversion efficiency of the RFCAFC system. The *EERh* of the combined system is calculated according to Equation (10).

$$EER\_h = \frac{Q}{3.6 \times 10^6 E\_h} \tag{10}$$

where *EERh* is the hourly cooling performance coefficient; *Q* is the cooling capacity, W; and *Eh* is the system energy consumption, kWh.

#### **4. Results**

Experimental data under similar outdoor meteorological conditions were grouped to assess the different operating strategies for the combined radiant floor and fan coil cooling system in high humidity environments. The variations in indoor temperature (*Tin*), radiant floor surface temperature (*Tf*), and PMV-PPD values over time differed among different operation strategies and different outdoor meteorological conditions. The operational status and energy consumption of the combined cooling system under different operation strategies were analyzed using the *EERh* and power consumption of the system. The analysis verified that the RFCAFC system was effective and energy efficient for different outdoor meteorological conditions in a high humidity environment. These results provide a strong theoretical basis for the promotion and application of the combined RFCAFC system in high humidity environments in Jinan, China.

#### *4.1. Indoor Temperature and Humidity*

Figures 8 and 9 show the variation of indoor humidity (*ϕ*), outdoor temperature (*TO*), indoor temperature (*Tin*), and dew point temperature (*Td*) over time using different operating strategies. It can be seen from Figure 8 that *ϕ* varied greatly during the test period. The *ϕ* tended to decrease significantly 0–3 h after turning on the combined cooling system, but after that the *ϕ* gradually reached a relatively stable state. This was because higher outdoor humidity resulted in higher indoor humidity as well, which affected the operation of the system. In general, the combined cooling system had a significant dehumidification effect in the high humidity indoor environment.

**Figure 8.** *ϕ* during the experimental period.

**Figure 9.** Temperature field showing (**a**) *Tin* change on the LH, (**b**) *Tin* change on the HL, and (**c**) *Tin* change on the MM.

As can be seen in Figure 9, the outdoor temperature fluctuated between 23 ◦C and 40 ◦C during the test period in Jinan, China. The average outdoor temperatures were 27.4 ◦C, 33.9 ◦C, and 28.7 ◦C during operating hours (8:00–18:00). The room temperature fluctuated between 25.5 ◦C and 27.5 ◦C during operation after the combined system reached a relative steady state. The average indoor temperatures were 26.1 ◦C, 26.6 ◦C, and 25.4 ◦C, respectively, during working hours. These values met the indoor temperature requirements for occupants. The standard deviations of *Tr* were low, i.e., 0.45 ◦C, 1.04 ◦C, and 0.58 ◦C, and the maximum standard deviation of *Tin* was only 1.04 ◦C. These values indicated that *Tin* was relatively stable and the operation strategy was effective in regulating *Tin* under different outdoor meteorological conditions.

Figure 9 shows that *TO* was significantly lower than the indoor temperature during the test period 7:00–9:00. Therefore, the room temperature appeared to be reduced by natural cooling during that time period. By comparing the operating strategies for different outdoor meteorological conditions, it was found that the indoor temperature (*Tin*) under the operation strategy for HL showed a decreasing trend during the test period, while the *Tin* under the LH and MM operating strategies exhibited oscillating changes during the test period. The reason for this phenomenon was that there was a larger heat load entering the room through the envelope structure with continuously high outdoor temperatures. At the same time, the combined cooling system operated intermittently and the low outdoor temperature environment at night further improved the stability of the indoor thermal environment by reducing the internal surface temperature of the building envelope and offsetting part of the heat load accumulated by the building during the day. Therefore, to further improve the stability of the indoor thermal environment, intermittent operation of the combined cooling system at night can be considered. From the perspective of energy saving, natural cooling can effectively be used from 7:00 to 9:00.

#### *4.2. Vertical Air Temperature Difference*

Figure 10 shows the *Tin* change curves at different heights from the ground over time. It can be seen that the indoor vertical temperature fluctuated between 24.5 ◦C and 33.1 ◦C and at 1100 mm above the ground the temperature fluctuated between 23.4 ◦C and 26.3 ◦C under different operating strategies. Under the same operational strategy, the trends of *Tin* change were similar and relatively uniform among different heights. Under the LH operating strategy, the maximum temperature difference between different locations in the room was 0.3 ◦C at the same height, while under the LH and MM operating strategies the maximum temperature differences between locations in the room at the same height were 0.7 ◦C and 0.5 ◦C, respectively. The figure also shows that the room was colder at the lower points and hotter at higher points, and the temperature gradient was slightly more pronounced above 1.5 m. Under the LH, HL, and MM operating strategies, the maximum temperature differences at different heights in the room were about 1.7 ◦C, 1.6 ◦C, and 1.8 ◦C, respectively.

According to the ASHRAE552017 standard, the maximum temperature gradient should not exceed 3 ◦C in an area with major human activities. Therefore, the operating strategies used under all different outdoor meteorological conditions met the indoor temperature gradient requirements. Due to the influence of solar radiation, the indoor local maximum temperature occurred at the height of 1.1–1.5 m from 13:00 to 15:00, which could cause thermal discomfort at specific locations. Therefore, shading measures should be considered in the building construction process to reduce thermal discomfort caused by the local high temperature phenomenon.

**Figure 10.** *Cont*.

**Figure 10.** Vertical temperature differences under the (**a**) LH, (**b**) HL, and (**c**) MM operating strategies.

#### *4.3. Radiant Floor Surface Temperature*

The changes in radiation floor surface temperature under different operating strategies are shown in Figure 11. The black curve indicates the change trend of *S* and the red line indicates the fluctuation range of *S*. Under the HL operating strategy, *S* remained basically stable at about 1.0, which was a desirable level. Under the LH and MM operating strategies, the *S* was low, mostly within the 0.7–1.0 range, which can be considered the normal range. However, under the MM operating strategy, the *s* appeared to fluctuate more. So, under the MM operating strategy, the radiation floor had worse stability, but still remained within the normal range. This may have been because the indoor humidity was too high under this environmental condition. From the following Figure 11, it can be found that *S* was more stable when the radiant floor operated continuously than when it operated intermittently. Therefore, when the combined cooling system is artificially regulated, the radiant floor should be set to a constant open state to reduce fluctuation and improve comfort.

**Figure 11.** *S* range for the (**a**) LH, (**b**) HL, and (**c**) MM operating strategies.
