Performance Evaluation of Thermal Bridge Reduction Method for Balcony in Apartment Buildings

: Most apartment buildings in South Korea use internal insulation systems to reduce building energy demand. However, thermal bridges such as balcony slabs in apartment buildings still lead to significant heat loss in winter, because the internal insulation system is not continuous in the balcony slab structure, and floor heating systems are commonly used in residential buildings. Therefore, this study investigates two types of thermal break elements, namely thermal break (TB) and thermal break-fiber glass reinforced polymer (TB-GFRP), to improve the thermal resistance of a balcony thermal bridge. To understand the effects of balcony thermal bridges with and without thermal break elements, the linear thermal transmittances of different balcony thermal bridges were analyzed using Physibel simulations. Then, the heating demand of a model apartment under varying thermal bridge conditions was evaluated using TRNSYS simulations. To understand the effect of insulation systems on heat loss through a balcony thermal bridge, apartments with internal and external insulation systems were studied. Whether the apartment was heating was also considered in the thermal transmittance analysis. Thus, the linear thermal transmittance of the thermal bridges with thermal break elements was reduced by more than 60%, and the heating energy demands were reduced by more than 8%.


Introduction
Persistent environmental problems, such as the lack of fossil fuels and global warming, are still major challenges for humans. Therefore, policies for energy saving and emission reduction have been adopted worldwide. The EU has a target to improve energy efficiency to 32.5% by 2030 to reduce energy consumption [1]. South Korea plans to reduce CO2 emissions by 37% by 2030 compared with business-as-usual (BAU) [2]. Thus, reducing energy consumption is necessary to improve environmental problems. Among the total energy consumption, building energy consumption contributes over 30% [3], and 70.4% of the building energy is consumed by apartment buildings [4]. Apartment buildings have less internal load compared to other buildings, hence they are envelope-loaddominated buildings. Therefore, heating load accounts for a large part of the total residential load, and more energy is consumed for heating. To reduce apartment building energy consumption, the performance of building equipment systems is improved, and innovative building components are used. In addition, a part of the heating load comes from the heat loss through thermal bridges of the building envelope, which in winter can be up to 30% of the total apartment building energy consumption [5,6]. Therefore, apartment buildings have a large potential for reducing energy consumption by improving the effects of thermal bridges. Therefore, increased attention is being paid to construction details where thermal bridges are present. building insulation standards have increased significantly, and therefore the heat loss will be more concentrated at the thermal bridges. For this reason, different methods for reducing heat loss by balcony slab thermal bridges have been investigated in many studies. In addition to improving the insulation system, installing thermal break (TB) in the balcony thermal bridge is an effective way to prevent heat loss.
A TB is an element of low thermal conductivity placed in thermal bridges to avoid excessive heat loss between the indoor and outdoor environments. It is designed so that it can mitigate the direct heat flow through the balcony slabs while providing sufficient strength to resist the shear force at the wall-slab junctions. In the study conducted by Santos et al. [32], it was found that the inner or outer TB strips in the light steel-framed building components could effectively improve the thermal resistance, and double TB strips could lead to more enhanced thermal performance. However, a balcony slab with a conventional TB still has a large heat transfer because of the stainless-steel bars in the concrete slab. In general, a TB product for balcony slabs consists mainly of an insulation block and stainless-steel bars. To prevent heat transfer from the thermal bridge more effectively, a noncorrosive material for reinforcement bars can be an alternative to TB components. Thus, a glass fiber-reinforced polymer (GFRP), which has a relatively low cost to be implemented, has been used as an alternative material for producing reinforcing bars [33,34]. Wakili et al. [35] introduced a TB with GFRP (TB-GFRP) for the concrete balcony slab, and analyzed the thermal performance of the proposed balcony slab. This study showed that the TB-GFRP element reduced the additional heat loss compared to the balcony slab without TB-GFPR. Goulouti et al. [36] also showed that a building using TB-GFRP can reduce the heating load by 41%. Even if GFRP can effectively reduce the heat transmittance, the application of GFRP is still quite new. In addition, the performance of TB-GFRP was not evaluated for the residential building with radiant floor heating systems, wherein the heat transfer at wall-slab junctions would be higher than that of the buildings with air heating systems.
The objective of this study was to evaluate different methods for improving the balcony slab thermal bridge. To do this, TB and TB-GFRP elements were selected as TBs, and effects of these elements on balcony thermal bridges were analyzed with the heat transfer simulations. Because internal and external insulation systems applied in apartments also have different effects on the thermal bridge, the analytical conditions were divided into two types: apartments with internal insulation systems and external insulation systems. In addition, to understand the effect of the radiant floor heating system on the balcony thermal bridge, the heating conditions in this study were also divided into an air heating system and a radiant floor heating system. The impact on the thermal bridge reduction was evaluated by analyzing the linear thermal transmittances and heating energy demands. In general, the linear thermal transmittances of thermal bridges were analyzed using a 2D heat transfer models [37]. Considering that the rebar in the TB is installed along the thermal bridges, 3D heat transfer simulations were conducted using Physibel-TRISCO. In addition, heating energy demands were analyzed with TRNSYS simulations, wherein thermal bridge effects were represented with the linear thermal transmittance obtained from Physibel-TRISCO simulations. In order to compare the TB alternatives with the existing thermal bridges, the balcony slab without TB was designated as reference cases for the evaluation of linear thermal transmittance and heating energy demands.
The remainder of this paper is organized as follows: Section 2 introduces the analysis cases and evaluation methods of the effects of TBs on the balcony slab thermal bridge and the heating energy demand; Section 3 describes the analysis results of the linear thermal transmittance and heating energy demand for different cases; Section 4 discusses the implications and limitations of this study; and Section 5 presents the conclusions of this study.

Evaluation Cases
To understand the balcony thermal bridges, a typical balcony slab attached to an exterior wall was modeled as the object of the analysis. There are three analytical conditions: a wall with external and internal insulation systems, an interior with an air heating system or radiant floor heating system, and a balcony slab with and without a TB. The original concrete balcony slab without any TBs was used as the reference case. As shown in Figure  1a, the thermal conductivity of steel reinforcing bar used in the balcony slab was 50 W/mK [38]. The analysis data of the reference cases were used to compare the results of the balcony thermal bridge with TBs. Two design options for thermal breaks were evaluated. To improve the effect of the balcony thermal bridge, the TB element was chosen first. As shown in Figure 1b, the TB element is mainly composed of an insulation block (insulation material composed with polystyrene hard foam) and stainless bars. The conductivities of the insulation block and stainless bars of the TB element are 0.031 W/mK [38] and 15 W/mK [38], respectively. When the TB element is installed in a balcony slab, the continuous steel reinforcing bars with high thermal conductivity are replaced by the stainless-steel bars in the TB element. Therefore, the TB element is effectively used for thermal break because the excessive heat flow through the reinforcing bars can be mitigated. However, the balcony slab with the TB element still has more potential to reduce heat loss, because the remaining stainlesssteel reinforcing bars in the balcony slab can become a heat flow path.
Therefore, a TB consisting of a GFRP rebar (TB-GFRP) was suggested to minimize heat flow through reinforcing bars, as described in Figure 1c. As mentioned in Section 1, GFRP rebar is a kind of non-conductive reinforcing bar, which has thermal conductivity of 0.35 W/mK [39], it has a low cost-to performance ratio, and is noncorrosive in nature [33]. Furthermore, the high tensile strength of the GFRP rebar makes it suitable for use as a steel reinforcing replacement. In the TB-GFRP element, the GFRP rebar replaced all the steel reinforcing bars of the concrete slab. To explain the thermal resistance of the GFRP rebar, the conductivity of the insulation block in the TB-GFRP element was set to the same value as that of the TB element. Table 1 shows the parameter comparison between the steel and GFRP, and Table 2 shows the material properties of the external wall and balcony slab. The parameter values were determined based on the data from the actual residential buildings. In addition, various insulation systems applied in buildings can affect the heat transmittance of thermal bridges. Internal insulation systems have been commonly applied to apartment buildings in Korea because of easy installation and low construction cost. However, the application of external insulation systems is steadily increasing to achieve energy saving target for building sectors [43]. Therefore, the analysis cases were divided into balcony thermal bridge models with internal and external insulation systems ( Figure 2 and Table 3). The standards of evaluating the thermal bridge are usually based on the assumption that the indoor condition is maintained by air heating systems [19]. However, most of residential buildings in Korea are equipped with radiant floor heating systems [44], wherein the indoor condition is maintained by heated floor structure. Thus, simulation models were set with and without a radiant floor heating system in order to investigate the effect of the radiant floor heating system on the balcony thermal bridge. The model without the floor heating system was assumed to be heated with an air heating system. Under different simulation conditions, the slab balcony with steel reinforcing bars was used as the reference case. Details of the analysis cases are listed in Table 3.

Simulation Modeling
Thermal bridges at the wall-balcony slab junction can be considered as linear thermal bridges, which can be evaluated using linear thermal transmittance defined in ISO 10211 [45]. Linear thermal transmittance is defined as the heat transfer per unit length when the temperature difference between the indoor and outdoor environments is 1 K under steady conditions. It can be calculated by dividing the heat flow from the thermal bridge by the length of the thermal bridge and the difference between the indoor and outdoor temperatures, as formulated by Equation (1).
where, ψ is the linear thermal transmittance Psi of the linear thermal bridge separating the two environments being considered (W/mK), is the thermal coupling coefficient obtained from a 2-D calculation of the component separating the two environments being considered (W/mK) is the thermal transmittance of the 1-D component j separating the two environments being considered (W/m 2 K), is the length within the 2-D geometrical model over which the value applies (m), and Nj is the number of 1-D components. Although and can be estimated through simple calculation, the thermal coupling coefficient needs to be calculated with a numerical method because two-dimensional heat transfer should be analyzed. In this study, Physibel-TRISCO simulation was used to obtain the thermal coupling coefficient, which is calculated by dividing the total heat flow through the entire structure by temperature difference (indoor and outdoor temperature difference). With the Physibel simulation result, the linear thermal transmittance ψ was calculated using Equation (1). Figure 3 shows the material properties and boundary conditions used in the Physibel simulations. The same structure layer and materials were applied to all simulation cases, except for the case of TBs. Internal and external insulation systems were also analyzed assuming the same material and thickness. For boundary conditions, the indoor and outdoor temperatures were set to 20 °C and 0 °C, respectively. Hot water temperature and flow rate for radiant floor heating system was assumed as 70 °C and 0.014 kg/s, which corresponds to the internal heat transfer coefficient of 806 W/m 2 K. The linear thermal transmittance was adopted to evaluate the heating energy demands as well as the impact of the TBs on the thermal bridge reduction. Heating energy demands of apartments with different balcony slab structures were analyzed using TRN-SYS simulation, wherein the linear heat transmittances of the thermal bridges were inputted as parameter of external walls. To evaluate the heating energy demand, a simulation model for a typical apartment house in Korea was developed as shown in Figure 4. The investigated apartment was composed of four heated zones: a living room and three bedrooms, which is one of the most common configurations in Korea. The location of the building was assumed as Incheon, which is located in the central region of South Korea. It was assumed that the main orientation of the building is the South and the balcony is attached to the South of the building. TRNSYS simulation models were developed in accordance with the above-mentioned Physibel simulation models, including the structure layers, external and internal thermal insulation systems, and two types of TB elements. Therefore, the heating energy demand of the apartment was also analyzed with radiant floor heating and air heating system. The simulation of heating energy demand was performed for a heating period of one year, with hourly time step. The heating period was defined based on the input weather data. EnergyPlus weather (EPW) data of Incheon [46] was applied to utilized for the annual simulation. In order to maintain the same boundary conditions with Physibel simulation, the indoor heating temperature was set to 20 °C. The occupancy schedule was set from 6 p.m. to 9 a.m. of the next day. In this study, the peak heating load and seasonal heating load (heating energy demand) were analyzed to illustrate the energy performance of different types of balcony slab structures. Table 4 shows the summarized TRNSYS simulation conditions.  Table 5 shows the simulation results of heat transfer at the wall-slab junction under air heating condition. It is shown that the reference cases (Cases 1 for external insulation and Case 4 for internal insulation) resulted in relatively large heat flow through the balcony slab, and slightly lower surface temperature at the wall-slab junction. In Cases 2, 3, 5, and 6 wherein the thermal break was applied, heat flow through the balcony slab was significantly reduced regardless of the insulation method.  Figure 5a shows the heat flow through the construction including the thermal bridge. With an external insulation system, the total heat transfer of the reference case (Case 1), TB (Case 2), and TB-GFRP (Case 3) was 25.41 W, 17.48 W, and 16.32 W, respectively. Compared to the reference case, the reduction rate of the heat loss at the thermal bridge was 31.2 % for TB and 35.8% for TB-GFRP case. It was found that the TB-GFRP can reduce the heat loss from the balcony thermal bridge by 4.6% compared to TB. With an internal insulation system, the reference case showed relatively larger heat flow (26.75 W) than that with external insulation system. The difference in the heat flow seems to be caused by the different heat flow path between external and internal insulation systems. The application of the TBs to the internal insulation system also showed the dramatic reduction in the heat flow. With TB (Case 5) and TB-GFRP (Case 6), the heat loss through the thermal bridge was reduced to 17.5 W and 16.32 W, respectively. The reduction rate was 34.6 for TB and 39.0% for TB-GFRP case, indicating that the TB-GFRP is more effective in mitigating the heat loss at thermal bridges. Figure 5b shows the results of linear thermal transmittance (ψ-value) with thermal breaks. The ψ-values of reference thermal bridge (Case 1), the thermal bridge with TB (Case 2) and TB-GFRP (Case 3) were 0.629 W/mK, 0.232 W/mK, and 0.174 W/mK, respectively, when the external insulation system was applied. In case of the internal insulation system, ψ-values were 0.696 W/mK (Case 4), 0.233 W/mK (Case 5), and 0.174 W/mK (Case 6), respectively. It is shown that the ψ-value was reduced using TBs, regardless of the external or internal insulation systems. The reduction rate of ψ-value was 63.1-72.3% for the external insulation, and 66.5-75.0% in the internal insulation. Table 6 shows the simulation results of heat transfer at the wall-slab junction under radiant floor heating condition. Compared to the results with air heating condition, the temperature inside the floor structure was considerably raised due to the heat supply by radiant floor heating system, which elevated the temperature difference between inside and outside balcony slab. As a result, the thermal bridge with radiant heating system showed much larger heat transfer than that with air heating system, as shown in Figure  6. Figure 6a shows that the thermal breaks are also effective in reducing heat flow through the balcony thermal bridge, even under radiant floor heating condition. In cases of external insulation system, the heat flow without thermal break was 36.01 W (Case 7), while the application of thermal breaks reduced the heat flow to 21.3 W (Case 8) and 19.43 W (Case 9). The reduction rate of the heat flow was 40.9% for TB and 46.0% for TB-GFRP case. With the internal insulation system, the heat flow of the reference case was 39.26 W (Case 10), while the application of thermal breaks reduced the heat flow down to 23.17 W (Case 11) and 20.73 W (Case 12), respectively. The reduction rate of heat flow was 41.0% for TB and 47.2% for TB-GFRP case.  The effect of thermal breaks on the heat flow reduction can be clearly evaluated using ψ-values. With the external insulation system, the ψ-value was 1.159 without thermal break (reference case), while it was reduced down to 0.423 W/mK (−63.5%) for TB and to 0.330 W/mK (−71.6%) for TB-GFRP. With the internal insulation system, the ψ-value of the reference case (1.321 W/mK) diminished to 0.517 W/mK (−60.9%) for TB and to 0.395 W/mK (−70.1%) for TB-GFRP. Table 7 summarizes the result of linear thermal transmittance with the thermal breaks at the balcony slab. It is obviously shown that the thermal break has significant impact on the reduction in the heat loss through the balcony slab. The reduction effect was consistently found regardless of insulation method and heating system. Although the radiant floor heating system leads to the substantial increase in the ψ-value, thermal breaks effectively lessen the ψ-value, which would be efficient to improve the heating system performance. In addition, it should be noted that the TB-GFRP can reduce the ψ-value by more than 10%, compared to the TB. Thus, the thermal conductivity of reinforcing bars in thermal break block needs to be reduced in order to achieve better insulation performance at the thermal bridges.  Figure 7 shows the simulation result of annual heating energy demands with air heating system. In both cases of external and internal insulation, the heating energy demand reduced when the TB (Case 2, 5) and TB-GFRP (Case 3, 6) were applied. For external insulation systems, the reduction rates of Cases 2 and 3 were 5.6% and 6.5%, respectively, compared to the heating energy demand without thermal breaks (Case 1, 46.99 kWh/m 2 ). For internal insulation system, the heating energy demand without thermal breaks (Case 4) was 50.61 kWh/m 2 . The application of TB (Case 5) and TB-GFRP (Case 6) reduced the heating energy demand by 6.1% and 7.0%, respectively. Peak heating loads for the analyzed cases were also evaluated, as shown in Figure 8. Similar to the heating energy demand results, the peak heating load also decreased with the application of thermal breaks. The reduction rate of peak heating load was up to 2.8% for the external insulation system and 3.1% for the internal insulation system. In both heating energy demand and peak heating load analysis, the TB-GFRP showed the biggest energy saving potentials at the balcony thermal bridge.  Figure 9 shows the heating energy demand under the radiant floor heating condition. It is shown that the radiant floor heating condition resulted in more heating energy demand than the air heating condition, because the heated floor structure increased the heat flow through the balcony thermal bridge. Nonetheless, the heating energy demand reduced with application of the TB and TB-GFRP. The reduction rates were 8.0% with TB and 9.3% with TB-GFRP in case of internal insulation system. For external insulation system, the reduction rates were 7.7% with TB and 9.2% with TB-GFRP. Internal insulation. Figure 10 shows the peak heating load under the radiant floor heating condition. Thermal breaks were effective in reducing the peak heating load as well as heating energy demand. The reduction rate of peak heating load was up to 4.0% for external insulation system and 4.2% for internal insulation system. It was found that the TB-GFRP can provide better energy saving performance under radiant floor heating condition, compared to the simple thermal break (TB).

Discussion
Simulation results showed that the thermal break at balcony slabs is crucial to mitigate the heat loss through the balcony thermal bridge. It was also shown that the thermal break using reinforcing bars with the reduced thermal conductivity (i.e., GFRP) can provide additional energy saving effect. In addition, the thermal breaks have more impact on the energy saving for residential buildings equipped with radiant floor heating system, because the radiant floor heating system causes more heat transfer through the balcony slab, compared to conventional air heating systems. Figure 11 summarizes the performance evaluation results of thermal breaks at balcony thermal bridge. It can be found that the heating energy demand has a linear relationship with the linear thermal transmittance at the balcony thermal bridge. Therefore, it is significant to reduce the linear thermal transmittance for the improvement of energy saving performance. Thermal breaks would be more significant for a residential building with more balconies, because the slope of the linear relationships between heating energy demand and linear thermal transmittance will be steeper. It is also found that the radiant floor heating and internal insulation systems lead to relatively larger linear thermal transmittance and heating energy demand, compared to air heating and external insulation system, respectively. Thus, the air heating system with external insulation system can be considered as a more energy-efficient solution. If it is inevitable to apply the radiant floor heating system with internal insulation system, thermal breaks should be applied to minimize the excessive heat loss thorough the balcony slabs. The application of the TB-GFRP rather than TB is recommended to improve the energy saving performance.
In this study, it was found that the thermal break is effective in reducing not only heating energy demand but also peak heating load. Thus, it would be possible to reduce the capacity of heating plants (e.g., boiler, heat exchanger) when the peak heating load is reduced by the application of thermal breaks. A further study is suggested to examine whether the heating capacity can be down-sized, considering the reduced peak heating load due to the thermal break.
This study was conducted to evaluate the impact of thermal breaks on heating energy demand. As the heating energy demand (or heating load) can be calculated without modeling HVAC system, this study implemented a so-called energy rate control approach [48] of TRNSYS simulation, wherein an ideal heat source supplies the required heat to maintain the set-point temperature. Nonetheless, a further study needs to be conducted with the HVAC systems modeled in order to investigate the energy consumption or economic feasibility of thermal breaks.

Conclusions
The objective of this study is to analyze the effect of a balcony slab thermal bridge on the heating energy demand of apartment buildings. To evaluate the heat loss caused by the thermal bridge, the linear thermal transmittance of the balcony slab thermal bridge was analyzed first. To reduce the heat loss at thermal bridges, two types of thermal breaks for balcony slabs (TB and TB-GFRP) were evaluated in terms of linear thermal transmittance and heating energy demand. The evaluation was conducted for different insulation methods (external and internal insulation) and heating conditions (air heating and radiant floor heating system).
The results showed that the TB and TB-GFRP reduce the linear thermal transmittance by 60.9-66.5% and 70.1-75.0%, respectively, depending on the insulation method and heating condition. The TB with low-conductivity reinforcing bars (TB-GFRP) showed the much reduction rate by 8.0-9.2% compared to the TB. It was clearly shown that the insulation performance of conventional TBs can be improved by replacing the steel reinforcing bars with the low-conductivity material such as GFRP. Annual simulation results showed that the TB and TB-GFRP can reduce the heating energy demand by 5.6-8.0% and 6.6-9.3%, respectively, depending on the insulation method and heating condition. It was also found that the TB-GFRP can save 0.9-1.5% more heating energy demand, compared to the TB. In particular, the impact of the TB-GFRP on the heating energy reduction was larger for a building with radiant floor heating system.
In conclusion, the investigated thermal break has potential for reducing heat loss through balcony thermal bridges. It is necessary to implement low-conductivity reinforcing bars (e.g., GFRP bars) for the thermal break in order to maximize the insulation effect of balcony thermal breaks. In particular, the thermal break needs to be applied to buildings with radiant floor heating system, because the heated slab structure can cause excessive heat loss at the thermal bridge.