The Impact of Shading Type and Azimuth Orientation on the Daylighting in a Classroom– Focusing on Effectiveness of Façade Shading, Comparing the Results of DA and UDI

There are many kinds of façade shading designs which provide optimal indoor daylighting conditions. Thus, considering combinations of different types of façade shading systems is an essential aspect in the optimization of daylighting in the building design process. This study explores (1) how the pattern and different characteristics are evaluated by varying façade shading types and considering their impact on daylighting metrics; and (2) the relative relationships between Daylight Autonomy (DA) and Useful Daylight Illuminance (UDI) with changes of the façade shading types, input parameters, and azimuth orientations. A typical high-school classroom has been chosen as a base model, and seven different façade shading types: vertical louver, horizontal louver, eggcrate louver, overhang, vertical slat, horizontal slat, and light shelf have been applied to eight azimuth orientations for the building. As tools for parametric design and indoor lighting analysis, Design Iterate Validate Adapt (DIVA)-for-Grasshopper has been used to obtain DA and UDI for comparison. Based on the simulation, (1) the effectiveness of the installation of façade shading compared to a non-shading case; and (2) design considerations for façade shading are presented. The result shows that there are some meaningful differences in DA and UDI metrics with the variation of orientation and façade shading types, although all cases of façade shading show some degree of decrease in DA and increase in UDI values. The types of shading devices which produce a dramatic decrease in DA values are the light shelf, horizontal slats, horizontal louvers, and eggcrate louvers. On the contrary, the types of shading devices which produce a dramatic increase in UDI values are the light shelf, horizontal slats, horizontal louvers, and eggcrate louvers. In the case of the vertical and vertical slat shading, the improvements of UDI values are significant in the east and west orientations. This demonstrates that the application and design of shading devices in certain façade orientations should be carefully considered for daylight control. Also, the results show that UDI explains relatively well the daylight performance in the case of the installation of a shading device.


Introduction
The design process plays an important role in developing sustainable buildings. In the case of different design processes, outcomes such as human comfort, reduction of energy consumption, and utilization of solar energy are of utmost importance. Thus, the design process should be highly In this research, two main objectives are explored. First, the pattern and different characteristics are evaluated by varying façade shading types and considering their impact on daylighting metrics. Various types of façade shadings, including vertical louver, horizontal louver, eggcrate louver, overhang, vertical louver slat, horizontal louver slat, and light shelf can be affected differently depending on the change in orientation and design. This research starts by analyzing the effect of façade shading types on daylighting using a parametric design tool. Secondly, the research compares and analyzes DA and UDI. There are two major differences between DA and UDI: (1) UDI concerns daylight illuminance in the range of 100 to 300 lux, which the DA range excludes, and which is considered effective even though the DA does not consider the range; (2) UDI perceives that a range over 2000 lux is undesirable and intolerable in terms of overlighting and glare. This study explores the changes in DA and UDI values with the application of various façade shading types and their characteristics of controlling the light condition. Ultimately, the research results can offer an organized façade shading design strategy through combined engineering technology and comprehensive understanding of manual façade shading design methods.

Basic Simulation Model
For simulation, a typical classroom space was used as a base model, which consisted of 5 m (16.4 ft) depth by 7.2 m (23.6 ft) width, with a floor-to-floor height of 3 m (9.8 ft), as shown in Figure 1. Input materials for the floor, wall, and ceiling were implemented as default values which are supported by DIVA in Table 1. The selected glazing type was 8 mm double-pane low-E glass. The analysis surface for the sensor nodes was set at the height of 0.75 m as the workplace height distance from the floor and divided into 70 modules, as shown in Figure 2. This simulation model does not consider any artificial lighting inside. The simulation was based on Incheon's (South Korea, Latitude 37) weather data, which is classified as a warm continental climate/humid continental climate (Dwa) in the Köppen climate classification.
Energies 2017, 10,635 3 of 20 In this research, two main objectives are explored. First, the pattern and different characteristics are evaluated by varying façade shading types and considering their impact on daylighting metrics. Various types of façade shadings, including vertical louver, horizontal louver, eggcrate louver, overhang, vertical louver slat, horizontal louver slat, and light shelf can be affected differently depending on the change in orientation and design. This research starts by analyzing the effect of façade shading types on daylighting using a parametric design tool. Secondly, the research compares and analyzes DA and UDI. There are two major differences between DA and UDI: (1) UDI concerns daylight illuminance in the range of 100 to 300 lux, which the DA range excludes, and which is considered effective even though the DA does not consider the range; (2) UDI perceives that a range over 2000 lux is undesirable and intolerable in terms of overlighting and glare. This study explores the changes in DA and UDI values with the application of various façade shading types and their characteristics of controlling the light condition. Ultimately, the research results can offer an organized façade shading design strategy through combined engineering technology and comprehensive understanding of manual façade shading design methods.

Basic Simulation Model
For simulation, a typical classroom space was used as a base model, which consisted of 5 m (16.4 ft) depth by 7.2 m (23.6 ft) width, with a floor-to-floor height of 3 m (9.8 ft), as shown in Figure 1. Input materials for the floor, wall, and ceiling were implemented as default values which are supported by DIVA in Table 1. The selected glazing type was 8 mm double-pane low-E glass. The analysis surface for the sensor nodes was set at the height of 0.75 m as the workplace height distance from the floor and divided into 70 modules, as shown in Figure 2. This simulation model does not consider any artificial lighting inside. The simulation was based on Incheon's (South Korea, Latitude 37) weather data, which is classified as a warm continental climate/humid continental climate (Dwa) in the Köppen climate classification.     Default material for standard external venetian blind slats and curtain wall frames, reflectivity of 52% * Reflectance: uniform diffuse reflection is applied; * Transmittance: the fraction of incident light, or other radiation, that passes through a substance; * Transmissivity: the measure of the capacity of a material to transmit radiation (the ratio of the amounts of energy transmitted and received); * Material properties' definitions from material.rad file in DIVA.
Energies 2017, 10, 635 4 of 20 5 Façade shading Matte_Silver Default material for standard external venetian blind slats and curtain wall frames, reflectivity of 52% * Reflectance: uniform diffuse reflection is applied; * Transmittance: the fraction of incident light, or other radiation, that passes through a substance; * Transmissivity: the measure of the capacity of a material to transmit radiation (the ratio of the amounts of energy transmitted and received); * Material properties' definitions from material.rad file in DIVA.

Simulation Program and Basic Concept
In this research, DA and UDI are used for the daylighting metrics. First, DA is conceived as annual daylight metrics, now commonly referred to as "dynamic daylight metrics" [26]. It is represented as a percentage of the annual daytime hours that a given point in a space is above a specified illumination level and at a threshold of 300 lux. Secondly, UDI is a modification of DA conceived by Mardaljevic and Nabil in 2005 [32,33]. UDI is defined as the annual occurrence of daylight illuminances across the work plane where the illuminances are within the range of 100-2000 lux, and are within a range considered "useful" by occupants [33]. The authors compare UDI as measuring excessive levels of illuminance which might cause glare, and which is ultimately targeted for a human factors-based metric, against DA, which is often used for minimizing artificial illuminance level [32,34].
As tools for parametric design and indoor lighting analysis, Grasshopper and DIVA are used [15] to devise alternatives by inputting different probable variations of façade shading types in relation to façade orientation. Grasshopper is a graphical algorithm editor that is a plug-in for Rhinoceros. Rhino (Rhinoceros) is a 3-dimensional (3D) Non-Uniform Rational Basis Spline (NURBS) modeler developed and distributed by McNeel. Grasshopper allows a designer or 3D modeler to program Rhino modeling. DIVA is one of Grasshopper's add-ons, which helps Grasshopper to conduct sustainability simulations such as daylight using Radiance open source tools [35].
Grasshopper is an effective parametric tool, especially when calculating a large number of cases, in this case 1,400 tries. It is utilized to access the relationships between input and output parameters after constructing algorithms. DIVA is an add-on program of Grasshopper, which can be applied to analyze DA and UDI. Thus, the principle and procedure of the simulation process is shown in Figure  3. The base model is drawn in Rhino and this information transfer into Grasshopper and (1) alternative modeling; (2) simulation; and (3) data sorting is performed through automation.
For the simulation, an algorithm is created to make shading combinations with three parameters, and each alternative takes about five minutes of lighting environment simulation time. Each louver has 200 alternatives, requiring a total of 1000 minutes of simulation time. Thus, it is more timeefficient than adjusting the variables manually.

Simulation Program and Basic Concept
In this research, DA and UDI are used for the daylighting metrics. First, DA is conceived as annual daylight metrics, now commonly referred to as "dynamic daylight metrics" [26]. It is represented as a percentage of the annual daytime hours that a given point in a space is above a specified illumination level and at a threshold of 300 lux. Secondly, UDI is a modification of DA conceived by Mardaljevic and Nabil in 2005 [32,33]. UDI is defined as the annual occurrence of daylight illuminances across the work plane where the illuminances are within the range of 100-2000 lux, and are within a range considered "useful" by occupants [33]. The authors compare UDI as measuring excessive levels of illuminance which might cause glare, and which is ultimately targeted for a human factors-based metric, against DA, which is often used for minimizing artificial illuminance level [32,34].
As tools for parametric design and indoor lighting analysis, Grasshopper and DIVA are used [15] to devise alternatives by inputting different probable variations of façade shading types in relation to façade orientation. Grasshopper is a graphical algorithm editor that is a plug-in for Rhinoceros. Rhino (Rhinoceros) is a 3-dimensional (3D) Non-Uniform Rational Basis Spline (NURBS) modeler developed and distributed by McNeel. Grasshopper allows a designer or 3D modeler to program Rhino modeling. DIVA is one of Grasshopper's add-ons, which helps Grasshopper to conduct sustainability simulations such as daylight using Radiance open source tools [35].
Grasshopper is an effective parametric tool, especially when calculating a large number of cases, in this case 1,400 tries. It is utilized to access the relationships between input and output parameters after constructing algorithms. DIVA is an add-on program of Grasshopper, which can be applied to analyze DA and UDI. Thus, the principle and procedure of the simulation process is shown in Figure 3. The base model is drawn in Rhino and this information transfer into Grasshopper and (1) alternative modeling; (2) simulation; and (3) data sorting is performed through automation.
For the simulation, an algorithm is created to make shading combinations with three parameters, and each alternative takes about five minutes of lighting environment simulation time. Each louver

Façade Shading Types and Input Parameters
As simulation alternatives, seven façade shading types: vertical louver, horizontal louver, eggcrate louver, overhang, vertical louver slat, horizontal louver slat, and light shelf were tested. In addition, the louver variation parameters consisting of count (number of shading devices), depth, direction (orientation), and angle are shown in Tables 2 and 3. There are vertical-installed façade shading types; vertical louver and vertical slat and horizontal façade shading types; horizontal louver and horizontal slat. The main difference between the vertical, horizontal, and eggcrate louver, and the overhang, vertical slat louver, horizontal slat louver and light shelf is that the former group has variation in the louver depth, whereas the latter group modifies the angle of the shading.

Façade Shading Types and Input Parameters
As simulation alternatives, seven façade shading types: vertical louver, horizontal louver, eggcrate louver, overhang, vertical louver slat, horizontal louver slat, and light shelf were tested. In addition, the louver variation parameters consisting of count (number of shading devices), depth, direction (orientation), and angle are shown in Tables 2 and 3. There are vertical-installed façade shading types; vertical louver and vertical slat and horizontal façade shading types; horizontal louver and horizontal slat. The main difference between the vertical, horizontal, and eggcrate louver, and the overhang, vertical slat louver, horizontal slat louver and light shelf is that the former group has variation in the louver depth, whereas the latter group modifies the angle of the shading.

Façade Shading Types and Input Parameters
As simulation alternatives, seven façade shading types: vertical louver, horizontal louver, eggcrate louver, overhang, vertical louver slat, horizontal louver slat, and light shelf were tested. In addition, the louver variation parameters consisting of count (number of shading devices), depth, direction (orientation), and angle are shown in Tables 2 and 3. There are vertical-installed façade shading types; vertical louver and vertical slat and horizontal façade shading types; horizontal louver and horizontal slat. The main difference between the vertical, horizontal, and eggcrate louver, and the overhang, vertical slat louver, horizontal slat louver and light shelf is that the former group has variation in the louver depth, whereas the latter group modifies the angle of the shading.

Façade Shading Types and Input Parameters
As simulation alternatives, seven façade shading types: vertical louver, horizontal louver, eggcrate louver, overhang, vertical louver slat, horizontal louver slat, and light shelf were tested. In addition, the louver variation parameters consisting of count (number of shading devices), depth, direction (orientation), and angle are shown in Tables 2 and 3. There are vertical-installed façade shading types; vertical louver and vertical slat and horizontal façade shading types; horizontal louver and horizontal slat. The main difference between the vertical, horizontal, and eggcrate louver, and the overhang, vertical slat louver, horizontal slat louver and light shelf is that the former group has variation in the louver depth, whereas the latter group modifies the angle of the shading.

Façade Shading Types and Input Parameters
As simulation alternatives, seven façade shading types: vertical louver, horizontal louver, eggcrate louver, overhang, vertical louver slat, horizontal louver slat, and light shelf were tested. In addition, the louver variation parameters consisting of count (number of shading devices), depth, direction (orientation), and angle are shown in Tables 2 and 3. There are vertical-installed façade shading types; vertical louver and vertical slat and horizontal façade shading types; horizontal louver and horizontal slat. The main difference between the vertical, horizontal, and eggcrate louver, and the overhang, vertical slat louver, horizontal slat louver and light shelf is that the former group has variation in the louver depth, whereas the latter group modifies the angle of the shading.

Façade Shading Types and Input Parameters
As simulation alternatives, seven façade shading types: vertical louver, horizontal louver, eggcrate louver, overhang, vertical louver slat, horizontal louver slat, and light shelf were tested. In addition, the louver variation parameters consisting of count (number of shading devices), depth, direction (orientation), and angle are shown in Tables 2 and 3. There are vertical-installed façade shading types; vertical louver and vertical slat and horizontal façade shading types; horizontal louver and horizontal slat. The main difference between the vertical, horizontal, and eggcrate louver, and the overhang, vertical slat louver, horizontal slat louver and light shelf is that the former group has variation in the louver depth, whereas the latter group modifies the angle of the shading.  Façade shading changes are based on the modification of parameters such as count, depth, and angle. When determining the criteria, we changed the parameter for the louver design based on the DA and UDI values in each direction (orientation) per louver. More specifically, the input parameter's variable ranges were selected so that the difference between the DA average value and the baseline value ranges from 2% to 30% in each azimuth orientation. The baseline DA value without shading is used as the common platform for the variation of each variable such as the angle, count, and depth. Therefore, each shading type has different parameter variable values. Accordingly, rather than  Façade shading changes are based on the modification of parameters such as count, depth, and angle. When determining the criteria, we changed the parameter for the louver design based on the DA and UDI values in each direction (orientation) per louver. More specifically, the input parameter's variable ranges were selected so that the difference between the DA average value and the baseline value ranges from 2% to 30% in each azimuth orientation. The baseline DA value without shading is used as the common platform for the variation of each variable such as the angle, count, and depth. Therefore, each shading type has different parameter variable values. Accordingly, rather than comparing the absolute mean values between different shading types, we have applied the relative  Façade shading changes are based on the modification of parameters such as count, depth, and angle. When determining the criteria, we changed the parameter for the louver design based on the DA and UDI values in each direction (orientation) per louver. More specifically, the input parameter's variable ranges were selected so that the difference between the DA average value and the baseline value ranges from 2% to 30% in each azimuth orientation. The baseline DA value without shading is used as the common platform for the variation of each variable such as the angle, count, and depth. Therefore, each shading type has different parameter variable values. Accordingly, rather than comparing the absolute mean values between different shading types, we have applied the relative Table 3. Façade shading changes are based on the modification of parameters such as count, depth, and angle. When determining the criteria, we changed the parameter for the louver design based on the DA and UDI values in each direction (orientation) per louver. More specifically, the input parameter's variable ranges were selected so that the difference between the DA average value and the baseline value ranges from 2% to 30% in each azimuth orientation. The baseline DA value without shading is used as the common platform for the variation of each variable such as the angle, count, and depth. Therefore, each shading type has different parameter variable values. Accordingly, rather Energies 2017, 10, 635 7 of 20 than comparing the absolute mean values between different shading types, we have applied the relative comparison within each shading for the main study. In terms of the number of sets, there are two types of parameters. There are eight different types of direction (orientation) and two parameters depending on shading type in terms of shape such as the angle, count, and depth. These make 25 cases in each direction (orientation) and total 200 combinations of variables in each façade shading type. In terms of UDI values, 100 to 2000 lux was set to compare the effectiveness of shading on daylight based on previous research [36][37][38].

Result and Discussions
In this section, the results based on the simulation will be presented. Most of the results show that the DA values decrease and UDI values increase with the installation of the shading system. In addition, basic descriptions and abbreviated terms that this paper uses are listed below.
-"Baseline": the basic value of each simulation result without façade shading based on the azimuth orientation -"Case": each 25 result values of each simulation based on the azimuth orientation and façade shading type -"Average", "Ave": the average value of each simulation "Case" based on the azimuth orientation and façade shading type -"Maximum", "Max": the maximum value of each simulation "Case" based on the azimuth orientation and façade shading type -"Minimum": the minimum value of each simulation "Case" based on the azimuth orientation and façade shading type -"Range": the difference between the "Maximum" and "Minimum" value -"Average-baseline": the difference between the "Average" and "Baseline" -"Orient": Azimuth orientation These values are used to explain the inherent characteristics of each shading and azimuth orientation. The basic process for the analysis of the results is: (1) to compare the "Average" and "Baseline" ("Average-baseline") which explains the relationship between DA and UDI in application of the shading system; and (2) to analyze "Range" which explains the sensitivity of the cases in modifying the shading system.

Vertical Louver
The simulation of the vertical louver was performed, and the results are shown in Figures 4 and 5, and Table 4. The average values of DA with the vertical louver are consistent as a whole in every direction (85.0-88.7%). The average values of DA are reduced (2.4-4.1%) compared to the baseline cases without a louver. This implies that more indoor daylight levels of 100-2000 lux (UDI range) and less than 300 lux (out of DA range) are created with the installation of the vertical louver. On the contrary, the improved average values of UDI are relatively consistent compared with the baseline values. This means that the UDI values within the range of 100-2000 lux daylighting conditions increase. In terms of façade shading applications in façade orientation, the south improved the most (7.7%) compared to the baseline, while the northwest the least, the difference of which was 1.6%. Also, when the vertical louver was applied to the south, the value shows 1.2 times improvement than to the west. Furthermore, DA decreases and UDI increases as the depth and count increase.

Horizontal Louver
The simulation of the horizontal louver was performed, and the results are shown in Figures 6 and 7, and Table 5. The average values of DA with the horizontal louver are the highest to the south (76.6%), and the values decrease further toward the north (64.1%). The average values of DA with the horizontal louver are less than the baseline values without a louver (14.5-25%). The difference between the maximum and minimum is large (55.5-85.1%) in each case, which implies that the louvers have a significant effect on the lighting environment. On the contrary, the average values of UDI in each azimuth orientation are following a changing trend of baseline values without the louver by having higher values when the orientation moves toward the north and lower values toward the south. The values improved compared to the baseline values and the improvement is highest to the south (16.2%). Also, when the horizontal louver is applied to the south, the value shows 1.57 times improvement than to the west. Furthermore, the DA value decreases and UDI value increases as the depth and count increase.

Horizontal Louver
The simulation of the horizontal louver was performed, and the results are shown in Figures 6  and 7, and Table 5. The average values of DA with the horizontal louver are the highest to the south (76.6%), and the values decrease further toward the north (64.1%). The average values of DA with the horizontal louver are less than the baseline values without a louver (14.5-25%). The difference between the maximum and minimum is large (55.5-85.1%) in each case, which implies that the louvers have a significant effect on the lighting environment. On the contrary, the average values of UDI in each azimuth orientation are following a changing trend of baseline values without the louver by having higher values when the orientation moves toward the north and lower values toward the south. The values improved compared to the baseline values and the improvement is highest to the south (16.2%). Also, when the horizontal louver is applied to the south, the value shows 1.57 times improvement than to the west. Furthermore, the DA value decreases and UDI value increases as the depth and count increase.

Horizontal Louver
The simulation of the horizontal louver was performed, and the results are shown in Figures 6 and 7, and Table 5. The average values of DA with the horizontal louver are the highest to the south (76.6%), and the values decrease further toward the north (64.1%). The average values of DA with the horizontal louver are less than the baseline values without a louver (14.5-25%). The difference between the maximum and minimum is large (55.5-85.1%) in each case, which implies that the louvers have a significant effect on the lighting environment. On the contrary, the average values of UDI in each azimuth orientation are following a changing trend of baseline values without the louver by having higher values when the orientation moves toward the north and lower values toward the south. The values improved compared to the baseline values and the improvement is highest to the south (16.2%). Also, when the horizontal louver is applied to the south, the value shows 1.57 times

Eggcrate Louver
The simulation of the eggcrate louver was performed, and the results are shown in Figures 8 and  9, and Table 6. The average values of DA with the eggcrate louver are highest to the south (71.3%) and decreasing to the north (56.8%). The average values of DA are less than the baseline values without a louver and the difference range is from 13.1% to 22.5%. The difference between the maximum and the minimum is large in each case of the eggcrate louver, which indicates that the façade shading affects the lighting environment. However, the average values of UDI with the eggcrate louver follow the trend of baseline values. The values improved compared to the baseline values, and the improvement is highest to the south (15.6%). Also, when the eggcrate louver is applied to the south, the value shows 1.58 times improvement than to the west. Furthermore, DA decreases and UDI increases as the depth and count increase.

Eggcrate Louver
The simulation of the eggcrate louver was performed, and the results are shown in Figures 8 and  9, and Table 6. The average values of DA with the eggcrate louver are highest to the south (71.3%) and decreasing to the north (56.8%). The average values of DA are less than the baseline values without a louver and the difference range is from 13.1% to 22.5%. The difference between the maximum and the minimum is large in each case of the eggcrate louver, which indicates that the façade shading affects the lighting environment. However, the average values of UDI with the eggcrate louver follow the trend of baseline values. The values improved compared to the baseline values, and the improvement is highest to the south (15.6%). Also, when the eggcrate louver is applied to the south, the value shows 1.58 times improvement than to the west. Furthermore, DA decreases and UDI increases as the depth and count increase.

Eggcrate Louver
The simulation of the eggcrate louver was performed, and the results are shown in Figures 8 and 9, and Table 6. The average values of DA with the eggcrate louver are highest to the south (71.3%) and decreasing to the north (56.8%). The average values of DA are less than the baseline values without a louver and the difference range is from 13.1% to 22.5%. The difference between the maximum and the minimum is large in each case of the eggcrate louver, which indicates that the façade shading affects the lighting environment. However, the average values of UDI with the eggcrate louver follow the trend of baseline values. The values improved compared to the baseline values, and the improvement is highest to the south (15.6%). Also, when the eggcrate louver is applied to the south, the value shows

Overhang
The simulation of the overhang was performed, and the results are shown in Figures 10 and 11, and Table 7. The average DA values of the overhang are consistently high as a whole (82.7-86.1%), and the values are less than the baseline values without a louver (5-6.4%). On the contrary, the average UDI values with the overhang follow the trend of baseline values without a louver. The values are consistently better than the baseline (4.3-11.8%). This means that the values within the range of 100-2000 lux have risen. In terms of façade orientation, the values are better toward the south. The south shows the biggest improvement (11.8%). The improvement to the south is 1.53 times higher than to the west. Furthermore, the DA decreases and UDI increases as the depth and count increase.

Overhang
The simulation of the overhang was performed, and the results are shown in Figures 10 and 11, and Table 7. The average DA values of the overhang are consistently high as a whole (82.7-86.1%), and the values are less than the baseline values without a louver (5-6.4%). On the contrary, the average UDI values with the overhang follow the trend of baseline values without a louver. The values are consistently better than the baseline (4.3-11.8%). This means that the values within the range of 100-2000 lux have risen. In terms of façade orientation, the values are better toward the south. The south shows the biggest improvement (11.8%). The improvement to the south is 1.53 times higher than to the west. Furthermore, the DA decreases and UDI increases as the depth and count increase.

Overhang
The simulation of the overhang was performed, and the results are shown in Figures 10 and 11, and Table 7. The average DA values of the overhang are consistently high as a whole (82.7-86.1%), and the values are less than the baseline values without a louver (5-6.4%). On the contrary, the average UDI values with the overhang follow the trend of baseline values without a louver. The values are consistently better than the baseline (4.3-11.8%). This means that the values within the range of 100-2000 lux have risen. In terms of façade orientation, the values are better toward the south. The south shows the biggest improvement (11.8%). The improvement to the south is 1.53 times higher than to the west. Furthermore, the DA decreases and UDI increases as the depth and count increase.

Vertical Slat
The simulation of the vertical slat was performed, and the results are shown in Figures 12 and  13, and Table 8. The average DA values with the vertical slat louver are consistently high as a whole (79.8-86%). The values with the louver decrease compared with the baseline values without a louver (5.1-9.3%). On the contrary, the average UDI values with the vertical slat louver follow the trend of baseline values without a louver. The values are consistently better than the baseline (9.8-12.8%). This means that the values in the range of 100-2000 lux have risen. In terms of façade orientation, the values are better further south, which shows the biggest improvement (12.8%). The improvement of the UDI value to the south is 1.2 times more than to the west. Furthermore, the slat angle perpendicular to the window shows the maximum DA value and the minimum UDI value.

Vertical Slat
The simulation of the vertical slat was performed, and the results are shown in Figures 12 and  13, and Table 8. The average DA values with the vertical slat louver are consistently high as a whole (79.8-86%). The values with the louver decrease compared with the baseline values without a louver (5.1-9.3%). On the contrary, the average UDI values with the vertical slat louver follow the trend of baseline values without a louver. The values are consistently better than the baseline (9.8-12.8%). This means that the values in the range of 100-2000 lux have risen. In terms of façade orientation, the values are better further south, which shows the biggest improvement (12.8%). The improvement of the UDI value to the south is 1.2 times more than to the west. Furthermore, the slat angle perpendicular to the window shows the maximum DA value and the minimum UDI value.

Vertical Slat
The simulation of the vertical slat was performed, and the results are shown in Figures 12 and 13, and Table 8. The average DA values with the vertical slat louver are consistently high as a whole (79.8-86%). The values with the louver decrease compared with the baseline values without a louver (5.1-9.3%). On the contrary, the average UDI values with the vertical slat louver follow the trend of baseline values without a louver. The values are consistently better than the baseline (9.8-12.8%). This means that the values in the range of 100-2000 lux have risen. In terms of façade orientation, the values are better further south, which shows the biggest improvement (12.8%). The improvement of the UDI value to the south is 1.2 times more than to the west. Furthermore, the slat angle perpendicular to the window shows the maximum DA value and the minimum UDI value.

Horizontal Slat
The simulation of the horizontal slat was performed, and the results are shown in Figures 14 and  15, and Table 9.

Horizontal Slat
The simulation of the horizontal slat was performed, and the results are shown in Figures 14 and  15, and Table 9.

Horizontal Slat
The simulation of the horizontal slat was performed, and the results are shown in Figures 14  and 15, and Table 9. improvement (17.9%) and the north produces the least (4.6%). The improvement of the UDI value to the south is 1.63 times more than to the west. Furthermore, the DA increases and UDI decreases as the angle of the slat approaches the solar altitude angle.
Energies 2017, 10,635 13 of 20 the south is 1.63 times more than to the west. Furthermore, the DA increases and UDI decreases as the angle of the slat approaches the solar altitude angle.

Light Shelf
The simulation of the light shelf was performed, and the results are shown in Figures 16 and 17, and Table 10    the south is 1.63 times more than to the west. Furthermore, the DA increases and UDI decreases as the angle of the slat approaches the solar altitude angle.

Light Shelf
The simulation of the light shelf was performed, and the results are shown in Figures 16 and 17, and Table 10

Light Shelf
The simulation of the light shelf was performed, and the results are shown in Figures 16 and 17, and Table 10. The average DA values of the light shelf are the highest to the south (74.1%) and the lowest to the north (61.6%). The average DA values decreased compared with the baseline cases (17-27.5%). The differences between the minimum and the maximum values are large in each case (67.4-84.2%). This means that the light shelf affects the daylighting environment. The average UDI values of the light shelf follow the trend of the baseline values without a louver. The further south, the better the values are, which shows the biggest improvement (16.8%). The north is the opposite, with the lowest improvement (5.5%). The differences between the minimum and the maximum are large (19.6%-51.1%). The improvement of the UDI value to the south is 1.5 times more than to the west. lowest improvement (5.5%). The differences between the minimum and the maximum are large (19.6%-51.1%). The improvement of the UDI value to the south is 1.5 times more than to the west.

The Effectiveness of Façade Shading Installation Compared with the Non-Shading Case
Based on the simulation results of seven different shading types, the effectiveness of shading installation is compared with non-shading cases. The graphs shown in Figures 18 and 19 describe the decreased average value or increased average value compared to the non-shading installation case (base case) in each DA and UDI. The DA values at the baseline of each azimuth orientation are between 89.1-90.9%, and the maximum difference was very small, within 1.8%. However, all cases with façade shading show decreased DA and increased UDI values. More specifically, the types of shading devices which produce a dramatic decrease in the DA values are the light shelf, horizontal slats, horizontal louvers, and eggcrate louvers. In the baseline case without façade shading, the DA values are greater than 89% in every direction.  lowest improvement (5.5%). The differences between the minimum and the maximum are large (19.6%-51.1%). The improvement of the UDI value to the south is 1.5 times more than to the west.

The Effectiveness of Façade Shading Installation Compared with the Non-Shading Case
Based on the simulation results of seven different shading types, the effectiveness of shading installation is compared with non-shading cases. The graphs shown in Figures 18 and 19

The Effectiveness of Façade Shading Installation Compared with the Non-Shading Case
Based on the simulation results of seven different shading types, the effectiveness of shading installation is compared with non-shading cases. The graphs shown in Figures 18 and 19 describe the decreased average value or increased average value compared to the non-shading installation case (base case) in each DA and UDI. The DA values at the baseline of each azimuth orientation are between 89.1-90.9%, and the maximum difference was very small, within 1.8%. However, all cases with façade shading show decreased DA and increased UDI values. More specifically, the types of shading devices which produce a dramatic decrease in the DA values are the light shelf, horizontal slats, horizontal louvers, and eggcrate louvers. In the baseline case without façade shading, the DA values are greater than 89% in every direction. The DA values decrease with the application of shading devices, which means an indoor daylight level below 300 lux is created. The decrease of the DA values are small in the case of the vertical louver, overhang, and vertical slat. The differences of daylighting conditions, which decrease the DA values, are not large depending on these types of shading devices, which means that these shading devices controlling daylight below the level of 300 lux do not affect the indoor lighting environment. shading devices, which means an indoor daylight level below 300 lux is created. The decrease of the DA values are small in the case of the vertical louver, overhang, and vertical slat. The differences of daylighting conditions, which decrease the DA values, are not large depending on these types of shading devices, which means that these shading devices controlling daylight below the level of 300 lux do not affect the indoor lighting environment. On the contrary, in the case of the UDI, where 100-2000 lux light levels are measured, the measurement is made excluding excessive light inflow. Therefore, when the façade shading is installed, the UDI values improve compared with the baseline case without shading.
The types of façade shading which produce a dramatic increase in UDI values are horizontal slats, light shelf, horizontal louvers, and eggcrate. These façade shading types improve UDI performance when installed in the south orientation, but the value decreases in the east, north, and west orientations. The overhang type has a lower performance in improving UDI value compared with the other façade shading systems, except for the vertical louver. In the case of the vertical louver and vertical slat, the improvements of UDI values are significant in the east and west orientations. This demonstrates that the application and design of shading devices in certain façade orientations should be carefully considered for daylight control. DA values are small in the case of the vertical louver, overhang, and vertical slat. The differences of daylighting conditions, which decrease the DA values, are not large depending on these types of shading devices, which means that these shading devices controlling daylight below the level of 300 lux do not affect the indoor lighting environment. On the contrary, in the case of the UDI, where 100-2000 lux light levels are measured, the measurement is made excluding excessive light inflow. Therefore, when the façade shading is installed, the UDI values improve compared with the baseline case without shading.
The types of façade shading which produce a dramatic increase in UDI values are horizontal slats, light shelf, horizontal louvers, and eggcrate. These façade shading types improve UDI performance when installed in the south orientation, but the value decreases in the east, north, and west orientations. The overhang type has a lower performance in improving UDI value compared with the other façade shading systems, except for the vertical louver. In the case of the vertical louver and vertical slat, the improvements of UDI values are significant in the east and west orientations. This demonstrates that the application and design of shading devices in certain façade orientations should be carefully considered for daylight control. On the contrary, in the case of the UDI, where 100-2000 lux light levels are measured, the measurement is made excluding excessive light inflow. Therefore, when the façade shading is installed, the UDI values improve compared with the baseline case without shading.
The types of façade shading which produce a dramatic increase in UDI values are horizontal slats, light shelf, horizontal louvers, and eggcrate. These façade shading types improve UDI performance when installed in the south orientation, but the value decreases in the east, north, and west orientations. The overhang type has a lower performance in improving UDI value compared with the other façade shading systems, except for the vertical louver. In the case of the vertical louver and vertical slat, the improvements of UDI values are significant in the east and west orientations. This demonstrates that the application and design of shading devices in certain façade orientations should be carefully considered for daylight control.

Design Considerations for Façade Shading Systems
Based on the simulation data (the differences between each maximum and minimum; maximum value − minimum value in DA and UDI results), the design consideration for façade shading systems is presented with the DA and UDI value range graphs shown in Figures 20 and 21. In the case of the vertically installed façade shading, already known for its effectiveness of shading east and west exposures [39,40], the result shows that the vertical slat in particular has advantages in the UDI measurement compared to the other shading devices in the east, northeast, north, northwest, and west exposures. Also, the vertical louver and slat increased the UDI value evenly regardless of the azimuth orientation, which demonstrates that they are more effective in the east and west compared with the other shadings. In horizontally installed façade shading, known for being useful for installations in the south orientation, both the horizontal louver and slat and eggcrate louver have advantages in UDI values compared with the other shading types. The overhang type shows a relatively similar pattern between vertically and horizontally installed louver types considering the DA and UDI values. Lastly, the light shelf shows relatively peculiar characteristics because it has a different style of shading properties based on the material's reflectance, and it also shows the greatest value in terms of fluctuation between the maximum and minimum scale. Thus, careful consideration for selecting suitable properties such as depth, angle, and azimuth orientation should be implemented. Furthermore, in most cases, there might be a chance to solve overlight problems, such as glare, in not only the south, east, and west, but also in the north, since the installation of façade shading decreases DA and increases UDI in all orientations.

Design Considerations for Façade Shading Systems
Based on the simulation data (the differences between each maximum and minimum; maximum value − minimum value in DA and UDI results), the design consideration for façade shading systems is presented with the DA and UDI value range graphs shown in Figures 20 and 21. In the case of the vertically installed façade shading, already known for its effectiveness of shading east and west exposures [39,40], the result shows that the vertical slat in particular has advantages in the UDI measurement compared to the other shading devices in the east, northeast, north, northwest, and west exposures. Also, the vertical louver and slat increased the UDI value evenly regardless of the azimuth orientation, which demonstrates that they are more effective in the east and west compared with the other shadings. In horizontally installed façade shading, known for being useful for installations in the south orientation, both the horizontal louver and slat and eggcrate louver have advantages in UDI values compared with the other shading types. The overhang type shows a relatively similar pattern between vertically and horizontally installed louver types considering the DA and UDI values. Lastly, the light shelf shows relatively peculiar characteristics because it has a different style of shading properties based on the material's reflectance, and it also shows the greatest value in terms of fluctuation between the maximum and minimum scale. Thus, careful consideration for selecting suitable properties such as depth, angle, and azimuth orientation should be implemented. Furthermore, in most cases, there might be a chance to solve overlight problems, such as glare, in not only the south, east, and west, but also in the north, since the installation of façade shading decreases DA and increases UDI in all orientations.
In summary, the shading types which have the least impact on increasing the UDI value are vertical louvers, overhang, and vertical slats in the order of magnitude of the improved values.

Conclusions
This study reveals some meaningful differences between DA and UDI metrics with the variation of the orientation and façade shading types. Although all cases of façade shading show some degree of decrease in DA and increase in UDI values, the results show the characteristics of controlling the light. The DA only considers the conditions of 300 lux or more, whereas the UDI considers the conditions between 100 and 2000 lux. The decrease of DA creates the environment of 300 lux or less. At the same time, the increase of UDI creates the environment of 100 to 300 lux.
In particular, the types of shading devices which produce a dramatic decrease in the DA values are the light shelf, horizontal slats, horizontal louvers, and eggcrate louvers. On the contrary, the types of shading devices which produce a dramatic increase in the UDI values are the light shelf, horizontal slats, horizontal louvers, and eggcrate louvers. In the case of the vertical louver and vertical slat, the improvements of the UDI values are significant in the east and west orientations.
This study also finds that the effectiveness of shading devices using DA values are not sufficient enough because they only use daylight levels of more than 300 lux, whereas for indoor lighting conditions, i.e., 100-300 lux of the daylight level, the excessive daylight level needs to be considered. It is known that UDI is effective in terms of shading performance evaluation [38]. This study reconfirms the effectiveness of UDI in the case of Incheon (South Korea, Latitude 37), South Korea. Namely, the DA conditions are only for lights with 300 lux or more, so that the values of DA are higher without a shading device than with. However, in the case of UDI, where the range of the minimum and maximum values is defined as 100-2000 lux, UDI is effective when excessive daylight needs to be controlled by shading devices.
Although the criteria may vary in certain environments, a minimum of 300 lux in illuminance level is generally recommended for classrooms, based on the IESNA (Illuminating Engineering Society of North America) guide [41]. At the DA condition (above 300 lux), when the classroom is illuminated sufficiently, the electric lighting use can be reduced and lead to energy saving. Additional effects when DA and UDI are used together include the facts that DA provides energy saving related data and UDI provides data on useful daylight range (minimum lighting source and glare possibility). Therefore, when the two metrics are used together, it can help to examine both the amount of energy and useful daylight at the same time.
Considering that LEED encourages using climate-based analyses such as DA, this comparison between DA and UDI in the change of shading and azimuth orientation might justify reconsideration for shading devices and their rating systems. To explain, the analytic technique for daylight analysis changes from static to dynamic conditions such as DA and UDI, and it is shown that there are huge differences in the DA and UDI values in the application of façade shading when we use those daylight metrics.

Conclusions
This study reveals some meaningful differences between DA and UDI metrics with the variation of the orientation and façade shading types. Although all cases of façade shading show some degree of decrease in DA and increase in UDI values, the results show the characteristics of controlling the light. The DA only considers the conditions of 300 lux or more, whereas the UDI considers the conditions between 100 and 2000 lux. The decrease of DA creates the environment of 300 lux or less. At the same time, the increase of UDI creates the environment of 100 to 300 lux.
In particular, the types of shading devices which produce a dramatic decrease in the DA values are the light shelf, horizontal slats, horizontal louvers, and eggcrate louvers. On the contrary, the types of shading devices which produce a dramatic increase in the UDI values are the light shelf, horizontal slats, horizontal louvers, and eggcrate louvers. In the case of the vertical louver and vertical slat, the improvements of the UDI values are significant in the east and west orientations.
This study also finds that the effectiveness of shading devices using DA values are not sufficient enough because they only use daylight levels of more than 300 lux, whereas for indoor lighting conditions, i.e., 100-300 lux of the daylight level, the excessive daylight level needs to be considered. It is known that UDI is effective in terms of shading performance evaluation [38]. This study re-confirms the effectiveness of UDI in the case of Incheon (South Korea, Latitude 37), South Korea. Namely, the DA conditions are only for lights with 300 lux or more, so that the values of DA are higher without a shading device than with. However, in the case of UDI, where the range of the minimum and maximum values is defined as 100-2000 lux, UDI is effective when excessive daylight needs to be controlled by shading devices.
Although the criteria may vary in certain environments, a minimum of 300 lux in illuminance level is generally recommended for classrooms, based on the IESNA (Illuminating Engineering Society of North America) guide [41]. At the DA condition (above 300 lux), when the classroom is illuminated sufficiently, the electric lighting use can be reduced and lead to energy saving. Additional effects when DA and UDI are used together include the facts that DA provides energy saving related data and UDI provides data on useful daylight range (minimum lighting source and glare possibility). Therefore, when the two metrics are used together, it can help to examine both the amount of energy and useful daylight at the same time.
Considering that LEED encourages using climate-based analyses such as DA, this comparison between DA and UDI in the change of shading and azimuth orientation might justify reconsideration for shading devices and their rating systems. To explain, the analytic technique for daylight analysis changes from static to dynamic conditions such as DA and UDI, and it is shown that there are huge differences in the DA and UDI values in the application of façade shading when we use those daylight metrics.
This study has a certain degree of limitations. It suggests simulation results based on weather data of a specific location, therefore it may not be applied to all the cases. Since, UDI thresholds are set as 2000 lux based on the default value of the DIVA program setting, the UDI threshold of 2000 lux is used in this experiment. However, if the threshold is set differently i.e., 2500 or 3000 lux, the value and trend of the result graph can vary. Therefore, the interpretation may be different depending on the location, building type, or the threshold setting, which should be considered in future studies. Also, the input shading parameter values were selected based on the minimum 2% or maximum 30% DA range in each azimuth orientation. Therefore, it is relatively difficult to compare and prove the effectiveness of the respective shading types, and this study focused on the relative comparison within each shading, rather than comparing the absolute mean values between different types of shading. Thus, the results may vary depending on the parameter variations selected for each shading type. Nevertheless, with these simulation techniques and methods, a designer and an engineer can establish more detailed metrics for daylight availability measurement as well as realize the inherent benefits of each shading system on human comfort.