Due to the unique structural and illuminance conditions of old industrial buildings, a daylighting strategy should be reconsidered when beginning a remodelling effort [7].

The original structure of old industrial buildings generally featured small and narrow windows. However, some owners or designers of old factory buildings chose to keep the initial window size for illuminance, which may not reach the daylighting requirement when given a new function. Even with some windows fixed on the south façade, daylight cannot penetrate a deep factory space. Therefore, during the renovation of old factory buildings, the size of the windows should be appropriately increased through a proper method.

The position and orientation of windows can affect the uniformity of daylighting performance in renovation design. In the current art studio renovation projects occurring in old factory buildings, many select a single window type, such as a small horizontal window, without considering design, leading to nonuniform daylighting performance of the interior space and daylight failing to arrive at a suitable area in need of light [8].

Multiple window types can solve this problem by installing clerestories or skylights to supplement the insufficient light space and ensure that the illuminance performance is more uniform and stable.

During renovations, some designers have blindly enlarged the window size to maximize daylighting efficiency without carefully considering the potential glare problem of daylighting performance. However, art studios prefer diffuse sunlight rather than intensive and directional sunlight. In the 798 Art District, the extensive glazing area can provide enough daylight for the interior environment. However, a potential glare problem also emerged, especially on sunny days. In some cases, changing the visual transmittance of glazing and adding shading or a diffusion device can alleviate and even solve this problem [9].

DIVA-for-Rhino is a highly optimized daylighting and energy modeling plug-in for the Rhinoceros—NURBS modeler [10]. The plug-in was initially developed at the Graduate School of Design at Harvard University and is now distributed and developed by Solemma LLC. DIVA-for-Rhino allows users to conduct a series of environmental performance evaluations of individual buildings and urban landscapes, including radiation maps, photorealistic renderings, climate-based daylighting metrics, annual and individual time step glare analysis, LEED and CHPS daylighting compliance, and single thermal zone energy and load calculations.

The project needs to judge the building lighting from multiple dimensions of illumination, so the following lighting parameters shown in Table 2 are adopted for the project: daylight factor [11] [12], illumination uniformity (U0) [12],

daylight autonomy [13], and daylight glare probability [14].

Guangzhou is the capital and largest city of Guangdong Province in South China. Located on the Pearl River, approximately 120 km northwest of Hong Kong, Guangzhou is an important national transportation hub and trading port. One of the five National Central Cities, it holds subprovincial administrative status. Figure 5 shows that the site is situated in the Tianhe district near two universities, which have convenient transportation and a good scholarly atmosphere.

Located just south of the Tropic of Cancer, Guangzhou has a humid subtropical climate influenced by the East Asian monsoon. Summers are wet with high temperatures, high humidity, and a high heat index. As shown in Figure 6, we can see that the solar altitude angle changes from time to time, which will affect the shading design of the building to avoid overheating in summer and heat gain in winter.

The sky cover range in Guangzhou shows that most of the sky conditions in Guangzhou are overcast. In November, as the cold wind blows from the north, the sky will be clearer than in other months.

From the above illumination range diagram shown in Figure 7, we can see that the illumination is relatively high during September and December; according to the sky cover condition shown in Figure 8, the more the sky clears, the more illumination the building gains.

After renovation design, the building has become a workshop with artistic functions and atmosphere, including an art studio, exhibitions, catering, offices and accommodations. Since the art studio is the most essential part of the building, Figure 9 shows that it is placed on the south side of the building, where it receives more daylight and a good view.

The excessive depth makes it difficult for the building to allow light access, and the window size causes a low occupied time percentage of daylight autonomy (Figure 11). Only 48.08% of the occupied time reaches 300 lux.

On the west facade, there is a two-storey building next to the factory with no gap at all, causing the west facade to not receive enough daylight hours horizontally.

Through the simulation of the daylight factor by DIVA, as shown in Figure 12, the mean daylight factor of the art studio is 1.29%, which falls short of the minimum daylight factor requirement of CIBSE for art studios.

If all the walls and doors are open, Figure 13 shows the maximum amount of daylight that the building will receive. The mean daylight factor can reach 19.2% if all the walls and columns are gone.

The next step is to determine the window-to-wall ratio (WWR). From the

references and case studies, there are three optional WWRs: 1. 1:5; 2. 1:4; 3. 1:3.

1) WWR = 1:5

From the simulation result a 1:5 WWR, we can see that the window near the art studio is 2000 × 8000 mm (Figure 14), and the mean daylight factor is 4.68, as shown in Figure 15(a), which fails to reach the mean daylight factor criteria of CIBSE (mean daylight factor: 5). In Figure 15(b), the daylight autonomy graph that sets 300 lux as the target illuminance, 84.43% of the time occupied, can reach the minimum illuminance requirement of CIBSE for the art studio. The daylight glare probability (DGP) in Figure 16 is concentrated from 8:00 am to 10:00 am and focuses on the south direction. Figure 16 shows that the DGP level can still be tolerated.

2) WWR = 1:4

From the simulation result of a 1:4 WWR, we can see that the window near the art studio is 3000 × 8000 mm (Figure 17). Figure 18(a) shows that the mean daylight factor is 5.95, and the minimum daylight factor is 2.1, which is higher than the 1:5 WWR and reaches the mean daylight factor criteria of CIBSE but still cannot meet the requirement of the minimum daylight factor (mean daylight factor: 5; minimum daylight factor: 2.5). The uniformity ratio of this option is 0.35, which cannot fulfil the requirement of CIBSE (Uniformity ≥ 0.7). In Figure 18(b), the daylight autonomy graph that sets 300 lux as the target illuminance, 89.01% of the time occupied, can reach the minimum illuminance requirement of CIBSE for the art studio. The daylight glare probability is concentrated from 8:00 am to 10:00 am and focuses on the south direction. Figure 19 shows that the DGP level is slightly higher than the previous option, but some shading devices can still reduce this kind of glare.

3) WWR = 1:3

From the simulation result of a 1:3 WWR, we can see that the window near the art studio is 3500 × 10,000 mm (Figure 20). Figure 21(a) shows that the mean daylight factor is 6.66, and the minimum daylight factor is 2.6, which is higher than the 1:4 WWR and reaches the daylight factor criteria of CIBSE. However, the uniformity ratio of 0.4 still cannot reach the standard of 0.7. In Figure 21(b), the daylight autonomy graph that sets 300 lux as the target illuminance, 91.25% of the time occupied, can reach the minimum illuminance

requirement of CIBSE for the art studio. The DGP is concentrated from 8:00 am to 16:00 am and focuses on the south direction. Figure 22 easily illustrates that the DGP level is much higher than the other two options, and the period time is much longer, which can be improved by choosing different window types in the next step, as well as the problem of uniformity.

After comparative analysis, Table 3 shows that the 1:3 WWR would be the best option for an adequate daylight factor and daylight autonomy, while the uniformity and glare problem needs to be fixed in the next step.

From the previous option, although the WWR of 1:3 is a proper choice, it still cannot reach the requirement of the uniformity ratio standard of CIBSE, and the glare problem also needs to be solved. Therefore, the window’s type and the position may be another aspect that deserves more consideration when solving this

problem.

In this step, three types of windows can be chosen: 1) Horizontal windows (including clerestories); 2) Skylights; 3) Multiple.

1) Horizontal Windows

According to the window-to-wall ratio, the primary window size should be 3500 × 8500 mm. Considering the uniformity of the daylight, this could be separated into a 3000 × 8500 mm horizontal window and a 500 × 8500 mm clerestory window, as shown in Figure 23. From the daylighting factor (Figure 24),

we can see that the mean daylight factor is nearly 6, which reaches the standard of CIBSE, but the daylight is concentrated near the window and makes a dark corner on the part that is far away from the light source; therefore, the uniformity of this option is just 0.3, which fulfils the requirement of CIBSE for art studios (0.7). Although the glare problem still remains (Figure 25), it has been slightly reduced due to the use of a clerestory.

2) Skylights

In this option, the window type is a skylight. Therefore, the horizontal window on the south façade will be blocked, and the only daylighting source will be the skylight.

The skylight is directed towards the sun, and the height of the building is just

9000 mm, which does not have enough height to diffuse the daylight. Therefore, the size of the skylight will be 3200 × 15000 mm (Figure 26); at the same time, a 1800 × 13,000 mm translucent coating shown in Figure 26(b) will be fixed under the skylight to diffuse the daylight.

From the daylight factor data shown in Figure 27(a), the mean daylight factor is 5.43, and the uniformity ratio is 0.71, which fulfils the requirement of CIBSE. The daylight autonomy in Figure 27(b) is also relatively high, higher than the horizontal window.

Since there are no horizontal windows on the south façade, the perspective views of glare have changed towards the skylight, and from the DGP results in Figure 28 and Figure 29, we find that the intolerable glare starts from 8:00 am to 10:00 am and extends longer in summer.

However, while the simulation results are satisfactory, there is still a problem with the skylight, which means the lack of a natural view of the room that the occupants need.

3) Multiple

The combination of a horizontal window with a skylight may be the optimal option for daylighting performance, with an excellent natural view and uniformity and sufficient daylight. As the space has two sources of daylight, the horizontal window size should be modified to some extent to prevent excess daylight and maintain the window-to-wall ratio. As a result, the window size was reduced to 2500 × 8000 mm, including a 500 × 8000 mm clerestory (Figure 30).

Figure 31(a) shows that the mean daylight factor is 6.22, and the uniformity ratio is 0.72, which reaches the daylight standard of CIBSE for an art studio. The daylight autonomy in Figure 31(b) also has 100% area to reach the 300-lux minimum illuminance requirement. Compared to the previous options, this is the most uniform option. The daylight glare shown in Figure 32 still needs to be solved; although it is better than other options, the glazing type and visible transmittance data should be chosen in the next step.

After the simulation and comparison, Table 4 shows that the multiple window type is the optimal option among the three, which not only can fulfil the requirement of the standard but can also allow the occupants to have good vision,

which most artists and art workers truly need.

By ameliorating the window-to-wall ratio and window type, daylight issues are almost solved except for the glare problem. The glazing type and its visible light transmittance index can address this problem by reducing the visible light from the outside.

There are two options for glazing and visible transmittance: 1) single-panel glazing with 88% visible transmittance and 2) double-panel low-e glazing with 65% visible transmittance.

1) Single-panel Glazing with 88% Visible Transmittance

Single-panel glazing is the glass that is commonly used in buildings. The high visible transmittance allows daylight to access the space to the maximum extent. The daylight factor figure shown in Figure 33(a) shows that the daylight factor is 6.18, and the uniformity ratio is 7.2, reaching the standard of CIBSE for an art studio. Although the glare problem in Figure 34 still has a slight influence in the morning, as it is not a compulsory requirement for daylight performance, it can be solved by some shading devices.

2) Double-panel Low-e Glazing with 65% Visible Transmittance

Although the double-panel low-e glazing can protect the occupants from the

glare problem (Figure 35), the material also reduces the daylight factor to 3.31 (Figure 36), which cannot fulfil the compulsory requirement of CIBSE for an art studio.

From the simulation analysis and the data comparison, Table 5 shows that although the low-e glazing can effectively reduce the glare, it also makes the daylight factor fail to reach the criteria of CIBSE. Therefore, single-panel glazing with 88% visible transmittance can be chosen for the art studio, and another shading device may improve the glare issues.