Daylighting assessment of window layouts and architectural elements in early design stages

Abstract

While window design evolves throughout the design process, establishing an initial concept is crucial. Architects often prioritize aesthetics when deciding on window layouts, yet considering daylighting performance early can significantly improve occupant health, well-being, and productivity. This study conducts a parametric analysis to assess the daylighting performance of different window layouts and architectural elements, including room layout, orientation, window to wall ratio, and visible transmittance (VT). Five window layouts were examined: square, rectangular, horizontal, vertical, and window wall. A total of 1,872 cases was simulated for Hanoi’s climate using Ladybug Tools. Results reveal substantial differences in daylighting performance among the layouts. Rectangular windows provide deeper illumination, while square windows distribute light more broadly near the window. Horizontal windows achieve greater daylight availability than vertical windows. Appropriately designed room dimensions and VT allow horizontal or vertical windows to maximize spatial daylight autonomy (sDA), while the window-wall configuration offers no substantial advantage. Window to wall ratio and VT are critical variables, capable of shifting sDA from 40 to 100%. Orientation also affects daylighting, with sDA differences of up to 15% between north- and west-facing windows. Based on these findings, early-stage window design guidelines are proposed to support effective daylighting performance.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-23389-x.

Introduction

Windows play a crucial role in connecting indoor and outdoor environments. Designing windows is a complex challenge for architects, especially during the early design stages. While aesthetic considerations often take precedence, effective window design must also account for daylighting and thermal performance, both of which are influenced by key architectural factors such as room layout, window orientation, window-to-wall ratios (WWR), glazing properties, and window layouts^1,2. Daylighting, in particular, is frequently overlooked despite its proven links to occupant health, well-being, and productivity, making it a critical factor in achieving high-quality indoor environments³.

Although common window layouts—including window-wall, rectangular, vertical, horizontal, and square—have been studied with regard to occupant preferences⁴, their impacts on meeting occupants’ daylight need remains largely unexplored.

This study evaluates daylighting performance across various window layouts in combination with different architectural elements, including room layout, orientation, window size, and glazing properties. Simulations are conducted using validated Ladybug tools, with daylight availability quantified by spatial daylight autonomy (sDA). Using Hanoi—a low-latitude city—as the simulation context, the findings provide practical insights for similar climatic regions and support evidence-based window design in early architectural decision-making.

Literature review

Key architectural elements affecting daylight

In the initial stages, architects usually establish the room layout first to meet the needs of occupants. While it significantly affects the functionality of the space, it also plays a crucial role in shaping window design to enhance daylighting performance. Several studies have explored how room depth influences daylight. Reinhart and LoVerso⁵ indicated that greater room depth can limit daylight penetration and recommended that room depth not exceed 2.5 times the window-head-height. Vanhoutteghem et al.⁶ demonstrated that achieving visual comfort in deep or narrow south-oriented rooms can be challenging and may lead to overheating. Cammarano et al.⁷ observed a decline in the percentage of daylit areas as room depth increased, with a more substantial reduction occurring between depths of 3 m to 6 m compared to the depths beyond 6 m. Kose and Kazanasmaz⁸ found that to meet the minimum requirements set by the Illuminating Engineering Society (IES), deeper rooms require larger glazing areas. Specifically, a 12-meter-deep room should have a window with a WWR of 67%, a 9-meter-deep room should have a WWR of 43%, and a 6-metereter-deep room should have a WWR of 30%. Consequently, room layouts that are excessively deep or narrow are generally less optimal for daylighting.

Window orientation is another architectural element that is determined early in the design process and is sometimes constrained by the site’s orientation. For locations in the Northern Hemisphere, south-facing windows receive the most sunlight, while north-facing windows receive the least^7,9–11. Dubois and Flodberg⁹ and Acosta et al.¹⁰ observed a moderate amount of sunlight on west-facing windows, while Cammarano et al.⁷ found that west-facing windows received a similar level of sunlight to south-facing ones. These orientation-based differences in solar exposure have important implications for daylighting potential.

Window-to-wall ratio (WWR) also has a substantial impact on daylighting^12,13. However, this impact is not linear. A previous study showed that increasing WWR from 30% to 40% results in a more noticeable improvement in daylight availability than increasing it from 40 to 60%⁷. Cammarano et al.⁷ and Shen and Tzempelikos¹⁴ found no significant increase in useful daylight illuminance (a daylighting metric) for WWR values above 40%, suggesting that a 40% WWR is generally sufficient to ensure “useful” daylight. Other studies, however, have indicated that the optimal WWR varies by orientation; for north-facing windows, larger WWRs are recommended to compensate for lower solar radiation^15,16.

WWR not only affects daylighting but also influences a building’s energy performance and aesthetics; therefore, it is usually determined early in the design process. Once decided, however, it constrains the range of potential window layouts.

Window layout is determined by three key factors: window size, window shape, and window position. The choice of window size and placement can be influenced by the room dimensions^17,18. You et al.¹⁹ demonstrated that vertically adjusting window position can reduce the need for artificial lighting. Common window layouts include window-wall, rectangular, vertical, round, horizontal, and square configurations. Acosta et al.¹⁰ found that horizontal windows provide more uniform illumination than square or vertical windows. In a study of Mediterranean summer (Csa) and continental arid summer (Dsa) climates (classified according to the Koppen-Geiger system²⁰, Kharvari²¹ found horizontal windows to be more effective for daylighting than vertical windows.

The selection of window glazing is another critical factor affecting both building energy performance and daylight availability, although it is often determined at a later design stage. Among glazing properties, visible transmittance (VT) is most directly related to daylight, representing the proportion of the visible light spectrum transmitted through the glazing²². For example, A glazing system with a VT of 82% transmits 15% more visible light (380–750 nm) than one with a VT of 74%²³. However, higher VT values are not always preferable, as excessive transmittance can lead to glare issues²⁴.

Daylighting metrics and simulation

Daylight availability is commonly quantified using daylight factor (DF) and dynamic metrics such as daylight autonomy (DA) and useful daylight illuminance (UDI). DF is a static metric measured at a specific time and does not account for location or orientation, while dynamic metrics address these limitations. Spatial daylight autonomy (sDA), an advanced form of DA defined by the IES LM-83-12 standard²⁵, is widely adopted in green building certifications like LOTUS in Vietnam²⁶ and LEED in the US²⁷. This metric measures whether a workplane, represented by a sensor grid, receives adequate natural lighting. If this workplane does not meet the acceptable glare threshold, measured by annual sunlight exposure (ASE), applicants may submit a plan to address glare, such as installing roller shades. Therefore, this study did not include ASE.

Many parametric studies have evaluated daylighting performance metrics using building simulation software. Alwetaishi and Taki²⁸ employed Tas²⁹, a tool widely used in the UK³⁰, to calculate daylight factor (DF). Chi et al.³¹ calculated DF by using Ecotect, however, which has not been updated since 2015 and is now part of the Revit environment³², though not all original functions were retained. Several studies^33,34 had suggested that the accuracy of Ecotect is unreliable. For example, Vangimalla et al.³⁴ found that Ecotect overestimates illuminance levels.

For dynamic metrics, Goia¹⁶ calculated daylight autonomy (DA) and useful daylight illuminance (UDI) using EnergyPlus³⁵. However, the daylighting simulation results from EnergyPlus showed no variation from the middle to the back of the room, suggesting limitations in modeling internal reflectance and reduced accuracy for deep-room simulations³⁶. When validating indoor illuminance at various simulation points, the mean bias error (MBE) ranged from 81.5% to 463.4%, with the highest MBE occurring at the deepest point, 5 m from the window³⁷. This large MBE was attributed to very low measured illuminance, while EnergyPlus tended to overestimate results.

Acosta et al.¹⁰ used Daysim³⁸ to simulate daylight autonomy, and its accuracy has been validated in several studies^39,40. To address areas with sudden changes in solar exposure, Daysim DDS was introduced, offering improved accuracy⁴¹. In the same room used to validate indoor illuminance with EnergyPlus, Yun and Kim³⁷ observed the highest MBE of observed the highest MBE of − 18.9% at the midpoint. However, Daysim DDS approximates the sun’s position, which can still affect precision. Subramaniam and Mistrick⁴² developed a method using precise sun locations, later integrated into the “Honeybee Annual Daylight” component in Ladybug Tools⁴³, which demonstrated superior performance compared to Daysim DDS.

Recently, several researchers have begun using AI and machine learning to predict daylight^44,45. For example, Wang et al.⁴⁶ employed a conditional generative adversarial network (cGAN) to predict daylighting performance based on room parameters including width, depth, ceiling height, and window locations. The method demonstrated high accuracy and significantly faster simulation speed—only 0.3 s compared to the 15 s required by conventional Climate-Based Daylight Modeling (CBDM). Han et al.⁴⁷ reported similarly high accuracy, with coefficients of determination (R²) of 0.988 and 0.996, while being approximately 250 times faster than a typical CBDM tool. However, the AI and machine learning-based models are not well-suited for parametric studies, since its reliability may be limited if the training dataset does not adequately cover the targeted variables or design conditions.

Methodology

Characteristics of the room model

To conduct a parametric study, room models with a flat roof, floor, and four walls—one of which includes windows—are built. The parameters studied are listed in Table 1, totaling 1,872 cases. The window orientations were assessed in the four cardinal directions: south, west, north, and east. This selection simplifies the parameter space and isolates the effect of orientation by focusing on the most distinct and commonly referenced cases in daylighting analysis.

Table 1.

Model parameters.

Parameter		Value
Orientation		South; West; North; East
Room dimensions (m)	Height	3
	Width	3; 6; 9
	Depth	3; 6; 9
Window-to-wall ratio (%)	Square and rectangular windows	20; 30; 40
	Horizontal and vertical windows	50; 60; 70
	Window-wall	100
Surface reflectance (%)	Indoor ceiling	80
	Indoor wall	50
	Indoor floor	20
	Outdoor ground	20
Visible transmittance (%)		29; 50; 73; 90
Office hours		8 AM − 5 PM

Figure 1 illustrates nine typical room layout scenarios, ranging from narrow to wide rooms. The room height is fixed at 3 m, while the width and depth vary between 3 m, 6 m, and 9 m. The ceiling height of 3 m was chosen as it reflects a common standard for residential and office buildings in Vietnam, such as those studied by Son and Huyen¹. This simplification allows the analysis to focus on the impact of floorplan dimensions and window layout.

Fig. 1.

Nine room sizes when installing window-wall.

Figure 2 presents five window layout types, previously evaluated by Dogrusoy and Tureyen⁴ in a study that examined only occupant preferences. However, these layouts have not yet been analyzed and compared for their impact on daylighting across different WWRs. As noted in the literature review, the chosen WWR restricted certain window layout options; therefore, this study examines a total of thirteen unique room—window layout combinations. Square and rectangular windows are the smallest, covering 20%, 30% or 40% of the wall area. For coverage of 50%, 60%, or 70%, horizontal windows replace square windows, and vertical windows replace rectangular ones. The fifth type is a window wall with a 100% WWR. Square, rectangular, and horizontal windows are installed at workplane height, 0.8 m from the floor.

Fig. 2.

Five window layouts for three room widths with different WWRs.

Four VT values, ranging from low to high, are selected from commercial productions in Vietnam, as shown in Table 1. Surface reflectance values for the ceiling, walls, and floor were set at 80%, 50%, and 20%, respectively, in accordance with LEED guidelines²⁷.

Location

The location under investigation is Hanoi (Vietnam), situated at 21º north latitude. The sun path of Hanoi is showed in Fig. 3. According to the Koppen-Geiger classification system, Hanoi’s climate is characterized as monsoon-influenced humid subtropical (Cwa)²⁰. Typical weather data, sourced from the Ladybug database⁴⁸, were used as input for the simulations conducted in this study. Hanoi experiences significant seasonal differences in solar radiation, with the high levels occurring from May to August—averaging approximately 620 Wh/m² of hourly global horizontal radiation—and the low levels from January to March, averaging about 340 Wh/m².

Fig. 3.

Hanoi’s sun path.

Daylighting simulation

The room model was constructed in Rhino⁴⁹ and simulated using Ladybug tools⁴³, which operate within the Grasshopper environment of Rhino. The simulation workflow is shown in Fig. 4. The room geometries were imported into Grasshopper as Honeybee faces and integrated with daylighting and material properties described in Table 1 using the “Honeybee Room” component.

Fig. 4.

Daylighting simulation workflow.

Subsequently, a sensor grid as shown in Fig. 5, was created at a workplane height of 0.8 m above the floor using the “Honeybee Sensor Grid from Rooms” component. This grid contains sensor points spaced 300 mm apart to define the positions and orientations of the illuminance sensors in the simulations. The “Honeybee Sensor Grid from Rooms” component was then connected to the “Honeybee Annual Daylight” component, the core tool for running annual daylighting simulations.

Fig. 5.

Sensor points generated for a 6 m wide × 9 m deep room.

The simulation quality was set to high accuracy using ambient bounces (ab) of 6, ambient division (ad) of 25,000, and limit weight (lw) of 0.00007. Additional inputs included the weather data file and the selection of widow orientation.

The simulation output consisted of illuminance values at each sensor point for every office hour, which were used to calculate sDA. This metric was defined as the percentage of the sensor grid maintaining an illuminance of at least 500 lx for at least 50% of office hours throughout the year, ensuring sufficient daylight for activities such as reading and writing.

Results

All simulation results for sDA are provided in the Supplementary Information online (Tables A1-A3). The main findings and several representative cases are discussed below.

Effect of room layout

Table 2 presents a typical case in Hanoi used to compare the effects of varying room width and depth, featuring square windows covering 20% of the south-facing wall area and fitted with glazing with a VT of 50%. The results show that increasing the room width increases the percentage of daylit area, whereas increasing the room depth decreases it. Specifically, for every 3 m increase in width, sDA rose to 100–121% of its previous value. In contrast, for every 3 m increase in depth, sDA dropped to 42–68% of its previous value, with a more pronounced decline observed near the window.

Table 2.

Spatial daylight autonomy (sDA) for south-facing rooms with square windows (20% WWR, 50% VT glazing).

		sDA (%)			sDA change (%)
		3 m depth	6 m depth	9 m depth	From 3 m to 6 m depth	From 6 m to 9 m depth
sDA (%)	3 m width	53	25	17	47	68
	6 m width	64	30	17	47	57
	9 m width	76	32	20	42	63
sDA change (%)	From 3 m to 6 m width	121	120	100
	From 6 m to 9 m width	119	107	118

When increasing the room width from 3 m to 6 m for the 3 m deep room, sDA rose from 53% to 64%. Figure 6a,b illustrate that the daylit area extended toward the back wall. Correspondingly, the percentage of office hours during which illuminance reached at least 500 lx increased from 70% to 80%, measured at a depth of 2.5 m in the middle of the room.

Fig. 6.

Daylit areas for three room layouts with square windows facing south (20% WWR, and 50% VT glazing).

Comparing the two 3 m wide rooms shown in Fig. 6a,c, sDA dropped from 53% to 25% as room depth increased from 3 m to 6 m, consistent with the findings of Cammarano et al.⁶. This reduction resulted not only from unlit areas deeper than 3 m but also from decreased daylight in the first 3 m. At a point 1.5 m from the west wall and 2.5 m from the window, the percentage of acceptable office hours dropped from 70% to 50%.

Effect of window orientation

In Hanoi, north-facing windows received minimal sunlight year-round, resulting in the lowest sDA and necessitating larger windows to ensure adequate daylight. West-facing windows typically achieved the highest sDA, but this orientation is less desirable for daylighting due to the potential for overheating and glare. South-facing windows recorded the second-highest sDA, while east-facing windows performed significantly worse than their west-facing counterparts. Table 3 presents the observed trends. For example, in the case of vertical windows with a 50% WWR, sDA gradually increased with orientation—from 50% for the north-facing window to 55% (east), 58% (south), and 65% (west). The sDA difference between the north and west orientations reached up to 15%. These findings are consistent with studies conducted in Turin, Catania, and Berlin⁷.

Table 3.

Spatial daylight autonomy (sDA) for some 6 m wide × 9 m deep rooms with 73% VT glazing.

Window layout	WWR (%)	sDA (%) for each window orientation
		South	West	North	East
Rectangular	20	31	32	25	29
	30	43	47	38	41
	40	56	58	49	50
Vertical	50	58	65	50	55
	60	67	74	58	63
	70	75	94	67	71
Window-wall	100	100	100	97	100

Due to the sun’s seasonal path, the effects of window orientation vary over time. As shown in Fig. 7, south-facing windows provided broader illumination across the workplane during winter. Daylight patterns also shifted throughout the day for east- and west-facing windows: east-facing windows received higher illumination in the morning, while west-facing windows experienced peak sunlight in the afternoon. This pattern can result in insufficient morning light and excessive afternoon light in west-facing rooms, contributing to glare and thermal discomfort. East-facing rooms exhibited a similar effect, with morning glare and reduced light levels in the afternoon.

Fig. 7.

Illuminance in a 3 m wide × 6 m deep room at 9 AM and 3 PM on a sunny day in winter and summer.

Effects of WWR and window layout

As expected, higher WWR values generally resulted in higher daylight levels; however, in this study, WWR is treated primarily as a constraint for determining suitable window layouts rather than as the main variable of interest. With WWR from 20% to 40%, we can consider square or rectangular windows. With WWR from 50% to 70%, we can consider horizontal or vertical windows which will allow more daylight availability in the space. And with 100%, window-wall will be used with maximizing daylight availability. A detailed evaluation of the window layouts is provided below.

Square window and rectangular window

Considering the same WWR, as shown in Table 4, rectangular windows generally resulted in higher sDA values, indicating greater daylight availability. This is due to their ability to allow daylight to penetrate deeper into the room, as illustrated in Fig. 8. At a depth of 1.3 m in the middle of the room with 29% VT, the simulated illuminance remained above 500 lx for 80% of office hours with rectangular windows, compared to 70% with square windows—a difference of 10% points.

Table 4.

Simulation results for selected 3 m × 3 m rooms with south-facing windows.

WWR (%)	Window layout	VT (%)	Spatial daylight autonomy (sDA, %)
20	Square	29	25
		50	53
		73	89
		90	100
	Rectangular	29	22
		50	61
		73	100
		90	100
50	Horizontal	29	81
		50	100
		73	100
		90	100
	Vertical	29	67
		50	100
		73	100
		90	100
100	Window-wall	29	100
		50	100
		73	100
		90	100

Fig. 8.

Daylit areas for the 3 m wide × 3 m deep rooms with south-facing windows and 20% WWR.

However, square windows provided a wider daylighting area near the windows, resulting in brighter conditions under low daylight scenarios, such as when visible transmittance was low or the room was narrow. This is also evident in Fig. 8: at a depth of 0.5 m and with 29% VT, the room with rectangular windows showed larger underlit areas in the two corners, leading to lower sDA values in those areas compared to the room with square windows.

Horizontal window and vertical window

Vertical windows typically extend from floor to ceiling, but daylight entering below 0.8 m does not effectively illuminate the workplane. Consequently, horizontal windows provided better workplane illumination, as reflected in their higher or comparable sDA values compared to vertical windows, as shown in the online Supplementary Information. For example, Table 4 compares both types at the same WWR of 50% and VT of 29%, where horizontal windows achieved an sDA of 81%, while vertical windows reached only 67%.

Window-wall window

The use of a window wall increased sDA values compared to other window layouts due to its larger window area. However, when the room depth was not more than twice the room width, or when the glazing system’s VT was at least 73%, the effect of the window wall on sDA was negligible, as the sDA reached the maximum value of 100% with a WWR of 70%.

Effect of visible transmittance

VT demonstrated a substantial impact on daylight availability. As shown in Fig. 9, variations in glazing systems within a fixed WWR configuration resulted in notable differences in sDA. Specifically, replacing a high-performance glazing system with 29% VT by a standard system with 90% VT led to a sharp increase in sDA from 41 to 100%. The magnitude of this effect was comparable to that observed when increasing window area. For instance, with a glazing system of 90% VT, increasing the WWR from 20% to 100% led to an increase in sDA from 33 to 100%.

Fig. 9.

sDA values for the 6 m wide × 9 m deep rooms with east-facing windows when changing WWR and VT.

Discussion

Window design guidelines in early stages

When developing a design concept, many architects focus on integrating local culture, ensuring harmony with surrounding buildings and landscapes, and implementing a strong design vision—often overlooking aspects of visual comfort. Some may be unfamiliar with simulation tools or have limited knowledge of high-performance building design. However, as recommended by green building rating systems such as LOTUS and LEED, the concept design phase is the optimal time to adjust a project to meet energy efficiency and visual comfort standards. Changes made at later stages can be challenging or may require extensive modifications to achieve the desired performance.

To support architects in the early stages of window design, we developed a straightforward workflow, as showed in Fig. 10, to help ensure adequate daylighting. The process is as follows:

Fig. 10.

Workflow for window design in the early design stages.

• Design room layout (Step 1): When determining a room’s layout with daylighting in mind, architects should aim to maximize room or floor width while minimizing depth. However, constraints such as site shape, space usage requirements, prevailing wind conditions, structural considerations, and budget limitations may restrict the ability to achieve the ideal layout.

• Design window layout (Steps 2–4): The same constraints may also limit options for determining window orientation, but some general guidelines can be followed. In Hanoi, orienting windows to the west or south can enhance indoor illumination but may also introduce glare issues, and for west-facing windows, overheating concerns. Placing windows primarily on northern walls can reduce the need for shading devices.

Architects should consider the preferred window configuration and increase the WWR if greater daylight availability is desired. For WWRs under 50%, square or rectangular windows provide better daylighting performance. For WWRs between 50% and 100%, horizontal or vertical windows perform better. A full window wall (100% WWR) will maximize indoor illumination.

• Adjust glazing properties (Step 5): If a space experiences excessive or insufficient daylight, adjusting the VT of the glazing is an effective strategy for achieving optimal daylighting.

This workflow provides architects with practical guidance to balance daylighting requirements with architectural constraints effectively.

Trade-off between thermal loads and daylighting requirements

In building design, achieving an optimal balance between thermal loads and daylighting requirements is often a critical consideration. Larger window areas (larger WWR) and higher visible transmittance (VT) values can improve daylight availability but may also increase heat gains in summer and heat losses in winter, thereby affecting a building’s thermal performance.

However, the present study focuses specifically on the influence of window layout—rather than window size or glazing properties—on daylighting performance and daylight distribution patterns. Since the total WWR and glazing specifications are held constant in the comparative analysis, variations in window layout are expected to have only a minimal impact on thermal loads. Therefore, while the general trade-off between thermal and daylighting performance is important in integrated design, it lies outside the primary scope of this work.

Differences between user preferences and daylighting performance in selecting window layout

Dogrusoy and Tureyen⁴, in their survey of office buildings, reported that most occupants preferred large, uninterrupted windows—particularly full window-wall and horizontal layouts—due to their visual openness and connection to the outdoors. Interestingly, our daylighting simulations confirm that these two layouts generally deliver the highest daylight availability. However, the window-wall design did not always produce better results; in some cases, its performance plateaued, suggesting that “bigger” does not always mean “brighter”. Moreover, large glass areas can increase the risk of glare and unwanted heat gain, which may reduce comfort and energy efficiency.

When window size is limited, Dogrusoy and Tureyen⁴ identified square windows as the second-best option. In our study, however, rectangular windows often performed better, especially in narrow rooms, as they allow daylight to reach further into the space. This suggests that optimal window design should not be guided by occupant preference alone, but rather by an integrated evaluation of daylighting performance, room shape, and potential comfort trade-offs.

Limitations

To simplify the developed workflow and support its practical use in early-stage building form determination, this study focuses on ensuring adequate daylight availability while excluding the assessment of discomfort glare and excessive heat gain. Evaluating glare and heat gain across 1,872 simulation cases would be inefficient, particularly in the absence of shading devices or dynamic smart window technologies, which are essential for meaningful glare control. Incorporating such systems would add substantial methodological complexity, as their performance varies depending on whether they are static or dynamic and is highly influenced by building orientation. Moreover, glare can often be effectively managed during building operation through operational measures such as roller shades. According to LEED v4.1, daylight credits are awarded when the design meets sDA thresholds and includes a strategy for glare control using shading devices. Based on this rationale, discomfort glare, excessive heat gain, as well as shading devices and dynamic smart window technologies, are not considered in this study but can be addressed in future research as part of later design development stages.

In addition, this study focuses exclusively on rooms with a single window wall to isolate and compare the effects of different window layouts on daylighting performance. While this approach allows a clearer assessment of each layout’s impact, it does not account for corner rooms, which commonly have openings on multiple walls. As a result, the findings may not fully represent daylighting conditions in such spaces.

To enhance the practical applicability of this research, future studies should expand the investigation to include ordinal orientations, specifically examining daylighting performance for windows facing northeast, southeast, southwest, and northwest directions. Additionally, given that room geometry significantly influences indoor daylight availability, subsequent research should systematically evaluate the effects of varying room dimensions, including different room heights, widths, and depths beyond the 3-meter ceiling height and 9 floorplan combinations (3 widths × 3 depths) examined in this study. These expanded investigations would provide a more comprehensive understanding of how window design strategies perform across diverse spatial configurations and orientations, thereby offering more robust guidance for practical building design applications.

Conclusions

This parametric study assessed the daylighting performance of five window layouts—square, rectangular, horizontal, vertical, and window wall—highlighting their distinct effects on daylight distribution. Additional architectural elements, including room layout, orientation, window size, and visible transmittance (VT), were also analyzed.

The findings showed that increasing room width significantly enhanced spatial daylight autonomy (sDA), rising from 53% to 64% as the width increased from 3 m to 6 m. In contrast, increasing room depth reduced sDA to 40–70% of its original value. Rectangular windows delivered deeper daylight penetration, maintaining illuminance above 500 lx for 80% of office hours at a depth of 1.3 m—10% points higher than square windows (70%). Horizontal windows provided higher overall daylight levels than vertical windows. When room depth was less than twice the width or when glazing VT exceeded 73%, both orientations achieved full sDA (100%), reducing the need for window wall solutions.

Both WWR and VT were identified as dominant factors influencing daylight performance, with variations capable of shifting sDA values from approximately 40% to 100%. Daylight availability also varied by orientation, with north-facing windows yielding the lowest sDA and south- or west-facing orientations achieving the highest. The sDA difference between north and west orientations reached up to 15%. However, when designing with orientations and WWR, other constraints and energy considerations may also play a significant role.

This study offers a practical workflow for architects to evaluate window design options during the early design phase. By comparing the daylighting impacts of different window layouts and associated design parameters, practitioners can make informed decisions that enhance both visual comfort and daylight performance.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Cong Thanh Do developed the conceptualization, methodology, performed the data analysis, and wrote the original draft. Nguyen Thi Thuy Dung conducted the simulations and prepared the visualizations. Ying-Chieh Chan contributed to data analysis, revised and approved the final version for publication, and secured funding. All authors agree to be accountable for all aspects of the work.

Data availability

All data generated or analyzed during this study are included in this published article (and its Supplementary Information file).

Declarations

Competing interests

The authors declare no competing interests.