Daylight performance of post-disaster container housing: a comparative study of alternative spatial configurations

1 Introduction

Housing is one of the most essential human needs, providing shelter and security. In the aftermath of natural, human-made, or hybrid disasters, it is also among the most vulnerable components of the built environment, often subject to severe damage or complete destruction. The loss of housing can lead to population displacement, the fragmentation of social structures, and significant physical and psychological impacts on affected populations (Mustak, 2022; Terumoto, 2024; Asadi et al., 2025). In response, post-disaster housing recovery aims to reduce future vulnerabilities and strengthen community resilience (Peacock et al., 2018). This recovery process comprises emergency shelter, temporary shelter, temporary housing, and permanent housing (Quarantelli, 1995). A representative recovery timeline indicates that emergency sheltering is provided within the first 12–48 h, followed by temporary sheltering lasting from 2 to 30 days. This phase transitions to temporary housing, which may last from 3 months to 5 years, before culminating in the provision of permanent housing over the longer term (Montalbano and Santi, 2023; Gkoumas et al., 2025). As the temporary housing phase may extend over months or even years, it becomes a primary living environment rather than a short-term solution. Among the available solutions, container housing is widely adopted for its rapid deployment, modular construction, cost-effectiveness and simplicity of implementation, ease of transportation, and potential for reuse and scalability (Hong, 2017; Kim et al., 2025; Süalp and Coşgun, 2023). Container housing also offers advantages in terms of ownership flexibility and facilitates pre-disaster planning by authorities, accounting for the different characteristics of various disaster types (Zhang et al., 2014). Previous research has examined user-related aspects of container housing, including occupant satisfaction, social acceptability, and post-occupancy needs (Peternel and Podjed, 2026). However, its socio-cultural suitability remains limited, as such solutions often struggle to adequately address users’ cultural practices, social needs, everyday routines, and lived experiences (Félix et al., 2013). Its environmental and economic performance has also been investigated in relation to energy and carbon performance, thermal comfort, life-cycle assessment, and life-cycle costing (Atmaca and Atmaca, 2016; Atmaca, 2017; da Costa et al., 2023; Khaddour et al., 2024; Zafra et al., 2021).

Within this broader environmental performance context, daylight represents an important dimension of post-disaster housing sustainability. Effective use of daylight can reduce daytime dependence on artificial lighting and support operational energy efficiency. Beyond its energy-related benefits, daylight also influences circadian regulation, psychological wellbeing, and overall quality of life in residential environments (Javani et al., 2019; Nagare et al., 2021; Osibona et al., 2021). These benefits are particularly relevant in post-disaster housing, where units must be rapidly deployed while still providing energy-efficient, healthy, and habitable living environments for occupants who may remain there for extended periods.

Studies on daylight performance in container housing have largely focused on the scale of the individual unit. Schiavoni et al. (2017) evaluated the natural and artificial lighting performance of container modules proposed for emergency housing, as well as their energy and life-cycle performance. Noaman and El-Ghafour (2024) examined daylight factor and daylight uniformity as part of a passive-design approach to energy-efficient temporary container housing. Ibrahim et al. (2026) investigated the effects of external hybrid louver systems on daylight availability, glare, visual comfort, and energy performance in residential container units.

These studies demonstrate the importance of unit-level design parameters, including window configuration, glazing, and shading, in determining indoor daylight conditions. However, daylight performance is not determined solely by the architectural layout and envelope characteristics of individual units. It is also influenced by settlement-scale variables, including unit orientation, inter-unit spacing, mutual obstruction, and cluster geometry. The settlement spatial configuration shapes the overall quality of post-disaster residential environments by influencing not only daylight access and visual conditions but also privacy, personal space, accessibility, safety, territoriality, and opportunities for social interaction (Kim and Kim, 2020; Khozaei Ravari et al., 2022).

Post-disaster container housing settlements are commonly organized in linear orthogonal grid configurations, a pattern particularly evident in Türkiye, where earthquakes may require the rapid provision of large-scale temporary housing (Okutan and Müştak Sevindik, 2026). However, the spatial uniformity of such configurations may compromise spatial comfort and indoor daylight performance, thereby reducing the overall quality of living conditions.

Accordingly, this study adopts the linear orthogonal grid configuration as a reference and develops six alternative spatial configurations for post-disaster container housing settlements based on spatial organization principles intended to improve spatial comfort. It then comparatively evaluates the indoor daylight performance of all seven configurations through simulation-based analysis, with particular emphasis on daylight availability, sunlight exposure, and visual comfort. Istanbul was selected as the study context because its high seismic risk may necessitate the large-scale provision of temporary housing following a major earthquake. The significance of this study lies in positioning post-disaster settlement design within the broader framework of sustainability transitions. Ribeiro et al. (2025) emphasize that sustainability transitions are spatially embedded and shaped by local and regional conditions. Applying this perspective to post-disaster housing highlights the need to evaluate temporary settlements in relation to their specific climatic and spatial contexts. Accordingly, examining the effects of settlement configuration on daylight performance in Istanbul can contribute to the development of more context-responsive and sustainable post-disaster housing suited to the city’s specific conditions.

In line with the recommendations of EN 17037:2018 + A1:2021, advanced climate-based daylight metrics (CBDMs) are employed to provide a comprehensive assessment of daylight performance (European Committee for Standardization [CEN], 2021). In addition, glare caused by daylighting is also evaluated to assess visual discomfort caused by daylight. Accordingly, this study addresses the following research questions.

• How does settlement-scale spatial configuration influence the daylight performance of post-disaster container housing, as evaluated using climate-based daylight metrics and daylight glare indicators?
• Which spatial configuration provides the most balanced daylight environment by maximizing useful daylight while minimizing excessive sunlight exposure, glare, and associated visual discomfort?

2 Materials and methods

2.1 Development of the settlement spatial configurations

The post-disaster temporary container housing settlement configurations evaluated in this study were developed in accordance with the settlement planning principles outlined by AFAD (2015) and the Chamber of City Planners of the Union of Chambers of Turkish Engineers and Architects (TMMOB Şehir Plancıları Odası, 2023). The basic residential unit used in the study consisted of a 3.0 m × 7.0 m × 2.64 m container comprising a living room with an open kitchen counter, a bedroom, and a bathroom (Figure 1).

According to these guidelines, temporary settlements should be organized hierarchically into residential units, communities, and neighborhoods. Each community should comprise 16 container units served by shared sanitation facilities, while each container should provide an average enclosed area of 3.5–4.5 m² per person. Accordingly, each 21 m² container was assumed to accommodate five residents, providing 4.2 m² per person. A minimum distance of 2 m is required between individual units, while adjacent communities should be separated by roads at least 6 m wide. At the neighborhood scale, communities should be connected by secondary roads with a minimum width of 10 m, whereas primary circulation roads should be at least 15 m wide. Fire-safety corridors should be at least 30 m wide and provided at intervals not exceeding 300 m. The guidelines further recommend a maximum neighborhood population of 2,500 residents, a minimum land allocation of 45 m² per person, and an 8 m setback between the settlement boundary and the nearest residential unit. These planning principles established the common constraints for developing the seven post-disaster container-housing settlement configurations. Each settlement comprises multiple neighborhoods, with each neighborhood consisting of four communities and each community containing 16 identical container units. The configurations vary in unit orientation, mutual obstruction, façade exposure, courtyard geometry, cluster scale, and degree of open-space enclosure. These systematic variations provide a structured basis for evaluating how settlement morphology influences indoor daylight performance. The daylight assessment was therefore conducted at the community scale to account for interactions among the 16 units in each configuration. In this context, the seven configurations are treated as planning-constrained analytical scenarios rather than as an exhaustive or universally optimal set of post-disaster settlement solutions.

Three spatial organization categories, including linear, enclosed courtyard cluster, and an open courtyard cluster, provided the conceptual basis for developing the seven configurations. The linear category comprises Configuration A (Orthogonal Grid), Configuration B (Staggered), and Configuration C (Diagonal). In all three configurations, the units are arranged in a row, distinguished by regular alignment, lateral displacement, and diagonal orientation, respectively. Configuration A represents the conventional arrangement widely implemented in post-disaster container settlements. Configuration B was developed by laterally offsetting adjacent housing units within the conventional orthogonal arrangement. This modification was intended to reduce direct visual alignment between neighboring units, potentially enhancing privacy and creating a less rigid spatial organization. By interrupting the continuity of the rows, the staggered arrangement may also facilitate local airflow and wind-driven natural ventilation (Cheung and Liu, 2011), while retaining the linear structure. Configuration C rotates the units by 30° and introduces entrance-side front yards between the dwellings and the circulation routes. Unlike Configurations A and B, in which the units directly face the road, the diagonal arrangement creates a gradual transition between the private dwelling domain and the public circulation network. These semi-private outdoor spaces may support personalization, informal social interaction, territorial definition, natural surveillance, and sense of safety (Swapan et al., 2019; Kickert et al., 2025). They may also provide a buffer between the residential units and circulation routes, reducing direct exposure to traffic and creating more protected outdoor areas, particularly for children and other vulnerable users. Such spatial qualities may improve the residential experience and support community recovery in post-disaster settlements. Courtyard-type configurations are incorporated into the development of settlements because they combine private dwelling units with a shared communal open space and are widely recognized as an effective housing typology for both environmental and social performance. Previous studies have demonstrated that courtyards can enhance environmental performance through improved daylight access, natural ventilation, and solar control, with their effectiveness largely dependent on courtyard geometry and enclosure proportions (Diz-Mellado et al., 2023; Zamani et al., 2018). From a social perspective, courtyard housing supports social interaction, community cohesion, territoriality, and natural surveillance by providing a shared semi-private space that encourages daily encounters among residents while reinforcing a sense of ownership, safety, and belonging (Lee and Park, 2015; Kalfaoğlu Hati̇;Poğlu and Mohammad, 2021; Turner, 2022; Zhang, 2015). Within the enclosed courtyard cluster category, Configurations D–F form a systematic sequence based on the subdivision of the prescribed 16-unit community. Configuration D (Sixteen-Unit Enclosed Courtyard Cluster) organizes the entire community around a single central courtyard. Configuration E (Eight-Unit Enclosed Courtyard Cluster) divides the community into two smaller courtyard clusters, thereby reducing the scale of each courtyard and communal grouping. Configuration F (Four-Unit Enclosed Courtyard Cluster) further subdivides the community into four clusters of four units, producing the smallest courtyard scale within the sequence. These subdivisions were selected because they divide the 16 units into equal-sized groups and allow courtyard and cluster scale to vary systematically. In all three configurations, the unit entrances directly face their respective shared courtyards. Within the open courtyard cluster category, Configuration G (Pinwheel Open Courtyard Cluster) provides a contrasting spatial condition characterized by a pinwheel arrangement and mixed unit orientations. The units are organized to form L-shaped and U-shaped open courtyards, along with internal courtyards that are not directly faced by the unit entrances. Instead, the entrances are oriented primarily toward the surrounding circulation network, and the internal courtyards are accessed indirectly. Unlike Configurations D–F, in which the entrances directly address fully enclosed shared courtyards, the open courtyards in Configuration G remain visually and physically connected to the wider settlement. These partially enclosed spaces create semi-private outdoor areas associated with small groups of dwellings. Table 1 presents the seven spatial configurations developed for the study, grouped into linear, enclosed courtyard cluster, and open courtyard cluster typologies.

TABLE 1. Post-disaster container housing settlement and cluster layout typologies.

The detailed characteristics of the seven configurations are presented in Table 2. Configurations A and B represent the most compact linear conditions, occupying 560 and 596 m², respectively, without specifically designated resident-use open space. Configuration C is distinguished by its diagonal orientation and by providing a private front yard for each unit. The yards are organized in three sizes: 14.13, 21.07, and 29.47 m². They occupy 332 m², accounting for 33.0% of the community’s total area of 1,007 m². Configurations D–F form a systematic enclosed-courtyard sequence with broadly comparable total areas of 922–962 m², while varying in community subdivision and courtyard scale. In Configuration D, a single 455 m² courtyard, representing 49.3% of the community area, serves all 16 units and 80 residents. In Configuration E, the total courtyard area decreases to 406 m² (43.4%) and is divided between two eight-unit clusters, providing 203 m² for 40 residents in each cluster. Configuration F provides 308 m² of courtyard space (32.0%) across four four-unit clusters, with each 77 m² courtyard serving 20 residents. Configuration G represents the most spatially dispersed condition, occupying 1,188 m² and allocating 768 m² (64.6%) to shared open and internal courtyards. The two side internal courtyards, each measuring 56 m², serve 20 residents, while the 96 m² central courtyard serves 40 residents. The L-shaped and U-shaped open courtyards provide 63 m² and 77 m², respectively; because each is directly associated with only two container units, each courtyard is assumed to serve 10 residents. The configurations demonstrate a deliberate progression from compact, land-efficient layouts toward more resident-oriented spatial organizations. While Configurations A and B provide no designated resident-use open space, Configurations D–F introduce enclosed courtyards at different community scales, creating varying degrees of enclosure, accessibility, and interaction among smaller resident groups. Configuration G provides the most differentiated open-space hierarchy by combining courtyards serving small groups of units with larger shared open areas. These variations indicate that spatial comfort is shaped by both the total amount of open space and its subdivision, accessibility, degree of enclosure, and the number of residents it serves.

TABLE 2. Detailed characteristics of the seven spatial configurations.

Area values were calculated from the digital models and rounded to the nearest whole square meter.

Total area within the outer boundary of the 16-unit arrangement, excluding surrounding roads.

The total area excludes the area occupied by the containers [16 × (3 m × 7 m) = 336 m²] and the resident-use open-space area.

2.2 Simulation and 3D model settings

Daylight simulations were performed using Radiance-based ClimateStudio (version 2.3), an environmental performance plug-in for Rhinoceros 3D (Rhino), to evaluate the daylight performance of the proposed post-disaster container housing settlement configurations (Solemma LLC, 2025). Seven settlement layouts were modeled in Rhino based on the spatial configurations described in Section 2.1. All settlement layouts were modeled with identical building dimensions and construction properties, ensuring that only the settlement layout varied among the simulation scenarios. Annual climate-based daylight simulations were conducted using the Typical Meteorological Year (TMY) weather file for Istanbul, Türkiye (Climate.OneBuilding.org, 2026). The climatic characteristics of the simulation location are presented in Figure 2, including (a) monthly temperature and solar radiation profiles, (b) annual thermal stress conditions represented by the UTCI heatmap, and (c) the annual sun-path diagram. As shown in Figure 2, solar radiation levels are highest during the summer months, while seasonal variations in solar altitude and solar trajectories influence daylight availability throughout the year.

All simulations were performed using a Radiance-based annual daylighting workflow to ensure consistency and comparability between scenarios. For daylight analysis, Radiance simulation parameters were defined with six ambient bounces (ab), 4096 ambient samples (ad), and a weight limit (lw) of 0.01, based on the default daylight simulation configuration implemented in ClimateStudio v2.3.

A sensor grid consisting of 373 points was generated in accordance with EN 17037:2018 + A1:2021. Daylight performance was evaluated separately for the two principal occupied zones of each container unit: the living room, coded as Zone 1, and the bedroom, coded as Zone 2. The grid comprised 204 sensor points in Zone 1 and 169 sensor points in Zone 2, with a uniform spacing of 0.2 m and a work-plane height of 0.85 m above the finished floor level (CEN, 2021). Daylight glare analysis was performed in ClimateStudio v2.3 using the Spatial Disturbing Glare (sDG) workflow. DGP was calculated at viewpoints positioned 1.2 m above the finished floor level, representing the eye height of a seated occupant, for eight equally spaced viewing directions. A total of 53 glare-analysis points were considered, including 28 points in Zone 1 and 25 points in Zone 2. Figure 3 illustrates the daylight and glare analysis grids, while Table 3 presents the Zone 1 and Zone 2 coding applied to the seven spatial configurations.

TABLE 3. Daylight performance evaluation zones and coding within the container unit across the evaluated post-disaster housing configurations (A–G).

For the annual daylight simulations, an occupancy schedule of 08:00–18:00 was adopted throughout the year, with an occupancy value of 1.0 during occupied hours, including weekends. This schedule was applied consistently across all simulation scenarios. Although the investigated building represents a post-disaster housing unit, such shelters frequently serve as multifunctional living spaces where occupants carry out daytime activities including working, studying, administrative tasks, and other daily routines in addition to residential use. Therefore, the selected occupancy schedule was considered for evaluating daylight performance during the period when visual comfort and daylight availability are most critical and to ensure a consistent comparison among the proposed design alternatives. Figure 4 gives the graphical expression of occupancy schedule adopted for the annual daylight simulations in ClimateStudio.

The optical properties assigned in ClimateStudio were defined using Radiance material parameters. Material assignments were applied to all modeled surfaces, including walls, roofs, glazing, and exterior ground surfaces. Optical properties were specified according to the selected material characteristics, and the corresponding spectral power distribution (SPD) data were assigned where applicable. Interior surface reflectances were set to 50% for walls, 82% for the ceiling, and 35% for the floor. Window glazing was modeled as double-glazed with a visible light transmittance (VLT) of 80%. The exterior ground surface was assigned a reflectance of 18.1% representing concrete exterior floor. The material specifications and optical properties adopted in the simulations are presented in Table 4.

TABLE 4. Material properties and spectral power distribution (SPD) characteristics of the simulated surfaces.

2.3 Evaluation criteria

In this study, advanced climate-based daylight metrics (CBDMs) are employed to provide a comprehensive assessment of daylight performance. Specifically, spatial daylight autonomy (sDA), useful daylight illuminance (UDI), annual sunlight exposure (ASE), and mean illuminance are evaluated to assess daylight availability and sufficiency, useful daylight conditions, excessive illuminance, and direct solar exposure within the evaluated container housing units. Spatial Daylight Autonomy (sDA300,50%) represents the percentage of the analyzed area receiving a minimum daylight illuminance of 300 lx for at least 50% of the annual occupied hours and is widely used as an indicator of daylight sufficiency and daylight availability (Illuminating Engineering Society of North America [IES], 2012). Higher sDA values indicate greater access to useful daylight throughout the year. Useful Daylight Illuminance (UDI) evaluates the annual occurrence of daylight illuminance levels within defined ranges. Following Mardaljevic et al. (2012), daylight levels were classified into four ranges: UDI fell-short (UDI-f; <100 lx), UDI supplementary (UDI-s; 100–300 lx), UDI autonomous (UDI-a; 300–3000 lx), and UDI exceeded (UDI-e; >3000 lx). A high occurrence of UDI-f indicates insufficient daylight availability, whereas a high occurrence of UDI-e may indicate overexposure and an increased risk of visual discomfort. Annual Sunlight Exposure (ASE1000,250) quantifies the percentage of the analyzed area receiving direct sunlight illuminance greater than 1000 lx for more than 250 occupied hours per year (Illuminating Engineering Society of North America, 2012). ASE is commonly used to identify areas susceptible to excessive direct sunlight and potential visual discomfort. Daylight glare performance was assessed annually using the Daylight Glare Probability (DGP) metric. Following the revised thresholds proposed by Wienold et al. (2019), glare conditions were classified as imperceptible (DGP < 0.34), perceptible (0.34 ≤ DGP < 0.38), disturbing (0.38 ≤ DGP < 0.45), and intolerable (DGP ≥ 0.45). Based on these annual evaluations, sDG was calculated as the percentage of occupied floor area with DGP values exceeding 0.38 for at least 5% of the occupied hours. These metrics provide a comprehensive evaluation of the daylighting characteristics of alternative post-disaster settlement configurations and enable a comparative assessment of the influence of spatial configuration on daylight performance.

3 Results

3.1 Daylight performance metrics across the evaluated spatial configurations

Table 5 presents the daylight performance of the seven spatial configurations in Zone 1 (living room) and Zone 2 (bedroom). All configurations achieved consistently high daylight availability. sDA values ranged from 99.9% to 100.0% in living rooms and from 99.8% to 100.0% in bedrooms. The limited variations of 0.1 and 0.2 percentage points, respectively, indicate that sDA provided little differentiation among the evaluated spatial configurations.

TABLE 5. Daylight performance metrics for Zones 1 and 2 across the evaluated spatial configurations (A–G).

Zone 1 refers to the living room; Zone 2 refers to the bedroom.

UDI-f, fell-short (<100 lx); UDI-s, supplementary (100–300 lx); UDI-a, autonomous (300–3000 lx); UDI-e, exceeded (>3000 lx).

UDI-a was the dominant daylight category across all configurations. Within the linear category, UDI-a ranged from 74.2% to 75.1% in living rooms and from 78.9% to 79.5% in bedrooms. The enclosed courtyard configurations recorded very similar UDI-a values, ranging from 75.7% to 75.8% in living rooms and from 78.2% to 78.5% in bedrooms. The open courtyard cluster recorded UDI-a values of 75.1% in living rooms and 78.9% in bedrooms. These narrow ranges indicate limited numerical variation in UDI-a among the evaluated configurations.

UDI-f and UDI-s accounted for only a limited proportion of occupied hours across all configurations. UDI-f ranged from 2.8% to 3.2% in living rooms and from 3.6% to 4.2% in bedrooms, indicating that illuminance levels below 100 lx occurred infrequently. UDI-s ranged from 5.2% to 6.3% in living rooms and from 7.6% to 9.2% in bedrooms, showing that daylight levels within the supplementary range of 100–300 lx were also relatively uncommon, consistent with UDI-a accounting for most occupied hours.

UDI-e also varied among the evaluated configurations. Within the linear category, UDI-e ranged from 16.9% to 17.7% in living rooms and from 9.2% to 9.9% in bedrooms, with the diagonal configuration recording the highest value in both zones. The enclosed courtyard configurations generally recorded the lowest UDI-e values, ranging from 14.9% to 15.1% in living rooms and remaining at 8.4% in bedrooms. Configuration G, representing the open courtyard cluster, recorded values of 16.5% in living rooms and 9.1% in bedrooms. In all configurations, UDI-e was consistently higher in living rooms than in bedrooms, indicating that illuminance levels above 3000 lx occurred during a greater proportion of occupied hours in living rooms.

ASE was consistently higher in living rooms than in bedrooms, ranging from 33.6% to 38.7% and from 26.0% to 29.2%, respectively. Within the linear category, Configurations A and B recorded the lowest living room ASE, both at 33.6%, while Configuration C recorded the highest value at 35.8%. In bedrooms, Configurations A and B likewise recorded the lowest ASE, both at 26.0%, whereas Configuration C recorded the highest value at 27.6%. Differences among the linear configurations were slightly more pronounced in living rooms than in bedrooms. Within the enclosed courtyard category, ASE increased progressively across Configurations D, E, and F, from 34.5% to 35.4% and 36.6% in living rooms and from 26.6% to 27.1% and 27.8% in bedrooms, respectively. Configuration G recorded the highest ASE values among all evaluated configurations, reaching 38.7% in living rooms and 29.2% in bedrooms, corresponding to the largest proportion of analyzed area exceeding the ASE criterion.

Mean illuminance was consistently higher in living rooms than in bedrooms, ranging from 2926.7 to 3135.5 lx and from 2092.2 to 2260.5 lx, respectively. Among the linear configurations, Configuration C recorded the highest mean illuminance in living rooms at 3135.5 lx, whereas Configuration B recorded the lowest at 3023.4 lx, followed by Configuration A at 3037.0 lx. In bedrooms, Configuration C recorded the highest mean illuminance at 2260.5 lx, whereas Configuration A recorded the lowest among the linear configurations at 2168.5 lx, followed by Configuration B at 2185.8 lx. Within the enclosed courtyard category, mean illuminance increased slightly across Configurations D, E, and F, from 2926.7 to 2937.8 and 2954.8 lx in living rooms and from 2092.2 to 2102.6 and 2115.5 lx in bedrooms, respectively. Configuration G recorded mean illuminance values of 3067.6 lx in living rooms and 2206.1 lx in bedrooms but did not record the highest value in either room type.

sDG was consistently higher in living rooms than in bedrooms, ranging from 33.26% to 39.51% and from 21.06% to 27.59%, respectively. Within the linear category, Configuration B recorded the lowest living-room sDG at 34.21%, followed closely by Configuration A at 34.68%, while Configuration C recorded the highest value at 38.81%. In bedrooms, Configuration B recorded the lowest sDG at 24.61%, followed by Configuration A at 24.63%, while Configuration C recorded the highest value at 26.94%. Within the enclosed courtyard category, sDG increased progressively across Configurations D, E, and F, rising from 33.26% to 34.24% and 35.83% in living rooms and from 21.06% to 21.53% and 22.59% in bedrooms, respectively. Configuration G recorded the highest sDG values among all evaluated configurations in both living rooms (39.51%) and bedrooms (27.59%), representing the largest proportion of analyzed floor area with DGP values exceeding 0.38 for at least 5% of the occupied hours.

3.2 Spatial distribution of UDI, ASE, and mean illuminance

The spatial distributions of UDI, ASE, and mean illuminance across the seven evaluated configurations are presented in Tables 6–8. Across all configurations, UDI-a was the predominant daylight category in both living rooms and bedrooms. In contrast, the ASE maps showed clearer spatial differences among the configurations, with areas exceeding the ASE criterion concentrated primarily near façade openings. The mean illuminance maps showed higher illuminance levels near the openings, which decreased toward the deeper portions of the rooms.

TABLE 6. Spatial distribution of UDI across the evaluated spatial configurations (A–G).

TABLE 7. Spatial distribution of ASE across the evaluated spatial configurations (A–G).

TABLE 8. Spatial distribution of Mean Illuminance across the evaluated spatial configurations (A–G).

Within the linear category, Configurations A and B exhibited broadly similar spatial patterns, whereas Configuration C showed more extensive areas exceeding the ASE criterion and higher illuminance levels, particularly in living rooms. Within the enclosed courtyard category, Configurations D–F displayed comparable UDI-a patterns. However, the extent of areas exceeding the ASE criterion increased progressively across Configurations D–F. Their mean illuminance distributions showed a similar decrease from the façades toward the room interiors, with differences occurring primarily in the extent of areas with higher illuminance levels. Configuration G, representing the open courtyard cluster category, exhibited a visually more heterogeneous spatial pattern among its individual units, with the location and extent of areas exceeding the ASE criterion and areas with higher illuminance levels varying among the units. The maps indicate that differences among the configurations were more evident in the extent and spatial distribution of areas exceeding the ASE criterion and in mean illuminance patterns than in UDI-a distributions.

3.3 Association between excessive daylight illuminance and spatial disturbing glare

In order to examine the unit-level association between excessive daylight illuminance and spatial disturbing glare, UDI-e and sDG values were paired for the same container unit and analyzed separately for Zone 1 (living room) and Zone 2 (bedroom). Each of the seven spatial configurations comprised 16 units, resulting in 112 paired observations per zone. Spearman rank-order correlation was used because the unit-level data exhibited clustered distributions and included tied values, and the analysis focused on monotonic rather than necessarily linear associations. To determine whether the pooled association remained after accounting for differences among Configurations A–G, a configuration-adjusted rank correlation was also calculated. Ranked UDI-e and sDG values were separately regressed on indicator variables representing the seven configurations, and the resulting residuals were correlated.

Strong positive associations were identified between UDI-e and sDG in both zones. In living rooms, the Spearman correlation coefficient was ρ = 0.916 n = 112. In bedrooms, the corresponding coefficient was ρ = 0.929 n = 112. The associations remained strong after accounting for spatial configuration, with adjusted rank-correlation coefficients of 0.923 for Zone 1 and 0.938 for Zone 2. The persistence of these associations indicates that the observed relationship was not attributable solely to differences in average daylight conditions among the seven configurations.

As illustrated in Figure 5, container units with a greater proportion of occupied hours exceeding 3000 lx generally also exhibited a greater proportion of occupied floor area exposed to disturbing or intolerable glare. UDI-e and sDG therefore tended to increase together within the evaluated configurations. However, the coefficients describe associations within the finite set of simulated units and should not be interpreted as evidence that higher UDI-e independently causes higher sDG or that UDI-e can substitute for direct glare assessment. The findings are consequently limited to the seven evaluated spatial configurations and are not generalized to all possible post-disaster container-housing settlements.

4 Discussion

The findings demonstrate that settlement-scale spatial configuration influenced the control and spatial distribution of daylight more strongly than its overall availability. All seven configurations achieved near-complete spatial daylight autonomy, while UDI-a varied only slightly. Under the standardized container design and Istanbul climatic conditions, insufficient daylight was therefore not the principal concern. Instead, the configurations were differentiated primarily by excessive illuminance, annual sunlight exposure, and spatial disturbing glare, indicating that glazed-façade orientation, mutual obstruction, courtyard enclosure, and community subdivision influenced whether abundant daylight remained useful or became a source of visual discomfort. This configuration-level variation was accompanied by a consistent difference between the two functional zones. Living rooms were generally more exposed than bedrooms, highlighting the importance of zone-specific evaluation. An average calculated for the entire unit would have concealed these differences in excessive daylight exposure and glare.

4.1 Daylight implications of linear configurations

Configurations A and B produced closely comparable daylight conditions. This similarity is consistent with their identical north–south distribution of glazed façades and comparable compact linear forms; the lateral displacement in Configuration B did not produce a substantial change in daylight exposure or mutual obstruction. Under the evaluated conditions, staggering therefore had limited daylight implications, although it may provide spatial benefits beyond the scope of the daylight analysis. Neither configuration provided designated resident-use open space. Configuration C exhibited a more exposed daylight condition. This outcome cannot be attributed to the 30° rotation alone, because the diagonal arrangement also reoriented the glazed façades toward the northwest and southeast, expanded the community footprint, altered inter-unit alignment and obstruction, and introduced 332 m² of semi-private front yards, representing 33.0% of the community area. These combined changes coincided with higher excessive illuminance, ASE, mean illuminance, and sDG than in Configurations A and B. Because daylight sufficiency was already near complete in the two compact linear configurations, the additional exposure did not meaningfully improve daylight availability. Configuration C therefore represents a trade-off between potential outdoor spatial benefits and less controlled indoor daylight conditions.

4.2 Daylight implications of enclosed courtyard cluster configurations

The enclosed courtyard cluster configurations recorded the lowest UDI-e values among the evaluated categories, although their performance varied with community subdivision, façade orientation, courtyard geometry, and mutual obstruction. Configuration D provided the most balanced outcome, combining near complete sDA and high UDI-a with the lowest sDG in both zones and a 455 m² courtyard, representing 49.3% of the community area and directly faced by the unit entrances. Its comparatively controlled performance is consistent with the predominance of north- and south-facing glazed façades and the organization of all 16 units around a single enclosed courtyard. Because Configurations A and B recorded slightly lower ASE, however, Configuration D should be understood as offering the strongest overall balance rather than the best performance for every individual metric. From a settlement planning perspective, these findings suggest that spatial configurations should not be optimized according to a single daylight performance indicator. Instead, planning decisions should balance daylight sufficiency, visual comfort, communal open-space provision, and land-use efficiency. The results indicate that the most appropriate settlement layouts are those that maintain useful daylight while limiting excessive sunlight exposure and glare without compromising the provision of communal open space, rather than those that maximize the performance of any single daylight metric. Across Configurations D to F, ASE, mean illuminance, and sDG increased progressively as the community was divided into a greater number of smaller courtyard clusters, whereas UDI-e remained nearly unchanged. This divergence indicates that the differences among these configurations were more evident in prolonged direct sunlight exposure and glare than in the annual occurrence of illuminance above 3000 lx. Because façade orientation, courtyard dimensions, cluster geometry, local access area, and mutual obstruction relationships changed simultaneously, the trend cannot be attributed to courtyard subdivision alone. The comparatively balanced performance of Configuration D is therefore interpreted as resulting from the combined influence of these interacting spatial characteristics rather than from any single design parameter. In particular, the predominance of north–south-facing glazed façades, the organization of all units around a single enclosed courtyard, and the resulting obstruction relationships appear to have moderated excessive solar exposure while maintaining high daylight availability. Although the present study does not isolate the individual contribution of each variable, their combined influence provides a plausible explanation for the observed performance balance. Configurations E and F may offer organizational benefits through smaller communal groups and more localized courtyards. However, under the evaluated geometries, these potential spatial advantages coincided with less favorable direct sunlight and glare conditions than in Configuration D. Courtyard subdivision should therefore be coordinated with façade orientation, obstruction relationships, and solar control strategies rather than treated as an isolated design variable.

Accordingly, settlement planning should consider courtyard subdivision, façade orientation, courtyard enclosure, and unit arrangement as interdependent design decisions rather than independent planning variables. Evaluating these parameters collectively during the early planning stages may help achieve a more balanced daylight environment while maintaining the spatial qualities required for temporary post-disaster settlements.

4.3 Daylight implications of open courtyard cluster configuration

Configuration G recorded the highest ASE and sDG values and exhibited the most heterogeneous daylight patterns among individual units. Its pinwheel arrangement distributed glazed façades equally across the four cardinal orientations and combined open and internal courtyards with varying degrees of enclosure and accessibility. Consequently, individual units experienced different relationships with surrounding open space, neighboring containers, direct sunlight, and mutual obstruction.

The performance of Configuration G cannot be explained by orientation distribution alone. Configuration F had the same proportions of north, south, east, and west facing glazed façades but recorded lower ASE and sDG. This comparison indicates that unit arrangement, courtyard enclosure, and obstruction relationships were also relevant. Moreover, Configuration G did not record the highest UDI-e or mean illuminance despite producing the highest ASE and sDG. Its distinctive daylight response therefore related more to the distribution and duration of direct sunlight and visual exposure than to the overall quantity of daylight. Configuration G provided 768 m² of resident-use open space, representing 64.6% of the community area and the largest provision among the evaluated alternatives. Compared with Configuration D, it offered a larger and more differentiated open space system but also greater variation in sunlight exposure and glare among the units. This comparison demonstrates that the daylight performance associated with resident use open space depends on its spatial organization and relationship to glazed façades, rather than on its area alone.

5 Conclusion

This study demonstrates that settlement configuration plays an important role in regulating daylight in post-disaster container housing, even when daylight availability is consistently high. Among the evaluated alternatives, Configuration D, which arranged the 16 units required for each community around a single enclosed courtyard, provided the most balanced outcome by combining useful daylight with comparatively controlled excessive illuminance, direct sunlight exposure, and glare. It also combined comparatively controlled indoor daylight conditions with a substantial and directly accessible communal courtyard, indicating that daylight regulation and communal open-space provision can be integrated at the settlement scale. The findings highlight daylight-responsive settlement planning as a relevant component of sustainable post-disaster container housing. Maintaining useful daylight while limiting excessive solar exposure and glare may support more visually comfortable indoor conditions for disaster-affected communities and reduce reliance on artificial lighting during occupied hours.

Although daylight performance represents only one aspect of post-disaster settlement planning, the findings demonstrate that settlement-scale spatial configuration can contribute to improving indoor visual comfort while supporting the provision of accessible communal open spaces. Accordingly, daylight-responsive spatial planning should be considered alongside broader planning priorities, including rapid deployment, construction cost, land availability, infrastructure provision, emergency accessibility, and social needs, when developing temporary post-disaster settlements.

The findings should, however, be interpreted in light of several limitations. The study was based on a standardized container unit under Istanbul climatic conditions, fixed material properties and occupancy assumptions, and seven predefined settlement configurations developed in accordance with existing planning guidelines. Furthermore, artificial lighting energy use and the potential social and cultural benefits of communal open spaces were not evaluated, and several spatial characteristics changed simultaneously, preventing the individual contribution of each design variable from being isolated.

Future research should therefore investigate orientation, spacing, courtyard geometry, and obstruction relationships using parametric approaches across different climatic conditions. Further studies should also examine whether the comparative performance patterns identified in this study remain consistent when façade-level design variables, including glazing transmittance, window size, and external shading devices, are evaluated together with settlement-scale spatial configuration. Integrating artificial lighting energy analysis with occupant-based evaluations of communal open spaces would provide a more comprehensive understanding of the interaction between settlement morphology, building-envelope design, and overall post-disaster housing performance.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI was used for grammar and language editing.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Pallav Dutta, Aliah University, India

Reviewed by: Pau Chung Leng, University of Technology Malaysia, Malaysia

Faris Ali Mustafa, Salahaddin University, Iraq

Hélder Lopes, University of Minho, Portugal

Merve Tuna Kayili, Karabuk Universitesi, Türkiye