Impact of glazed balcony design on daylight in Finnish apartments

2.3.1 Daylight factor (DF)

The DF is a common method (Figure 1) describing how much ambient light enters a room (Carlucci et al. 2015) as a proportion of the available outside natural light. It is the proportion of illuminance inside (E_i) to the unobstructed illuminance outside (E_o) at a certain point on a measuring plane, and expressed as a percentage. It includes sunlight’s diffuse component as well as reflected light both internally and externally, and was developed to avoid dependency on dynamic fluctuations of sky conditions (Reinhart et al. 2006). As such, it is measured or calculated with a standard overcast sky—the International Commission on Illumination’s (CIE) Standard Overcast Sky Model—and disregards building location or orientation. Using averaging rules, the minimum DF requirement in housing is 1.5–2.0%, although 5% is considered good for comfort (Pelsmakers 2015). UK national guidelines set room-specific daylight targets, prioritising living spaces, then kitchens and bedrooms (Littlefair et al. 2024).

Various authors suggest that DF has notable limitations (Carlucci et al. 2015; Mardaljevic et al. 2009; Vikberg et al. 2020). First, as it cannot represent non-overcast skies, the DF significantly differs from actual daylight conditions, though in Northern Europe skies are often cloudy. Additionally, expressed as a percentage, it does not account for absolute values at a given time, season or location. Also, aiming to maximise DF, the use of larger glazed areas may be encouraged, which could negatively impact both energy efficiency and IEQ aspects such as thermal comfort. Although based on simplified assumptions, DF remains the most widely used quantitative measure for daylighting (Reinhart et al. 2006) and ‘a good conceptual parameter, a measure of a building’s transparency to the sky’ (Baker & Steemers 2019: 132). Generally, the use of DF enables a consistent evaluation of the architectural impacts on indoor daylight provision and is particularly valuable for comparing in-situ measurements conducted in different homes.

2.3.2 Absolute illuminance levels using climate-based metrics

Due to the limiting factors associated with DF, actual illuminance levels (Figure 1) are considered a more realistic approach for calculations (Reinhart et al. 2006; Vikberg et al. 2020). Yet, their reliability without model validation remains uncertain. This is made possible using standardised climate models in validated software, which is referred to as climate-based daylight modelling (CBDM). Based on visual comfort requirements, 300 lx is commonly used as a quantitative daylight requirement to be achieved over a certain area for half of the daylight hours (CEN 2018).

CBDM uses different metrics to evaluate daylight performance (Figure 1) which differ in scope and thresholds, such as daylight autonomy (DA), spatial daylight autonomy (sDA) and useful daylight illuminance (UDI) (Reinhart et al. 2006). DA measures the percentage of occupied time—between 08.00 and 20.00 hours—during which a specific point receives sufficient daylight. sDA extends DA to a spatial level and measures the percentage of floor area that meet the DA autonomy criterion. UDI, on the other hand, evaluates the quality of daylight by categorising the horizontal daylight intensity into certain ranges, i.e. too dark (< 100 lx), useful illuminance (100–2000 lx) and too bright (> 2000 lx). Originally developed for workspaces, the upper category, referred to as UDI_e, assesses visual discomfort and overheating from overexposure. In housing, occupancy patterns and adaptive solutions often mitigate heat gains and overexposure. While the suitability of these thresholds for housing remains uncertain, UDI provides a nuanced assessment of daylight availability and potential discomfort, enabling meaningful design comparisons to be made.

3.1.1 Case studies

A total of 10 apartments, pseudonymised as CS1–CS10 (Table 1 and Figure 2), served as purposive selected case studies to evaluate the impact of balcony design and balcony glazing on daylight availability.

Figure 2

The apartments were selected for their diversity of glazed balconies and located in four main urban centres to ensure geographical representation, while ensuring practical accessibility for data collection (Turku, Helsinki, Tampere and Oulu, part of the continental climate group D) (Figure 3). Access was enabled through voluntary resident participation, facilitated by architects of suitable projects who connected researchers with housing companies to distribute invitations. Privacy notices about the research and informed consent forms were shared with participants before the study.

Figure 3

Two types of balcony glazing systems were featured in the case studies. First, a frameless and ventilated design with retractable glass panels that slide and turn to fully open and fold, usually against the wall (CS1–CS5, CS9, CS10). Panels are attached via upper and lower fixings with approximately 5 mm vertical gaps between glass panes. Second, a framed sliding-only design with no built-in ventilation gaps (CS6–CS8). Both systems are uninsulated with single toughened glass.

3.1.2 DF in-situ measurements

In-situ measurements were carried out for all 10 case studies four times during one full year, at similar times of the day (December 2024, March, June and September 2025). This frequency was necessary to ensure that measurements were conducted under appropriate weather conditions, specifically during fully overcast days, corresponding to the measurements presented in the study. Therefore, the data reflect different times of the year, which reinforces the relevance of using DF as a comparative metric.

For each case study, a 0.5 m measurement grid was drawn out over the entire reference plane to be measured, i.e. the apartment as well the balcony space, starting from the main windows and 0.5 m from the walls. Each point was named and marked on the floor using masking tape; points were named and noted in a table and the grid was also drawn out on a floor plan for replicability. During each visit, the illuminance was measured with a luxmeter Trotec BF06 (accuracy: ±5% ± 10 lx for < 10,000 lx; ±10% ± 10 lx for > 10,000 lx), placed 1 m above the floor at each point by a researcher. Simultaneously, another researcher measured outdoor lux levels in an unobstructed area outdoors. During measurements, all artificial lights were turned off and the blinds and curtains opened. The readings were repeated twice: once with balcony glazing fully closed and once fully opened. Based on these measurements, DF was calculated for each point and represented in a DF map (Figures 4 and 6; and see Appendix A in the supplemental data online).

Figure 4

3.1.3 Semi-structured interviews

Data obtained from in-situ measurements were contextualised with qualitative insights from semi-structured interviews (Andalib 2024; Vikberg et al. 2022), a common post-occupancy evaluation method for assessing occupants’ visual comfort (Elsayed et al. 2023). This followed the quantitative analysis to explore how residents’ experiences aligned—or differed—with the findings, strengthening interpretation in support of user-informed design principles. Residents were interviewed about balcony use, visual and thermal comfort, social and personal impacts, and seasonal changes. This paper focuses only on findings related to daylight, primarily based on two interview questions:

• How does your balcony affect the amount of daylight in the adjacent living space?

• How does the amount of daylight on your balcony affect your use of it?

Participants from 10 households were encouraged to reflect on seasonal daylight variations and related comfort. Interviews were recorded, transcribed and translated from Finnish to English by the first author. Responses were used to annotate and interpret the measurement results, providing a context for the observed trends. For further details of the interviews, see Appendix C in the supplemental data online.

3.2.1 Parametric modelling in IDA-ICE

The case study was modelled in IDA-ICE 5.1, a dynamic simulation tool validated for assessing indoor climate and energy performance (Hilliaho et al. 2015). Though not designed for glazed environments, it remains a viable option using modelling workarounds. At the time of the study, no validation report for the daylight add-on was available; however, the software relies on the Radiance engine, widely used in industry and scientifically validated (Mardaljevic 2000). For all base model input data, see Table S6 in Appendix B in the supplemental data online.

IDA-ICE 5.1 uses 2012 reference-year climate data per the Finnish Building Code. The localities used for this dataset are Sodankylä, Jyväskylä and Helsinki (Figure 3), representative of different regions and climates of Finland. The base model was initially matched to Helsinki as the most representative locality.

3.2.2 Comparative analysis

To evaluate daylight availability, visual and thermal comfort, this study adapts Hu’s et al. (2023) parametric approach for non-residential buildings to housing design, integrating the sDA and UDI_e metrics. By combining these metrics, the analysis aims to identify scenarios that maximise useful daylight while minimising overexposure across all seasons. Results were also analysed against the CEN (2018) criteria to verify compliance.

Regarding window sizing, Lehtinen et al. (2022) indicate that windows behind balconies should be considerably larger than those without shading. For the case study, the width is assumed to be at its maximum, and window height has been extended to 2.3 m. Other features—primarily light transmittance—known to influence daylight availability and overheating are acknowledged but remain outside of the scope of this study.

A measuring plane consisting of 17 points was created in the living space similarly as described in in-situ measurements (Figure 5). While IDA-ICE 5.1 typically allows a sensitivity analysis of balcony depth, this is not possible for glazed balconies due to zone modelling limitations. Therefore, depth was modified in separate base models and the results analysed in Excel. Repeated daylight simulations were performed, making successive changes to models based on defined parameters, i.e. balcony depth (1.5, 1.8 and 2.1 m) in relation to orientation (north, west, south and east), location (Helsinki, Jyväskylä and Sodankylä) and date stamp (21 December 2024; 21 March, 21 June and 21 September 2025), a total of 144 calculations. Tested balcony depth variations were based on commonly observed dimensions in the Finnish building stock and the case studies. A minimum depth of 1.5 m was selected as previous research (Jegard et al. 2024) has shown that sufficient balcony depth is a key factor for balcony use.

4.1.1 Overall daylight conditions: impacts of balcony structure

The results described below are presented in Figure 6 and in Appendix A and Table S1 in Appendix B in the supplemental data online. In this section, the analysis focuses on the worst-case scenario, i.e. the glazing being closed. Comparisons can offer valuable insights into overall daylight availability and spatial prioritisation of daylight within the case study apartments. However, in the comparative analysis of average DF between rooms, observed differences cannot be solely attributed to balcony presence as variations in room dimensions and window configurations also contribute significantly.

Figure 6

For rooms behind a balcony, a 2% average DF recommendation is reached for three of the 10 case studies (CS1 = 2.28%, CS2 = 2.58%, CS9 = 2.53%), but all others are under a 1.5% minimum recommendation. All the case studies are largely under the 5% recommended by Pelsmakers (2015) for good daylighting. By comparison, measured average DF in rooms without a balcony shading them and on the same facade ranged from 1.78% to 6.45% (CS3 = 4.02%, CS4 = 2.67%, CS5 = 6.45%, CS9 = 6.29%, CS10 = 1.78%). Only one case study showcased DF under 2% in these rooms (CS10), while all others were above 3.4%.

Overall, only CS1 and CS2 showcased overall good daylight provision (average DF = 2.28–4.92% and 2.58–4.17% across rooms, respectively), benefiting from a shallow plan depth and dual-orientation windows, one of which was unobstructed by a balcony (CS1) and large windows as well as high ceilings (CS2). Although CS1 was effective for good daylight provision, a facade is directly overlooking the building on the western side and blinds were kept closed to ensure privacy. Additionally, although CS2 demonstrates good daylight availability, residents’ feedback revealed that it suffered from significant overheating in summer. CS6 and CS7 were perceived by residents as having excellent daylight quality, largely due to floor-to-ceiling windows (Table 1) and dual orientations, despite having deep plans and low average DF for rooms behind the balcony (CS6 = 0.68%, CS7 = 0.29–0.33%). CS3–CS5 showed moderate to poor daylight provision, suffering from a deep layout but benefiting from a non-continuous balcony facade (average DF = 1.33–4.02%, 0.46–2.67% and 0.81–6.45% across rooms, respectively). Similarly, CS9 and CS10 exhibit deep plans depth (9–10 m) with single orientation and consequently poor daylighting conditions (average DF = 1.13–6.29% and 0.78–1.85% across rooms, respectively). Notably, CS9’s balcony was valued by residents as a usable extension of the living space, despite the negative impact on daylight. CS10, while featuring large openings, experienced severe overheating in summer and limited ventilation options, according to the residents. This was the only case study without integrated window blinds (Table 1), which may have exacerbated the issue. All residents reported that during summer, when direct sunlight enters the balcony, they avoided using it because of excessive heat, as reported in earlier research (Jegard et al. 2024). For CS3, when direct sunlight did not enter the apartment directly, at times sunlight still reached the interior through reflections from the opposite building’s glazed balconies. According to the resident, this reflected light not only partly compensated for the lack of direct sunlight but also created a pleasant effect, especially during winter.

For most case studies, excluding CS9, rooms behind a balcony had largely deeper plan depth than rooms without, resulting in a much smaller window-to-floor ratio already affecting average DF negatively. In this case, the addition of a balcony structure will have a bigger impact on average DF as well as overall daylight conditions. What is more, these rooms are predominantly living spaces which results in compounded disadvantages in the spaces where daylight is most needed. Specifically, the kitchen area—identified by UK standards as requiring a higher daylight level than bedrooms (Littlefair et al. 2024)—is the most negatively affected (CS3, CS6–CS10). In CS3, the resident observed a clear difference in daylight availability between the bedroom and living room, the latter being noticeably better. Despite this, the difference was considered acceptable by the current occupant.

Among the case studies, most followed the daylight penetration rule, except CS9 and CS10 which had a building depth of over 10 and 9 m, respectively. That said, the balcony is not accounted for in this calculation as suggested by RT-Kortistot (2023). Doing so would easily showcase the poor daylight zone left unaccounted for.

Across all cases, balcony depth—i.e. facade to outer edge distance—emerges as a critical factor. Balconies deeper than 1.8 m consistently obstruct daylight, with no zones exceeding 5% DF, confirming the previous findings of Lehtinen et al. (2022). Shallower balconies of 1.5 m allow for limited zones of acceptable daylight over 2% (Dahlan et al. 2009; Katunsky 2021; Lehtinen et al. 2022), particularly in CS2, CS3 and CS5.

In summary, given the varying room sizes, rooms behind balconies showed an average DF over 2% in three of the 10 cases (below the recommended 5%), while rooms without balconies showed higher DF values. Good daylight provision occurred only in shallow-plan layouts with dual-orientation windows, one of which was unobstructed by a balcony. Privacy concerns and overexposure underscored the need for thoughtful site layout as well as adaptable strategies to balance daylight provision and occupant comfort. Living spaces behind balconies face disadvantages due to greater depth; kitchens often located deepest would benefit most from improved daylight access. Finally, living spaces obstructed by balconies deeper than 1.5 m had poor daylighting conditions.

4.1.2 Measured DF versus CEN’s EN 17037 standard

To account for daylight availability, the results are also compared with the European standard for daylighting in buildings (CEN 2018) (Figure 7). While the standard specifies viewing level at 0.85 m height, in-situ measurements were performed at 1.0 m to reflect both the home environment and measuring in real-life conditions with furniture placement. Although this could create some discrepancies, this 15 cm difference is likely to be negligeable.

Figure 7

As shown in Figure 7, none of the case studies fulfilled both CEN (2018) criteria of (1) a DF of minimum 0.7% across 95% of the viewing plane and (2) a DF of minimum 2.2% across 50% of the viewing plane across all rooms. CS1 and CS3, where rooms are unobstructed by a balcony, met both requirements for adequate daylight access. These rooms are bedrooms, traditionally considered to have the lowest daylight requirements under UK standards—a notion that may be outdated given the rise of work-from-home practices. In CS1–CS3 and CS9, rooms unobstructed by a balcony met the second requirement. When applying the criteria to rooms obstructed by a balcony, none of the case studies meet either requirement (Figure 7). This outcome aligns with previous results and supports findings from Lehtinen et al. (2022) highlighting that the area behind a balcony typically does not meet recommended daylighting standards. This observation may be influenced by the compounding effect of balcony glazing and adjacent window. Although this study did not reveal a pronounced impact of balcony glazing alone (see the next section), it is plausible that the combined presence of both elements contributes to a poorly lit zone behind the balcony and is subject to further investigation. It is also worth noting that some argue that the minimum DF requirements specified in the standard may be excessive for housing (Hraska & Čurpek 2023), though further research is needed to understand reduced daylight provision and its implications on space use, energy use and occupant wellbeing.

4.1.3 Open versus closed balcony glazing

The impact of balcony glazing—closed versus open (Figure 6)—on the average DF can be attributed primarily to glazing properties and the shading effect of balcony glazing structures. Some variations may also result from real sky conditions during the balcony open–closed measurements, which may influence daylight availability, but any inconsistencies are considered small since indoor and outdoor measurements were taken simultaneously. For all case studies, actual light transmittances of balcony glazing were unavailable, except for the modelled case study where a manufacturer-specified value of 87% was applied. Given that all balcony systems consisted of clear toughened glass, it is considered reasonable to estimate a transmittance of 80–90%.

The absolute average DF differences between closed and open glazing ranged from –0.18 (CS9) to 0.37 (CS2) (see Table S2 in Appendix B in the supplemental data online). The highest percentage reduction of DF when glazing is closed was 32.38% (CS10) compared with open conditions. This correlates with the findings of Wilson et al. (2000) estimating a reduction of approximately 25–30% of DF simply from glazing and glazing structures. Note that this result is an isolated case in both Wilson et al. and the measured case studies. The highest percentage increase of DF when glazing was closed was 21.43% (CS8, bedroom). The stacked glazing in front of the space—when open—likely explains why DF is higher when the glazing is closed. This could also be due to unstable sky conditions or changed reflections. This is also the case for CS9 for which measurements also showed a significant increase of DF when the glazing is closed.

Overall, the mean absolute difference of average DF for all case studies was 0.08 and a mean percentage reduction of 3.02% when glazing is closed relative to open conditions (see Table S2 in Appendix B in the supplemental data online), which is within measuring error margins. Hence, no real difference could be observed across the studied cases (with some exceptions as above). Furthermore, no significant differences were observed in the results between framed and frameless balcony glazing structures nor with half- or full-length openable glass panels. Such minor differences, within the instrument’s measuring error margins, could arise from the in-situ method’s limited ability to capture the sensitivity of the glazing, for which more accurate and precise instruments are needed. However, this also suggests that the impact of the balcony glazing on indoor daylight conditions is low with the relatively minimal glazing systems observed in the case studies.

Relative to DF, adding glazing and associated structures can have a disproportionately large impact when baseline DF values are low (CS4–CS6, CS10). In such cases, the addition of glazing on a balcony can significantly affect overall daylight conditions and, by contrast, would have a less pronounced effect in a room where initial DF is already high, as is the case with CS1 and CS2 (see Table S2 in Appendix B in the supplemental data online).