Occupants visual comfort assessments: A review of field studies and lab experiments

Abstract

Many daylight metrics are developed to assess different aspects of visual comfort in recent years. However, there are still some doubts about the robustness of these metrics. This paper presents a comprehensive and thorough analysis of 58 field studies and lab experiments (2012–2020), focusing on metrics used to predict occupants' acceptance and visual preference in indoor environments. The review first covers the methodology of studies. Second, the results are discussed in four main categories: 1) confounding parameters in visual comfort studies; 2) metrics’ accuracy; 3) newly defined metrics or modified versions of the existing ones, and 4) metric thresholds. Finally, the lessons learned and limitations of the existing studies and recommendations for future studies on visual comfort are presented. Going through the results of those studies focusing on finding a proper metric for visual comfort evaluations revealed some of the previous studies had limitations in the experimental, statistical, and equipment. Ignoring these issues may have biased the studies’ findings. According to results, there is still a lack of consensus on adequate illuminance thresholds. Moreover, due to the complexity of the visual comfort concept and occupants' perception dependency on diverse factors, there is no global glare metric that could be applied to various conditions. To address these gaps, this study presents future researches trends, including the development of comprehensive study protocols, long-term glare assessment, and refining metrics thresholds. The findings of this paper could help researchers facilitate the application of filed and lab studies in visual comfort studies.

Introduction

Daylight perception is a significant factor in indoor environmental quality. Occupants’ acceptable light levels should be provided to guarantee comfortable and productive spaces (Chraibi et al., 2016). Daylight not only improves health, awareness, productivity, and sense of comfort but also results in energy consumption reduction (Jakubiec and Reinhart, 2016, Jakubiec and Reinhart, 2012). Therefore, it is necessary to understand which factors affect daylight quality and quantity.

Daylighting is a notoriously difficult building performance strategy to evaluate (Reinhart et al., 2006). Many researchers attempt to predict the daylight quantity under varying experimental situations. So far, there are several metrics available and little is known about their application. Daylight indices which have been developed to assess different aspects of visual comfort (i.e., amount of light, light uniformity, light quality, and glare), differ from each other for several features, e.g., their accuracy, simplicity, space, and time discretization (Carlucci et al., 2015). Daylight metrics are implemented in technical standards and rating systems, suggesting simplified methods and values for assessing daylight and glare performance in buildings (e.g., LEED, BREEAM, IES, DNGB, EN 17037, BIS 8206) (Committee, 2019, Konis and Selkowitz, 2017)

Among daylight and visual comfort studies, a group has focused on subjective evaluations via field and lab experiments to acquire information on occupants’ visual preferences (Velds, 2002), investigating the following issues:

• effective parameters on daylight perception

• metrics’ performance and robustness

• metrics’ thresholds

• visual comfort issues

• development of new metrics for predicting occupants’ visual comfort

Researchers found that not all the existing methods can accurately evaluate visual comfort, which resulted in poor consistency between laboratory-derived measures and occupants’ experience in buildings (Jakubiec and Reinhart, 2016), therefore many researchers are trying to across some issue through the study of human-level cognition. Moreover, satisfying light conditions differ significantly between individuals (Chraibi et al., 2016) due to the known inherent variability between occupants concerning visual comfort perception, preference, and acceptance (Van Den Wymelenberg, 2014). Therefore, there is no consensus yet as to what metric can accurately predict visual comfort, and it is necessary to understand the relationship between metric assessments relative to human evaluations through field studies (Nezamdoost and Van Den Wymelenberg, 2017b).

Previously, several review articles have been published dealing with different issues of visual comfort. Van Den Wymelenberg (2014) compared the existing glare metrics and synthesized the effect of glare and other factors on blinds operation based on several laboratory and field studies. Carlucci et al. (2015) reviewed and categorized visual comfort indices according to common features. The pros and cons of different daylighting assessment methods (e.g. numerical modeling, field measurement, scale model, and manual procedures) are compared by Wong (2017). Dogan and Park (2018), criticized the current CBDM metrics in residential buildings, identified shortcomings, suggested an improved daylight analysis procedure. Pierson et al. (2018) summarized window-based discomfort glare indices and discomfort glare perception definition methods. Allan et al. (2019) reviewed and highlighted the benefits and shortcomings of general lighting and glare measures. Ayoub (2019) reviewed daylight prediction fundamental trends based on weather datasets, sky models, geometry of buildings, and daylight calculation methods. A detailed description of fifty common simulation tools and metric performance are presented in this review. Hamedani et al. (2019) reviewed the existing literature to provide a holistic overview of implemented methods in measuring light-induced physiological responses (i.e. pupil size, eye movement, gaze direction, degree of eye-opening, and blink rate) to objectify perceived glare. Wasilewski et al. (2019) conducted a literature review of Spatio-temporal glare simulation to document the limitations of current methods and assess the potential to apply it in empirical studies.

This paper presents an overview of 58 field studies and lab experiments (2012–2020(, which assessed the relation between visual comfort metrics and occupants’ preferences (i.e., by questionnaires, simulations, and measurements). Since studies were conducted in different building types and geographical locations, the performance of the same metrics is dissimilar or even contradictory. Fig. 1 presents the dispersion of studied metrics.

To evaluate the prediction accuracy of widely used daylight metrics, a global visual comfort database is required. The limited number of studies prohibits building a precise meta-analysis model. In contrast to thermal comfort studies, in which researchers have built a database by gathering and harmonizing data from field studies completed around the world. Which might be used to support comfort metrics (Cheung et al., 2019). By categorizing the visual comfort field and experimental studies based on their methodology and results, the current work tries to address the lack of visual comfort guidelines and comprehensive protocols for evaluating visual performance in buildings.

The included publications are not suggested to comprise a comprehensive representation of the state of the art in this area. Relevant publications that more closely fit in the scope of this contribution were reviewed. The body of literature was made up of eligible papers according to the inclusion criteria such as journal quality, publication year, matching research question. All the reviewed papers are published in authentic and peer-reviewed journals (e.g., Building and Environment, Building and Energy, Applied Energy, Solar Energy, and international conference proceedings such as PLEA and IBPSA). On the other hand, in the daylight community, 2012 was the year IES published LM-83, which documented the definition and calculation procedures for the first two humans-based daylighting design metrics: spatial daylight autonomy and annual sunlight exposure, based on field studies. Since then, subjective studies were conducted to assess the viability of daylight metrics in different climates. Moreover, the purpose of the current study is to investigate only experiments with daylight settings. The authors are aware that there were more studies investigating discomfort glare before 2012. They were removed due to overlap with Wymelenberg’s (2014) review paper. In which, visual comfort data cross-examinations, metric robustness testing, and, data collection methods and results database, were highlighted as first priority for potential future research.

Among reviewed papers, 42 were published during 2016–2020. Furthermore, about half of the studies were conducted in America, while 36% and 26% were in Asia and Europe respectively, and two studies in Australia as presented in Fig. 2. Most studies were conducted in the northern hemisphere (middle latitude about 30-60°N), where seasonal variations are observed on an annual basis. While seasonal variation is relatively low in the tropic, and daylight availability is relatively high.

More than half of the articles investigated office spaces (60%), one third studied educational buildings (34%), four papers explored residential buildings (7%), and 4 studied other building types.

Several methods are known efficient in occupant behavior assessments. In the reviewed studies, subjective assessments are done in natural settings (field) or under controlled conditions (laboratory experiment). Field methods tend to observe, analyze, and describe what exists (e.g. (Mahić et al., 2017, Mangkuto et al., 2017)). While laboratory methods are tightly controlled investigations in which the researcher manipulates the particular factor under study to determine if such manipulation generates a change in the subjects (e.g. (Bian and Luo, 2017, Yamin Garretón et al., 2018)).

In the following sections, first, the methodology of each study type is presented, and the general user assessment methods are described afterward. Second, the results are discussed in four main categories: 1) Confounding parameters in visual comfort studies (i.e., architectural, personal, and lighting-related); 2) evaluation of metric accuracy; 3) newly defined metrics or modified versions of the existing ones, and 4) proposed metric thresholds. Finally, the limitations of existing studies and recommendations for future studies on visual comfort are presented. The authors are aware that many of the existing studies show severe problems in their methodology such as experimental conditions (sampling size, order effect, low stimuli range), statistical analysis (Type-I error and false statistical inferences), and equipment (HDR pixel overflow). Since they are published in highly ranked peer-reviewed journals, this study reports and discusses their methods to point out defects. However, authors do not attempt to synthesize the results in the way a systematic review would. Fig. 3 shows a research concept map.