Sage Journals: Discover world-class research

Abstract

Light affects many aspects of human physiology, through the non-image-forming (NIF) pathway. To account for this pathway, lighting design simulation tools need to combine several luminous and temporal factors to predict how architectural and lighting design decisions affect eye-level light exposure. Based on a systematic literature review, containing 55 journal and conference papers, the state-of-the-art towards implementing lighting beyond vision in computer simulation workflows for building design is presented. The review shows that, while interest in simulating the NIF effects of light on people is increasing, there is not a common method to perform these simulations. Gaps were identified in the currently available simulation workflows in relation to metrics, software and approaches for predicting NIF effects of light in the context of the building design.

1. Introduction

Light entering the eyes not only influences visual performance, but also affects many aspects of human physiology, through the non-image-forming (NIF) system.¹ The NIF responses include (but are not limited to) the regulation of the circadian system, the suppression of melatonin production, and acute alertness.² The photoreceptors that typically instigate these responses are the intrinsically photosensitive retinal ganglion cells (ipRGCs), which are sending signals to the central pacemaker of the human circadian rhythms.³ The ipRGCs have a peak spectral sensitivity at the short wavelength bluish light at approximately 480 nm,⁴ different to the known photoreceptors for vision (rods and three cone types). Moreover, the ipRGCs receive signals from rods and cones⁵ and knowledge about the contribution of each photoreceptor to the NIF system is growing.^6,7

The NIF responses to light are influenced by luminous and temporal factors, specifically light quantity, spectrum, directionality, timing, duration and previous light history (Figure 1).^8,9 Light quantity was conventionally measured through illuminance derived for photopic vision. The effect of increased light quantity on people was quantified through dose-response curves, where dose is the amount of light at eye-level and response is often alertness, melatonin suppression or circadian phase shift.^10,11 However, it is argued that photopic illuminance is not an appropriate metric for quantifying light considering its NIF effects¹² due to the differences in spectral sensitivities between the different photoreceptors. Therefore, new metrics have been introduced that account for the spectral sensitivity of the different photoreceptors. In regards to directionality (incidence of radiation to the retina), there are indications that melatonin suppression increases when light reaches the inferior (upper field of view) or the nasal (field of view from the nose side) retinal area.^13,14 Timing implies that light given at different phases of the circadian clock can reduce or extend a person’s internal day. Increasing the duration of light exposure increases the biological effect,¹⁵ although this relationship is non-linear, meaning that shorter duration light stimuli are more efficient per minute of exposure.¹⁶ Light history means that the adaptation conditions before a bright light exposure influence the biological response.¹⁷

Figure 1

The six luminous and temporal NIF light factors and the types of responses to them. The approximate duration of the response is indicated in parenthesis. Acute responses happen immediately after the light exposure (e.g. melatonin suppression). Circadian responses happen with a frequency of roughly a day (e.g. sleep-wake cycle). Long-term responses happen due to circadian disturbances for a prolonged period of time (e.g. seasonal affective disorder). The figure combines information from Khademagha et al.⁸ and Houser and Esposito¹⁸

The light exposure experienced by a person is mostly defined by architectural daylighting and artificial light sources, since people spend the majority of their time indoors.¹⁹ Living in urban environments is associated with decreased exposure to daylight and increased light levels at night, which delays the circadian clock.²⁰ While the main aim of building lighting design was previously to enhance visual performance, current knowledge from photobiology indicates that a new set of lighting criteria relevant for initiating NIF responses to light needs to be introduced. This adds complexity to architectural and lighting design projects and requires novel design tools.

Lighting simulation software is often used to assist decision-making in the architectural and lighting design process.^21,22 In response to the knowledge of the NIF pathway, new simulation workflows have been proposed to predict light in buildings.^23,24 These workflows should include the six above-mentioned luminous and temporal factors (Figure 1) that have been indicated as relevant for the NIF responses to light. When using lighting simulations to guide design decisions, it is recommended to consider both image-forming (IF) and NIF effects, a methodology now referred to by the International Commission on Illumination (CIE) as ‘integrative lighting’.²⁵ Daylight and electric light need to be considered separately as well as combined. This requires software that can model both light sources and can calculate eye-level light stimulus as an output.

The results of simulations need to provide useful information in the early and detailed phase of the architectural and lighting design practice. In the early design phase, it is important to take uncertainty into consideration in lighting simulations²⁶ and keep in mind that integrative lighting is a small piece of the puzzle which needs to be balanced with many other, sometimes contradicting, design criteria. In the detailed design phase, user behaviour might need to be considered as suggested by other studies.²⁷ Different workflows are applicable when zooming in to the detailed level of user behaviour or when zooming out to the level of building massing or urban design.

The aim of this paper is to identify the state-of-the-art in simulation workflows for daylighting and electric lighting design applicable for quantifying eye-level light stimulus for predicting NIF responses. Based on a systematic literature review, we present what are the gaps in the currently available simulation workflows in the context of building design.

2. Method

The literature review was conducted using four groups of keywords: ‘light’, ‘non-image-forming’, “simulation” and “building design”, and alternative terms for these (Table 1) in Web of Science, Scopus and Science Direct. The search was performed within the title, abstract and keywords and limited to publications after 2002, since this is approximately the time when ipRGCs were discovered.^3,28 As alternatives to “non-image-forming”, the terms “non-visual”, “circadian” and the more general term “health” were used. These terms were selected based on the commonly used keywords in the publications about this topic that we were familiar with prior to conducting the literature review. Whereas “non-image-forming” and “non-visual” are often used interchangeably (as in CIE S 026²⁹), “circadian” refers to a sub-set of these responses (Figure 1). Since the terminology around this topic varies and in order to avoid missing papers due to incomplete keywords, all eligible publications were forward-traced (by identifying the articles that cite them after they have been published). In addition, four relevant review papers found through the initial search that investigated the connection between architecture and NIF effects of light^8,30–32 were forward- and backward-traced (by searching their references). Eventually, 55 papers were included in the literature review (Figure 2). The final search was performed on 26th August 2022.

Table 1

Keywords for the literature search in the databases Web of Science, Scopus and Science Direct

	Connected by AND
Connected by OR	light*	“non image forming”	simulation*	“building design”
	daylight*	“non visual”		“architectural design”
		circadian		“built environment”
		health		“lighting design”
				“design support”
				“design process”
				“design parameters”

The asterisk * was used as a wildcard in Scopus and Web of Science. The quotation marks “ ” were used for search terms that consist of more than one word.

Figure 2

Search methodology. “Forward-traced” refers to identifying the articles that cite a paper after it has been published. “Backward-traced” refers to searching the references of an article

The following elements were extracted from each included publication:

if and how the previously described luminous and temporal factors were considered.

which software was used.

which type of light sources were simulated (electric light, daylight).

which metrics were used.

whether the workflow was suitable for simulating building massing models, rooms or users.

if IF effects of light (visual performance, comfort and/or interest) were also included.

3. Results

A summary of the analysed papers is presented in Table 2. Figure 3 shows the number of publications over the years.

Table 2

Summary of simulation methods in the analysed papers. The inclusion (or not) of the luminous and temporal factors relevant for NIF effects is denoted with “Yes” (or “No”). Quantity of light is not presented as a separate column because it was included in all reviewed publications (although by different metrics)

Study		Spectrum (source)	Spectrum (materials)	Directionality	Duration	Timing	History	Software		NIF metric		IF factors		Simulation level
Jakubiec and Alight³³		Yes	Yes	No	Yes	Yes	Yes	ALFA		mel. irrad., alertness, mel. sup., phase shift*		No
Alight and Jakubiec³⁴		Yes	Yes	No	Yes	Yes	Yes	ALFA, Lark, Radiance		mel. irrad., EML, M/P, nvR_D, alertness, mel. sup., phase shift*		No
Abboushi and Safranek³⁵		Yes	Yes	No	Yes	Yes	No	Honeybee, ALFA		EML		No
Danell et al.³⁶		Yes	No	No	Yes	Yes	Yes	Radiance		EML, nvR_D		No
Pierson et al.³⁷		Yes	Yes	No	Yes	No	Yes	Lark		mel. EDI, nvR_D		No
Amundadottir et al.³⁸		Yes	No	No	Yes	No	Yes	Radiance		nvR_D		Vis. interestVis. comfort
Saiedlue et al.³⁹		Yes	Yes	No	No	Yes	No	ALFA		EML		Vis. perform.Vis. comfort
Altenberg Vaz and Inanici⁴⁰		Yes	Yes	No	No	Yes	No	ALFA		EML		Vis. perform.Vis. comfort
Potocnik and Kosir⁴¹		Yes	Yes	No	No	Yes	No	ALFA		EML, CS		Vis. perform.Vis. comfort
Elsayed and Rakha⁴²		Yes	Yes	No	Yes	No	No	ALFA		EML		Vis. perform.Vis. comfort
Zeng et al.⁴³		Yes	Yes	No	No	Yes	No	Daysim		EML		Vis. perform.
Acosta et al.⁴⁴		Yes	No	No	Yes	Yes	No	Daysim		CS		Vis. comfort
Bellia et al.⁴⁵		Yes	Yes	No	No**	Yes	No	Daysim		CS		Vis. perform.
Salamati et al.⁴⁶		Yes	Yes	No	No	Yes	No	ALFA		EML, M/P		Vis. comfort
Rockcastle et al.⁴⁷		Yes	No	No	Yes	No	Yes	DIVA		nvR_D		Vis. perform.Vis. comfortVis. interest
Aguilar-Carrasco et al.⁴⁸		Yes	Yes	No	No	Yes	No	DIALux, Daylight Visualizer		CS		Vis. perform.
Zauner and Plishcke⁴⁹		Yes	Yes	No	No	Yes	No	Relux		mel. EDI		Vis. perform.
Safranek et al.⁵⁰		Yes	Yes	No	No**	Yes	No	ALFA		EML, CS		Vis. perform.
Brennan and Collins⁵¹		Yes	Yes	No	No	Yes	No	Daysim, Radiance, Lark		EML		No
Konis⁵²		Yes	Yes	No	No**	Yes	No	Lark		EML		No
Potocnik et al.⁵³		Yes	Yes	No	Yes	No	No	ALFA		CS		No
Acosta et al.⁵⁴		Yes	Yes	No	No	Yes	No	Daysim		CS		No
Lee and Boubekri⁵⁵		Yes	No	No	Yes	Yes	No	DIVA		illum.		No
Khademagha et al.⁵⁶		Yes	Yes	No	No	No	No	Radiance		effect. irrad.		Vis. perform.Vis. comfort
Yun et al.⁵⁷		Yes	No	No	No	Yes	No	Honeybee		EML		Vis. perform.Vis. comfort
Andersen et al.⁵⁸		Yes	No	No	Yes	No	No	Lightsolve		nvR_D****		Vis. interest
Inanici et al.⁵⁹		Yes	Yes	No	No	No	No	Radiance		circ. lum., circ. illum.		Vis. perform.
Ewing et al.⁶⁰		Yes	Yes	No	No	No	No	Radiance		5 α-opic irrad.		Vis. perform.
Potocnik and Kosir⁶¹		Yes	Yes	No	No	No	No	ALFA		EML, CS		Vis. comfort
Danell et al.⁶²		Yes	Yes	No	No	No	No	ALFA		EML		Vis. perform.
Hosseini and Sheikhansari⁶³		Yes	Yes	No	No	No	No	ALFA		EML		Vis. performVis. comfort
Song et al.⁶⁴		Yes	No	No	No**	Yes	No***	Honeybee		EML		Vis. perform.
Inanici et al.⁶⁵		Yes	No	No	No	Yes	No	ALFA, Lark, Radiance		EML		Vis. perform.		-
Pechacek et al.⁶⁶		Yes	No	No	No	Yes	No	Daysim, Relux		circ. potential		No
Geisler-Moroder and Dür⁶⁷		Yes	Yes	No	No	No	No	Radiance		CAF		No
Amundadottir et al.⁶⁸		Yes	No	No	Yes	No	No	Daysim		nvR_D****		No
Mardaljevic et al.⁶⁹		Yes	No	No	No	Yes	No	Radiance		circ. potential		No
Konis⁷⁰		Yes	No	No	No**	Yes	No***	Honeybee		EML		No
Kim et al.⁷¹		Yes	Yes	No	No	No	No	CQAT		CAF, CCT		No
Kim et al.⁷²		Yes	Yes	No	No	No	No	CQAT		CAF, CCT		No
Amirazar et al.⁷³		Yes	No	No	No**	Yes	No	DIVA		EML		No
Balakrishnan and Jakubiec⁷⁴		Yes	Yes	No	No	No	No	ALFA, Lark, Radiance		spect. irrad.*****		No		-
Busatto et al.⁷⁵		Yes	No	No	No	Yes	No	Radiance		CAF, CS, EML		No
Maskarenj et al.⁷⁶		Yes	No	No	No	Yes	No	OWL		mel. EDI, mel. irrad., CS		No		-
He et al.⁷⁷		Yes	Yes	No	No	No	No	ALFA		EML, CS, illum., M/P		No
Pierson et al.⁷⁸		Yes	Yes	No	No	No	No	ALFA, Lark, Radiance		spect. irrad.*****		No
Pierson et al.⁷⁹		Yes	Yes	No	No	No	No	ALFA, Lark, Radiance		spect. irrad.*****		No
Dai et al.⁸⁰		Yes	No	No	No	No	No	Radiance		EML, CS		Vis. perform.
Sanchez-Cano and Aporta⁸¹		Yes	No	No	No	No	No	DIALux		mel. EDI, EML, CS		Vis. perform.
Ezpeleta et al.⁸²		Yes	No	No	No	No	No	DIALux		EML, mel. EDI		Vis. perform.
Andersen et al.⁸³		Yes	No	No	No	No	No	Daysim		illum.		No
Khademagha et al.⁸⁴		No	No	Yes	No	No	No	Radiance		lum.		No
Yao et al.⁸⁵		Yes	No	No	No	No	No	Honeybee		CS		No
Lee and Boubekri⁸⁶		No	No	No	No	Yes	No	Unknown		illum.		No
Cai et al.⁸⁷		No	No	No	No	No	No	Daysim		illum.		No
Summary (number of papers)								ALFA	18	EML	25
								Radiance	16	CS	14
								Daysim	9	nvR_D	7
								Lark	8	mel.EDI	5	Vis. perform.	21		48
	47							Honeybee	5	CAF	4	Vis. comfort	12		4
	22	52	31	1	13	28	6	Others	13	Others	29	Vis. interest	3		3

: daylight; : electric light; : building level; : room level; : user level (see Figure 4 for definition); EML: Equivalent Melanopic Lux; CS: Circadian Stimulus; nvR_D: non-visual Direct Response; M/P: Melanopic/Photopic ratio; CAF: Circadian Action Factor; CCT: Correlated Colour Temperature; mel. EDI: melanopic Equivalent Daylight Illuminance; effect.: effective; irrad.: irradiance; circ: circadian; illum.: illuminance; lum.: luminance; spect.: spectral; mel. sup.: melatonin suppression; CQAT: Colour Quality Assessment Tool.

By using the Postnova et al.⁸⁸ and Tekieh et al.⁸⁹ models to predict Karolinska Sleepiness Scale, mean reaction time on a task and melatonin concentration.

But yearly duration or frequency is used.

***

Although history is mentioned, it can be understood as weekly frequency.

****

Earlier version of the nvR_D model.

*****

Metric not used as a NIF indicator, but to quantify spectral accuracy.

Figure 3

Number of publications in the topic of simulations of NIF effects of light in building design over the years

3.1 Luminous and temporal factors

Novel methodologies to simulate the effects of light on health and alertness were proposed. Pechacek et al.⁶⁶ combined light quantity, spectrum and timing in a preliminary simulation workflow to calculate a static “circadian potential” provided by daylight in a space. Andersen et al.²³ and Mardaljevic et al.⁶⁹ further developed that using photobiology knowledge about the time-dependent effect of light and they defined a lower and upper bound for the possibility that a given light dose will have an effect on people. They proposed to divide the day into three intervals: (1) early to mid-morning (06.00–10.00), when light advances the circadian phase, (2) mid-morning to early evening (10.00–18.00), when light can affect alertness, and (3) night-time (18.00–06.00), when light exposure that might trigger NIF responses is to be avoided. Amundadottir et al.³⁸ incorporated a dynamic mathematical model that additionally accounts for light duration and history to calculate the effect of a light exposure into a simulation framework. Their framework combined light quantity, spectrum, duration and history in a single model that aims to translate light stimulus to human response.

Geisler-Moroder and Dür⁶⁷ used the RGB channels from renderings with the Radiance software⁹⁰ to estimate the effect of light on the suppression of melatonin, therefore introducing low resolution (three channel) spectral information of sources and materials into simulations. Spectral information was also included by performing simulations of photometric quantities and post-processing the results based on the spectral power distribution (SPD) of the light source.^{35,36,38,44,47,66,68–70,73,75,80–83,85} This method disregards that the SPD that reaches the eye is modified by the spectral properties of materials, which might be a reasonable assumption in neutrally coloured spaces. For non-neutrally coloured spaces, a simplified approach was to multiply the SPD of the light sources with the spectral reflectance of the materials and use this to post-process photometric results.^45,48,54 This approach, however, has not been validated against measurements yet. Instead of post-processing the results, a pre-processing method was proposed by Zauner and Plischke.⁴⁹ Their method consists of creating a “melanopic simulation model” where the luminous flux of the luminaires and the luminous reflectance of materials are modified to a melanopic equivalent. The simulation result is directly interpretable in terms of melanopic quantities. One benefit of this method is that it can be applied to conventional lighting simulation software.

The duration of light exposure was either considered statically, assuming that a fixed duration of exposure of a few hours a day (1–5 hours) is enough to stimulate the NIF system,^42,44,55 or dynamically, considering that short duration light exposures are more effective than long ones, and intermittent light patterns are more effective than continuous per minute of exposure.^{32,36,38,47,58,68}

In addition, the directionality of light being projected on the retina of the eye was investigated by Khademagha et al.⁸⁴ that used renderings to estimate the effect of room design parameters by applying masks corresponding to the upper and lower field of view.

3.2 Software

Since the conventional lighting simulation software simplifies the visible spectrum into a three-dimensional RGB colour space, new spectral simulation software has been developed to process the spectral properties of light sources and materials. Inanici et al.⁵⁹ implemented a methodology to increase the spectral resolution of Radiance simulations to more than three channels. They used an n-step algorithm that divides the visible spectrum in n wavebands, where increasing the n increases the spectral resolution.^91,92 Since Radiance is a 3-channel (RGB) renderer, in order to implement, for example, a 9-step algorithm, three individual simulations need to be combined. Each of these three simulations accounts for a different part of the visible spectrum. The methodology by Inanici was translated into the Lark spectral lighting simulation tool.⁹³

Lark and ALFA⁹⁴ are two publicly available simulation tools that were developed for simulations of the NIF effects of light. Their innovation, in comparison to conventional simulation tools, is that the visible spectrum is divided into more than three channels (nine for Lark and 81 for ALFA) and they directly calculate metrics for quantifying NIF effects of light. Both tools can perform static point-in-time simulations. To expand their use to dynamic annual simulations, Jakubiec and Alight^33,34 implemented a method that uses 56 simulation timesteps throughout the year to interpolate to annual results (Lightsolve method⁹⁵).

The version 2.0 of Lark⁹⁶ also offers annual simulation using the Daylight Coefficient method for dynamic daylighting simulations.⁹⁷ However, one limitation is that it assumes a constant SPD for the entire simulation period. This can be overcome only with the use of a model that calculates annual SPDs from available weather data.⁶⁵ The recently developed OWL tool⁷⁶ is a step towards overcoming this limitation. OWL can be used to estimate the SPD of the sky dome using available weather data as input. It implements spatially and spectrally resolved sky models⁹⁸ that use the luminance of different patches of the sky to calculate SPD. Lark and OWL are both developed for the Grasshopper for Rhino environment and can potentially be combined.

A few studies were performed to validate the accuracy of these tools. One study found that Lark calculates spectral irradiance with most errors within a ±20% range in a neutrally coloured space under daylight.⁷⁸ The authors noted a reproducibility error in ALFA, unresolved at the time of their study. It was also shown that under clear sky conditions ALFA tends to overestimate the irradiance,⁹⁹ even though the spectral characteristics are well represented.⁷⁴ For electric light simulations, the errors in spectral irradiance were larger than the ±20% range, with ALFA being more accurate due to the higher spectral resolution.⁷⁹ However, when the spectral irradiance was used to calculate relevant NIF metrics, these differences vanished and most errors of both tools were within the ±20% range.⁹⁹

Another in-house spectral simulation software for electric light was developed by Kim et al.⁷¹ called Colour Quality Assessment Tool (CQAT), but it is limited to a simple geometry and diffuse materials.

3.3 Light sources

Out of the 55 papers, 33 studied only daylight, eight studied only electric light and 14 both daylight and electric light. Different methods to model the spectra of direct sunlight and diffuse skylight were adopted in the reviewed papers. Often, the assumption was made that the entire sky dome can be uniformly represented by a CIE standard illuminant (e.g. D65).^44,67,68,85 Other approaches were to use standard illuminants based on the window orientation of the simulated space,^48,64,66,83 differentiating between the spectrum of the direct and diffuse contribution,⁷⁰ or between overcast and clear sky conditions.^48,69 Bellia et al.⁴⁵ modelled typical average SPDs for cities located in the north, centre and south of Europe based on the percentage of overcast skies in each location (close to D50 for northern locations, close to D65 for middle locations and more blue than D65 in southern locations).

In ALFA the spectrum of the sun and the sky is generated from the library libRadTran,¹⁰⁰ using a US mid-latitude (45°N) summer atmospheric profile.⁷⁴ In Lark, the sun is neutrally coloured and the sky is uniformly coloured by a user selected SPD. In OWL, as mentioned, the SPD is calculated from available weather data. For this calculation, the luminance of different patches of the skydome is converted to Correlated Colour Temperature (CCT) using a model.⁹⁸ The CCT of each patch is then converted to SPD using the CIE 015 standard¹⁰¹ and is finally aggregated over the skydome. All three (ALFA, Lark and OWL) sky models have some limitations. A comparison of the ALFA-generated skies with measured data showed that ALFA did not show the wide variability in CCT that was observed in the measurements.⁶⁵ For Lark, the neutral sun assumption introduced errors in CCT calculation up to 41% when modelling clear skies (even though the error in NIF metrics was up to 17%).⁶⁵ In contrast, the luminance to CCT models of OWL correspond well with measured CCT under clear skies.⁹⁸ However, OWL applies these clear sky models for all sky types and does not discriminate between clear, intermediate and overcast skies.

3.4 Metrics

A variety of metrics was found in the literature to quantify the eye-level light stimulus (Table 2). The most used metrics were the Equivalent Melanopic Lux (EML) and the Circadian Stimulus (CS). EML was proposed by the building certification institute WELL¹⁰² based on the ipRGC spectral sensitivity curves from Lucas et al.¹⁰³ CS was introduced by Rea et al.¹⁰⁴ based on a model for nocturnal melatonin suppression, and was later adopted by the Underwriters Laboratory (UL) design guideline 24 480.¹⁰⁵ Melanopic Equivalent Daylight Illuminance (EDI), one of the five α-opic EDI metrics recommended by the CIE,²⁹ was used in five of the publications.⁸¹ Four (relatively older) publications used the Circadian Action Factor (CAF), a metric proposed by Gall and Bieske¹⁰⁶ as a ratio of circadian weighted and photometric quantities. Only photopic quantities (luminance, illuminance and annual metrics derived from these) were used in five publications.

A few papers used metrics that combine luminous and temporal factors to estimate a human response. Specifically, a dynamic metric called the “non-visual Direct Response” (nvR_D) model that was developed by Amundadottir²⁴ was used in seven papers. The model takes as input a time series of eye-level light stimuli and translates it to a predicted human alerting response. The mathematical models of Postnova et al.⁸⁸ and Tekieh et al.⁸⁹ to predict alertness, phase shift and melatonin suppression were used in two papers.^33,34 These models are based on the balance between homeostatic and circadian systems. The homeostatic system increases while a person is awake and the circadian system is defined by sleep history and dynamic light stimuli.

3.5 Modelling buildings, rooms and users

During early-stage daylight design, simulation workflows can go beyond modelling individual rooms, considering that parameters such as interior partitions or even specific materials might be unknown (here referred to as “building level”, Figure 4). The significance of using lighting simulations during this stage is clear, since decisions about the shape of the building and its relationship with the context largely affect the light that occupants will be exposed to, especially in dense urban contexts.¹⁰⁷ For this purpose, Konis⁵² developed a workflow to compare alternative building form options during early-stage design. His method takes as input a building massing model, it automatically divides it into floors, and evaluates an annual NIF potential of the entire building using Lark. A percentage of “circadian effective” zone is calculated for the building, which is the area where daylight over the year is considered sufficient for circadian entrainment (when EML exceeds a threshold for a number of hours).⁷⁰ Additionally, he proposed annual climate-based NIF daylight metrics⁷⁰ and suggested a grade-based system to evaluate spaces based on these metrics. These building-level simulations are suitable for early-stage daylighting design, where the goal is not to prove compliance with a specific standard but rather to do relative comparisons of design options and acquire general guiding directions.

Figure 4

Building, room and user level simulation workflows. Building level is more suitable for early-stage architectural daylighting design, while room and user level apply better to detailed daylighting and electric lighting simulations

Modelling of individual rooms is usually a next step of the design process, when space dimensions, window typology, material properties, layouts and electric lighting are explored (here referred to as “room level”, Figure 4). Most of the publications focused on the room level of detail. For the highest level of detail, user-focused simulations that account for occupant movement within a space and gaze behaviour can be included (here referred to as “user level”, Figure 4). The default assumption in simulations is to model users by selecting a position in the room where they are likely sitting (e.g. a desk), a height from the floor to represent the eye-level (often 1.2 m for a sitting person) and one or multiple vectors to indicate their static view direction. However, since occupants do not always stay at their desks looking towards a fixed direction, models indicating occupants’ movement within a space are useful in order to investigate their eye-level light stimulus. Since the eye-level light exposure is essential for evaluating a space based on its potential to induce NIF effects, the user level of detail can be considered as a baseline.

These user behaviour aspects were included in four publications. Specifically, in two publications, occupants were assumed to move randomly between selected zones.^68,83 In a paper that investigated light exposure in offices,³⁶ occupant behaviour profiles were defined based on their role in the office. The occupant profiles determined when and how much time a person would spend at their desk, in a meeting room, or in other spaces in the office. Gaze behaviour was included in a paper that proposed a human-centred approach to daylighting simulations,¹⁰⁸ applying a gaze responsive model that predicts eye movements based on luminance contrasts in the field-of-view.¹⁰⁹ Even though the user level approach presents practical difficulties, as it requires assumptions of how people move through a space, it needs further investigation to elaborate on the errors that arise from not considering it.

3.6 Integration of IF and NIF criteria

IF and NIF can to some extent be predicted with existing simulation software together or separately, and their combination was a topic of interest for almost half of the publications in this review. Andersen et al.⁵⁸ combined NIF potential with visual interest in an integrated simulation workflow. Amundadottir et al.³⁸ added glare considerations to that and demonstrated that NIF potential, visual interest and glare might be contradictory design goals. They indicated that a trade-off might be necessary depending on the space and occupant needs, for example by prioritizing visual comfort and NIF potential in spaces where occupants spend longer periods of time whereas visual interest in spaces where the duration of stay is shorter.

Rockcastle et al.⁴⁷ took that approach one step further and proposed a scoring method to balance visual, perceptual and NIF criteria. Their analysis demonstrated that evaluating a space based on horizontal illuminance could lead to different design decisions compared to designing based on eye-level light metrics (e.g. Daylight Glare Probability or nvR_D), suggesting that horizontal and eye-level metrics should be considered together. Several studies explored the effect of building design parameters on IF and NIF criteria,^{39,40,41,44,46,49,56,61,63,80,81} showing that an integrated design approach is needed to find the right balance between both. This underlines the need for having standards and guidelines that include these different aspects together and suggest what the trade-offs should be for various design applications. Though the daylight standard EN 17037¹¹⁰ includes considerations for daylight provision, sunlight exposure, glare and view, it notably leaves out recommendations for NIF effects. The revised lighting standard EN 12464-1¹¹¹ highlights the importance of daylight and the need for variation based on time of the day but does not provide concrete recommendations for NIF effects. Yet, the integration of all relevant criteria in industry standards might push the development of new simulation software that facilitates the incorporation in design practice.

4. Discussion

This review demonstrated that research interest in modelling the NIF effects of light on people within buildings is increasing (Figure 3). Studies focused on the development and validation of simulation workflows to investigate the circadian and acute responses to light. Some of the most recent studies used these workflows to investigate how the various design parameters affect the NIF responses and attempted to offer preliminary advice to designers. In this section we present the gaps in the available simulation workflows.

4.1 Gaps in metrics

In relation to metrics, a consensus is yet to be reached. The most commonly used metrics were EML and CS. CIE has defined five α-opic EDI metrics to describe light, based on the sensitivities of the five ocular photoreceptors and it is argued that melanopic EDI is a good predictor of circadian and acute light responses.^112,113 Melanopic EDI can be calculated from EML with a simple multiplication factor (mel. EDI = 0.91 × EML). A recent publication by a consortium of 18 experts in the field of light and health recommends for healthy, daytime working adults, to provide a vertical melanopic EDI of minimal 250 lx at eye level during daytime. For the evening this should be less than 10 lx and less than 1 lx for sleeping environments.¹¹⁴ It should be noted that the recommendations are based on data from mostly night-time laboratory studies.

Beyond the disputed nature of the metrics, a difficulty also lies in the fact that light quantities alone cannot comprehensively be used for the prediction of NIF responses because they do not include temporal dynamics. Human response metrics need to include luminous and temporal parameters together, since a single light quantity cannot predict the magnitude of a response without considerations for timing, duration and previous light history. The response metrics proposed by Amundadottir (nvR_D),²⁴ Postnova et al.⁸⁸ and Tekieh et al.⁸⁹ are a step towards that direction, but their applicability under a variety of (day)light conditions needs to be tested. One study that compared the nvR_D model with daytime alertness under daylight showed that there is a moderate correlation, but the model still has large prediction errors.¹¹⁵ This means that recalibration of this model is necessary, especially as we learn more about the contribution of daylight to the various effects on people.¹¹⁶

4.2 Gaps in software

The development of the ALFA, Lark and OWL software make spectrally resolved simulations more accessible, but they still have their limitations. All tools can perform static point-in-time simulations. Yet, dynamic simulations are necessary to enable the calculation of human response metrics that account for both luminous and temporal factors. The current tools can do that to some extent. Interpolating selected point-in-time results to get an annual result is possible using the Lightsolve method, but it is not directly implemented in ALFA, Lark or OWL. The Lark version 2.0⁹⁶ can directly perform dynamic simulations, but with the assumption of a constant yearly SPD.

This assumption of a constant SPD of a selected standard CIE illuminant was often made in the literature. Is it a reasonable simplification or is it important to dynamically vary the SPD of daylight? Inanici et al.⁶⁵ argue that we do not model annual sky luminance based on an average, so we should not model the sky spectrum based on an average as well. Indeed, Diakite-Kortlever and Knoop⁹⁸ tested the accuracy of using the CIE D65 illuminant to predict melanopic content outdoors. They found that it can represent sky SPD reasonably well under overcast sky conditions, but for clear skies it underestimates the melanopic content. They claim that for clear skies spectral information of each patch of the sky dome is needed for accurate daylight simulations. In contrast, Pierson et al.⁹⁹ found that the assumption of a constant D65 was reasonably accurate in simulating melanopic EDI indoors in a variety of sky conditions (clear, hazy, overcast, rainy). They note that, in the conditions that they tested, accurately simulating quantity of light (irradiance) had a larger impact to the result than accurately simulating spectrum. It is though difficult to claim that this is generalizable for other locations and during all times of the day and year. We need further comparisons of simulations with measurements from different locations, especially indoors, to know how the varying sky colour affects the simulation accuracy in terms of both quantity and spectrum.

A different issue is that these software tools might be difficult to access due to lack of user-friendliness or cost. Lark and OWL require familiarity with Rhino, Grasshopper and possibly basic programming knowledge (if modifications to the basic code need to be made). ALFA is easier to use, since the only prerequisite is the ability to use Rhino; however, it is a commercial licensed software. In addition, since Rhino (which is also a commercial software) might not be the default CAD program for all lighting and architectural designers, the use of such plugins is restricted. Simulation methods with pre- or post-processing steps^49,81 can also be used to calculate NIF metrics using freely available simpler tools (like DIALux¹¹⁷ or Relux¹¹⁸), but with limited spectral resolution and/or without accounting for the materials’ spectral reflectance. Moreover, the additional steps make the method more prone to user error.

4.3 Gaps in approaches for predicting NIF effects

In Table 2 one can see that none of the reviewed papers included all six relevant luminous and temporal parameters. This is probably because the exact relationship between light quantity, spectral composition, directionality, timing, duration, history and specific human responses still needs to be established,² which is an issue beyond simulation software and method. The two least frequently used factors relevant for NIF responses in lighting simulation studies are light history and directionality. Further research into these parameters should provide information on how to implement them in simulation studies.

Finally, a significant issue when aiming to predict the effects of light on people by using simulation, is that a real person’s complex behaviour in a space needs to be represented using a limited number of static positions and view directions. For the vast majority of the indoor simulation papers examined, the authors used a sensor in the vertical plane to model light received at the eye (with the exception of two papers by Lee and Boubekri^55,86 that used light on a horizontal plane as a proxy for that received at the eye). Thus, it is reasonable to conclude that there is widespread acceptance amongst researchers that light prediction on the horizontal plane is not sufficient to predict the light dose received at peoples’ eyes. Nevertheless, in reality, gaze direction of building occupants in a static position (e.g. seated at a desk) will vary in both azimuth (i.e. compass direction) and altitude (i.e. directed above or below the horizon) to a degree that is difficult to characterize.^119,120 Accordingly, the uncertainty in received light dose due to gaze behaviour is presently unknown. Furthermore, depending on their schedule and/or patterns of work, a person may spend considerable periods of the day not at their notional workstation. Simulation results do indeed support the hypothesis that the occupants’ view direction and their distance from window significantly affects the magnitude of the light dose they receive.^40,41,84 This suggests that further investigations into how people move in spaces and change their view direction could provide more reliable input for simulations.

4. Conclusion

This paper presented a systematic literature review of 55 journal and conference papers, aiming to identify the state-of-the-art and the gaps in simulation workflows for predicting NIF effects of light. In the introduction of this paper, we presented the relevant factors for simulating eye-level light stimulus, needed for predicting NIF responses. In the results section, we identified which of these factors are already included in simulation workflows and what methods were applied. In the discussion section we distinguished gaps in relation to metrics, software and approaches for predicting NIF light effects.

There are still many uncertainties about the mechanism with which light affects the range of biological responses, but with increasing research on these we will be able to move from light metrics (e.g. melanopic EDI) to human response metrics (e.g. predicted alertness or melatonin suppression). Temporally dynamic simulations are needed for this to be possible in practice. These are partly implemented in the available spectral simulation software, but improvements are needed especially to the spectral sky and sun models. Nonetheless, it is not clear how detailed these models need to be to accurately predict the quantity and spectrum of light indoors.

It is also unknown if light quantities incident on static vertical vectors are a good predictor of the actual light dose that a person receives. People’s dynamic behaviour in indoor environments in relation to their light exposure requires further investigation. It is understandable that this adds complexities to the task of optimizing lighting to support human needs, but only by exploring these complexities we can break them down to simple design guidelines.

Footnotes

Acknowledgements

We would like to thank professor John Mardaljevic for reading the manuscript and providing his feedback.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received funding from the European Training Network LIGHTCAP (project number 860613) under the Marie Skłodowska-Curie actions framework H2020-MSCA-ITN-2019 for the research and publication of this article.

ORCID iDs

M Gkaintatzi-Masouti

J van Duijnhoven

MPJ Aarts

References

Boyce

PR.

Human Factors in Lighting, 3rd edition. Boca Raton, FL: CRC Press, 2014.

Vetter

Pattison

Houser

Herf

Phillips

AJK

Wright

, et al. A review of human physiological responses to light: implications for the development of integrative lighting solutions. LEUKOS 2021; 18: 387–414.

Hattar

Liao

Takao

Berson

Yau

KW.

Melanopsin-containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 2002; 295: 1065–1070.

Zaidi

Hull

Peirson

SNN

Wulff

Aeschbach

Gooley

, et al. Short-wavelength light sensitivity of circadian, pupillary, and visual awareness in humans lacking an outer retina. Current Biology 2007; 17: 2122–2128.

MTH

. Melanopsin and the intrinsically photosensitive retinal ganglion cells: biophysics to behavior. Neuron 2019; 104: 205–226.

Gooley

Rajaratnam

SMW

Brainard

Kronauer

Czeisler

Lockley

SW.

Spectral responses of the human circadian system depend on the irradiance and duration of exposure to light. Science Translational Medicine 2010; 2: 31–33.

Rüger

Gordijn

MCM

Beersma

DGM

De Vries

Daan

Time-of-day-dependent effects of bright light exposure on human psychophysiology: comparison of daytime and nighttime exposure. American Journal of Physiology - Regulatory Integrative and Comparative Physiology 2006; 290: 1413–1420.

Khademagha

Aries

MBC

Rosemann

ALP

van Loenen

EJ.

Implementing non-image-forming effects of light in the built environment: a review on what we need. Building and Environment 2016; 108: 263–272.

Rea

Figueiro

Bullough

JD.

Circadian photobiology: an emerging framework for lighting practice and research. Lighting Research and Technology 2002; 34: 177–187.

10.

Cajochen

Zeitzer

Czeisler

Dijk

DJ.

Dose-response relationship for light intensity and ocular and electroencephalographic correlates of human alertness. Behavioural Brain Research 2000; 115: 75–83.

11.

Zeitzer

Dijk

Kronauer

Brown

Czeisler

CA.

Sensitivity of the human circadian pacemaker to nocturnal light: melatonin phase resetting and suppression. Journal of Physiology 2000; 526: 695–702.

12.

Spitschan

Stefani

Blattner

Gronfier

Lockley

Lucas

How to report light exposure in human chronobiology and sleep research experiments. Clocks and Sleep 2019; 1: 280–289.

13.

Khademagha

Aries

Rosemann

Loenen

E Van

. Why directionality is an important light factor for human health to consider in lighting design? International Journal of Sustainable Lighting 2016; 18: 3–8.

14.

Khademagha

Aries

MBC

Rosemann

ALP

Van Loenen

EJ.

A multidirectional spectral measurement method and instrument to investigate non-image-forming effects of light. Measurement Science and Technology 2018; 29: 8.

15.

Dewan

Benloucif

Reid

Wolfe

Zee

PC.

Light-induced changes of the circadian clock of humans: increasing duration is more effective than increasing light intensity. Sleep 2011; 34: 593–599.

16.

Chang

Santhi

St Hilaire

Gronfier

Bradstreet

Duffy

, et al. Human responses to bright light of different durations. Journal of Physiology 2012; 590: 3103–3112.

17.

Chang

Scheer

FAJL

Czeisler

CA.

The human circadian system adapts to prior photic history. Journal of Physiology 2011; 589: 1095–1102.

18.

Houser

Esposito

Human-centric lighting: foundational considerations and a five-step design process. Frontiers in Neurology 2021; 12: 630553.

19.

Klepeis

Nelson

Ott

Robinson

Tsang

Switzer

, et al. The National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to environmental pollutants. Journal of Exposure Analysis and Environmental Epidemiology 2001; 11: 231–252.

20.

Wright

McHill

Birks

Griffin

Rusterholz

Chinoy

. Entrainment of the human circadian clock to the natural light-dark cycle. Current Biology 2013; 23: 1554–1558.

21.

Ochoa

Aries

MBCC

Hensen

JLMM

. State of the art in lighting simulation for building science: a literature review. Journal of Building Performance Simulation 2012; 5: 209–233.

22.

Ayoub

100 years of daylighting: a chronological review of daylight prediction and calculation methods. Solar Energy 2019; 194: 360–390.

23.

Andersen

Mardaljevic

Lockley

SW.

A framework for predicting the non-visual effects of daylight-part I: photobiology-based model. Lighting Research and Technology 2012; 44: 37–53.

24.

Amundadottir

ML.

Light-driven model for identifying indicators of non-visual health potential in the built environment. PhD Dissertation, Ecole Polytechnique Federale de Lausanne, Switzerland, 2016.

25.

International Commission on Illumination. Position Statement on Non-Visual Effects of Light: Recommending Proper Light at the Proper Time. Vienna: CIE, 2019.

26.

Østergård

Jensen

Maagaard

SE.

Building simulations supporting decision making in early design - a review. Renewable and Sustainable Energy Reviews 2016; 61: 187–201.

27.

van Duijnhoven

Aarts

MPJ

Kort

HSM.

Personal lighting conditions of office workers: an exploratory field study. Lighting Research and Technology 2021; 53: 285–310.

28.

Berson

Dunn

Takao

Phototransduction by retinal ganglion cells that set the circadian clock. Science 2002; 295: 1070–1073.

29.

International Commission on Illumination. System for Metrology of Optical Radiation for ipRGC-Influenced Responses to Light. CIE S 026/E:2018. Vienna: CIE, 2018.

30.

Sithravel

Ibrahim

Identifying supportive daytime lighting characteristics for enhancing individuals’ psychophysiological wellbeing in windowless workplace in tropical Malaysia. Indoor and Built Environment 2021; 30: 298–312.

31.

Bellia

Fragliasso

Good places to live and sleep well: a literature review about the role of architecture in determining non-visual effects of light. International Journal of Environmental Research and Public Health 2021; 18: 1002.

32.

Andersen

Unweaving the human response in daylighting design. Building and Environment 2015; 91: 101–117.

33.

Jakubiec

Alight

Spectral and biological simulation methods for the design of healthy circadian lighting. Proceedings of Building Simulation 2021: 17th Conference of IBPSA, Bruges, Belgium, 1–3 September 2021.

34.

Alight

Jakubiec

JA.

Evaluating the use of photobiology-driven alertness and health measures for circadian lighting design. Proceedings of Building Simulation 2021, Bruges, Belgium, 1–3 September 2021.

35.

Abboushi

Safranek

Determining critical points to control electric lighting to meet circadian lighting requirements and minimize energy use. Proceedings of Annual Modeling and Simulation Conference (ANNSIM), San Diego, USA, 18–21 July 2022, pp. 559–568.

36.

Danell

Amundadottir

Rockcastle

Evaluating temporal and spatial light exposure profiles for typical building occupants. Proceedings of Symposium in Simulation in Architecture and Urban Design, 25–27 May 2020, pp. 539–546.

37.

Clotilde

Eugenia

SMV

Mariëlle

Marilyne

A conceptual simulation workflow to guide design decisions regarding the effects of daylight on occupants’ alertness. Proceedings of International Conference CISBAT 2021, Lausanne, Switzerland, 8–10 September 2021.

38.

Amundadottir

Rockcastle

Sarey Khanie

Andersen

A human-centric approach to assess daylight in buildings for non-visual health potential, visual interest and gaze behavior. Building and Environment 2017; 113: 5–21.

39.

Saiedlue

Amirazar

Place

Assessing circadian stimulus potential of lighting systems in office buildings by simulations. Proceedings of ARCC Conference Repository, Toronto, Canada, 29 May–1 June 2019, pp. 719–27.

40.

Altenberg Vaz

Inanici

. Syncing with the sky: daylight-driven circadian lighting design. LEUKOS 2020; 17: 291–309.

41.

Potočnik

Košir

Influence of geometrical and optical building parameters on the circadian daylighting of an office. Journal of Building Engineering 2021; 42: 102402.

42.

Elsayed

Rakha

A framework to simulate the non-visual effects of daylight on the cognitive health of mild cognitive impairment (MCI) people. Proceedings of 2020 Building Performance Analysis Simbuild Virtual Conference, 29 September–1 October 2020, pp. 119–29.

43.

Zeng

Sun

Lin

Optimized lighting energy consumption for non-visual effects: a case study in office spaces based on field test and simulation. Building and Environment 2021; 205: 108238.

44.

Acosta

Leslie

Figueiro

MG.

Analysis of circadian stimulus allowed by daylighting in hospital rooms. Lighting Research and Technology 2015; 49: 49–61.

45.

Bellia

Acosta

Campano

MÁ

Fragliasso

Impact of daylight saving time on lighting energy consumption and on the biological clock for occupants in office buildings. Solar Energy 2020; 211: 1347–1364.

46.

Salamati

Mathur

Kamyabjou

Taghizade

Daylight performance analysis of TiO2@W-VO2 thermochromic smart glazing in office buildings. Building and Environment 2020; 186: 107351.

47.

Rockcastle

Amundadottir

Andresen

The case for occupant-centric daylight analytics: a comparison of horizontal illumination and immersive view. Proceedings of Building Simulation 2019, Rome, Italy, 2–4 September 2019, pp. 1239–1246.

48.

Aguilar-Carrasco

Domínguez-Amarillo

Acosta

Sendra

JJ.

Indoor lighting design for healthier workplaces: natural and electric light assessment for suitable circadian stimulus. Optics Express 2021; 29: 29899–29917.

49.

Zauner

Plischke

Designing light for night shift workers: application of nonvisual lighting design principles in an industrial production line. Applied Sciences (Switzerland) 2021; 11: 10896.

50.

Safranek

Collier

Wilkerson

Davis

RG.

Energy impact of human health and wellness lighting recommendations for office and classroom applications. Energy and Buildings 2020; 226: 110365.

51.

Brennan

Collins

AR.

Outcome-based design for circadian lighting: an integrated approach to simulation and metrics. Proceedings of ASHRAE and IBPSA-USA Building Simulation Conference, Chicago, USA, 26 to 28 September, 2018: pp. 141–8

52.

Konis

A circadian design assist tool to evaluate daylight access in buildings for human biological lighting needs. Solar Energy 2019; 191: 449–58.

53.

Potočnik

Košir

Dovjak

Colour preference in relation to personal determinants and implications for indoor circadian luminous environment. Indoor and Built Environment 2020; 31: 121–138.

54.

Acosta

Campano

MÁ

Leslie

Radetsky

Daylighting design for healthy environments: analysis of educational spaces for optimal circadian stimulus. Solar Energy 2019; 193: 584–596.

55.

Lee

Boubekri

Introduction of new daylighting metrics for health, wellbeing, and feasibility: a study of the indoor building environment. Journal of Green Building 2022; 17: 105–126.

56.

Khademagha

Diepens

JFL

Aries

MBC

Rosemann

ALP

van Loenen

EJ.

Effect of different design parameters on the visual and non-visual assessment criteria in office spaces. Proceedings of International Conference CISBAT 2015, Lausanne, Switzerland, 9–11 September, 2015, pp. 363–368.

57.

Yun

Jeong

Choi

Photopic illuminance-based black-box model for regulation of human circadian rhythm via daylight control. Building and Environment 2021; 203: 108069.

58.

Andersen

Guillemin

Amundadottir

Rockcastle

Beyond illumination: An interactive simulation framework for nonvisual and perceptual aspects of daylighting performance. Proceedings of 13th Conference of the International Building Performance Simulation Association, Chambery, France, 26–28 August, 2013, pp. 2749–2756.

59.

Inanici

Brennan

Clark

Spectral daylighting simulations: computing circadian light. Proceedings of 14th International Conference of IBPSA - Building Simulation, Hyderabad, India, 7–9 December, 2015, pp. 1245–1252.

60.

Ewing

Haymaker

Edelstein

EA.

Simulating circadian light: multi-dimensional illuminance analysis. Proceedings of Building Simulation 2017, San Fransisco, USA 7–9 August, 2017, pp. 2363–2371.

61.

Potočnik

Košir

Influence of commercial glazing and wall colours on the resulting non-visual daylight conditions of an office. Building and Environment 2020; 171: 106627.

62.

Danell

Hartmeyer

Peterson

Davis

Andersen

Rockcastle

The impact of light distribution and furniture layout on meeting light exposure objectives in an office - A simulation case study. Proceedings of Building Simulation 2021, Bruges, Belgium, 1–3 September, 2021.

63.

Hosseini

Sheikhansari

A daylight assessment on visual and nonvisual effects of light shelves: a human-centered simulation-based approach. Journal of Daylighting 2022; 9: 28–47.

64.

Song

Jiang

Cui

A study on nonvisual effects of natural light environment in a maternity ward of a hospital in cold area. BioMed Research International 2022; 2022: 8608892.

65.

Inanici

Abboushi

Safranek

Evaluation of sky spectra and sky models in daylighting simulations. Lighting Research and Technology. First published 2 July 2022. DOI: 10.1177/14771535221103400

66.

Pechacek

Andersen

Lockley

SW.

Preliminary method for prospective analysis of the circadian efficacy of (day)light with applications to healthcare architecture. LEUKOS 2008; 5: 1–26.

67.

Geisler-Moroder

Dür

Estimating melatonin suppression and photosynthesis activity in real-world scenes from computer generated images. Proceedings of 5th European Conference on Colour in Graphics, Imaging, and Vision and 12th International Symposium on Multispectral Colour Science, Joensuu, Finland, 14–17 June 2010, pp. 346–352.

68.

Amundadottir

Lockley

Andersen

Simulation-based evaluation of non-visual responses to daylight: Proof-of-concept study of healthcare re-design. Proceedings of 13th Conference of the International Building Performance Simulation Association, Chambery, France, 26–28 August, 2013, pp. 2757–2764.

69.

Mardaljevic

Andersen

Roy

Christoffersen

A framework for predicting the non-visual effects of daylight - part II: the simulation model. Lighting Research and Technology 2014; 46: 388–406.

70.

Konis

A novel circadian daylight metric for building design and evaluation. Building and Environment 2017; 113: 22–38.

71.

Kim

Choi

Sung

MK.

Development of a colour quality assessment tool for indoor luminous environments affecting the circadian rhythm of occupants. Building and Environment 2017; 126: 252–265.

72.

Kim

Choi

Sung

MK.

Accuracy evaluation of a calculation tool based on the spectral colour property of indoor luminous environments. Building and Environment 2018; 139: 157–169.

73.

Amirazar

Azarbayjani

Zarrabi

Ashrafi

Cox

, et al. Assessing the circadian potential of an office building in the southeastern US. Proceedings of the Symposium on Simulation for Architecture and Urban Design, Delft, Netherlands, 4–7 June, 2018, pp. 24–30.

74.

Balakrishnan

Jakubiec

Spectral rendering with daylight: a comparison of two spectral daylight simulation platforms. Proceedings of Building Simulation 2019, Rome, Italy, 2–4 September, 2019, pp. 1191–1198.

75.

Busatto

Mora

Peron

Romagnoni

Application of different circadian lighting metrics in a health residence. Journal of Daylighting 2020; 7: 13–24.

76.

Maskarenj

Deroisy

Altomonte

A new tool and workflow for the simulation of the non-image forming effects of light. Energy and Buildings 2022; 262: 112012.

77.

Yan

Cai

Improving the accuracy of circadian lighting simulation with field measurement. Journal of Building Performance Simulation 2022; 15: 575–598.

78.

Pierson

Aarts

MPJ

Andersen

Validation of spectral simulation tools for the prediction of indoor daylight exposure. Proceedings of Building Simulation 2021, Bruges, Belgium, 1–3 September, 2021.

79.

Pierson

Gkaintatzi-Masouti

Aarts

MPJ

Andersen

. Validation of spectral simulation tools for the prediction of indoor electric light exposure. Proceedings of CIE 2021 Malaysia Midterm Meeting and Conference, 27–29 September, 2021.

80.

Dai

Huang

Hao

Lin

Chen

Spatial and spectral illumination design for energy-efficient circadian lighting. Building and Environment 2018; 146: 216–225.

81.

Sánchez-Cano

Aporta

Optimization of lighting projects including photopic and circadian criteria: a simplified action protocol. Applied Sciences (Switzerland) 2020; 10: 8068.

82.

Ezpeleta

Orduna-Hospital

Solana

Aporta

Pinilla

Sánchez-Cano

Analysis of photopic and melanopic lighting in teaching environments. Buildings 2021; 11: 439.

83.

Andersen

Gochenour

Lockley

SW.

Modelling “non-visual” effects of daylighting in a residential environment. Building and Environment 2013; 70: 138–149.

84.

Khademagha

Aries

Rosemann

van Loenen

New method for analysing a luminous environment considering non-image-forming effects of light. Proceedings of 33rd PLEA International Conference, Edinburgh, 2–5 July, 2017, pp. 3245–3252.

85.

Yao

Cai

Xue

Dai

Efficient circadian daylighting: a proposed equation, experimental validation, and the consequent importance of room surface reflectance. Energy and Buildings 2020; 210: 109784.

86.

Lee

Boubekri

Impact of daylight exposure on health, well-being and sleep of office workers based on actigraphy, surveys, and computer simulation. Journal of Green Building 2020; 15: 19–42.

87.

Cai

Yue

Dai

Hao

Lin

Shi

, et al. The impact of room surface reflectance on corneal illuminance and rule-of-thumb equations for circadian lighting design. Building and Environment 2018; 141: 288–297.

88.

Postnova

Lockley

Robinson

PA.

Prediction of cognitive performance and subjective sleepiness using a model of arousal dynamics. Journal of Biological Rhythms 2018; 33: 203–218.

89.

Tekieh

Lockley

Robinson

McCloskey

Zobaer

Postnova

Modeling melanopsin-mediated effects of light on circadian phase, melatonin suppression, and subjective sleepiness. Journal of Pineal Research 2020; 69: e12681.

90.

Ward

and Shakespeare R. Rendering with Radiance: The Art and Science of Lighting Visualization. San Francisco: Morgan Kaufmann Publishers, 1998.

91.

Ruppertsberg

Bloj

Rendering complex scenes for psychophysics using radiance: how accurate can you get?

Journal of the Optical Society of America 2006; 23: 759–767.

92.

Ruppertsberg

Bloj

Creating physically accurate visual stimuli for free: spectral rendering with radiance. Behavior Research Methods 2008; 40: 304–308.

93.

Inanici

, ZGF Architects. Lark spectral lighting. Retrieved 12 October 2022 from https://faculty.washington.edu/inanici/Lark/Lark_home_page.html

94.

Solemma. ALFA. Retrieved 12 October 2022 from https://www.solemma.com/alfa

95.

Bodart

Kleindienst

Andersen

Graphical representation of climate-based daylight performance to support architectural design. LEUKOS 2008; 5: 39–61.

96.

Gkaintatzi-Masouti

Pierson

van Duijnhoven

Andersen

Aarts

MPJ

. A simulation tool for building and lighting design considering ipRGC-influenced light responses. Proceedings of Building Simulation Nordic 2022, Copenhagen, Denmark, 22–23 August, 2022.

97.

Bourgeois

Reinhart

Ward

Standard daylight coefficient model for dynamic daylighting simulations. Building Research and Information 2008; 36: 68–82.

98.

Diakite-Kortlever

AKA

Knoop

Forecast accuracy of existing luminance-related spectral sky models and their practical implications for the assessment of the non-image-forming effectiveness of daylight. Lighting Research and Technology 2021; 53: 657–676.

99.

Pierson

Aarts

MPJ

Andersen

Validation of spectral simulation tools in the context of ipRGC-influenced light responses of building occupants. Journal of Building Performance Simulation. First published 26 September 2022. DOI: 10.1080/19401493.2022.2125582

100.

Emde

Buras-Schnell

Kylling

Mayer

Gasteiger

Hamann

, et al. The libRadtran software package for radiative transfer calculations (version 2.0.1). Geoscientific Model Development 2016; 9: 1647–1672.

101.

International Commission on Illumination. Colorimetry, 4th edition. CIE 015:2018. Vienna: CIE, 2018.

102.

International WELL Building Institute. WELL v2 Q4 2021: WELL building standard. Retrieved 12 October 2022 from https://v2.wellcertified.com/wellv2/en/overview

103.

Lucas

Peirson

Berson

Brown

Cooper

Czeisler

, et al. Measuring and using light in the melanopsin age. Trends in Neurosciences 2014; 37: 1–9.

104.

Rea

Figueiro

Bullough

Bierman

A model of phototransduction by the human circadian system. Brain Research Reviews 2005; 50: 213–228.

105.

Underwriters Laboratory. Design Guidelines for Promoting Circadian Entrainment with Light for Day-Active People. Northbrook, IL: UL Design Guideline 24480, 2019.

106.

Gall

Bieske

Definition and measurement of circadian radiometric quantities. light and health - non-visual effects. Proceedings of the CIE symposium 2004, Vienna, Austria, 30 September–2 October, 2004, 1:129–132.

107.

Nault

Waibel

Carmeliet

Andersen

Development and test application of the UrbanSOLve decision-support prototype for early-stage neighborhood design. Building and Environment 2018; 137: 58–72.

108.

Rockcastle

Amundadottir

Andersen

A simulation-based workflow to assess human-centric daylight performance. Proceedings of the Symposium on Simulation for Architecture and Urban Design, Toronto, Canada, 21–24 May, 2017, pp. 17–24.

109.

Sarey Khanie

. Human responsive daylighting in offices: a gaze-driven approach for dynamic discomfort glare assessment. PhD Dissertation, Ecole Polytechnique Federale de Lausanne, Switzerland, 2015.

110.

European Committee for Standardization. Daylight in buildings. EN 17037:2018. 2018.

111.

European Committee for Standardization. Light and lighting - Lighting of work places - Part 1: Indoor work places. EN 12464-1:2021. 2021.

112.

Brown

TM.

Melanopic illuminance defines the magnitude of human circadian light responses under a wide range of conditions. Journal of Pineal Research 2020; 69: e12655.

113.

Schlangen

LJM

Price

LLA.

The lighting environment, its metrology, and non-visual responses. Frontiers in Neurology 2021; 12: 624861.

114.

Brown

Brainard

Cajochen

Czeisler

Hanifin

Lockley

, et al. Recommendations for daytime, evening, and nighttime indoor light exposure to best support physiology, sleep, and wakefulness in healthy adults. PLoS Biology 2022; 20: e3001571.

115.

Soto Magan

. Alertness in work environments - on the role of indoor daylight exposure. PhD Dissertation, Ecole Polytechnique Federal de Lausanne, Switzerland, 2021.

116.

Münch

Wirz-Justice

Brown

Kantermann

Martiny

Stefani

, et al. The role of daylight for humans: gaps in current knowledge. Clocks and Sleep 2020; 2: 61–85.

117.

DIAL GmbH. DIALux. Retrieved 12 October 2022 from https://www.dialux.com/en-GB/.

118.

Relux Informatik

. Relux. Retrieved 12 October 2022 from https://relux.com/en/

119.

Sarey Khanie

Stoll

Einhäuser

Wienold

Andersen

. Gaze and discomfort glare, Part 1: Development of a gaze-driven photometry. Lighting Research and Technology 2017; 49: 845–865.

120.

van Duijnhoven

Aarts

MPJ

Kort

HSM.

The importance of including position and viewing direction when measuring and assessing the lighting conditions of office workers. Work 2019; 64: 877–895.

Simulations of non-image-forming effects of light in building design: A literature review

Abstract

1. Introduction

2. Method

3. Results

3.1 Luminous and temporal factors

3.2 Software

3.3 Light sources

3.4 Metrics

3.5 Modelling buildings, rooms and users

3.6 Integration of IF and NIF criteria

4. Discussion

4.1 Gaps in metrics

4.2 Gaps in software

4.3 Gaps in approaches for predicting NIF effects

4. Conclusion

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

ORCID iDs

References