Assessment of blue light hazards and correlated colour

Assessment of Blue Light Hazards and Correlated Colour Temperature for

Public LED Lighting

Professor Joanne Wood1 Dr Alexander Black1 Dr Gillian Isoardi2

1School of Optometry and Vision Science, Institute of Health and

Biomedical Innovation Queensland University of Technology, 2Light Naturally, Brisbane, Australia

March 2019

Initiated by: Delivered by:

Dedicated t b o a etter Brisbane

t . energex

Queensla d u . of Technonl niversity ogy

Page 2 of 84: Blue Light and Public LED Lighting

1 Table of Contents

1 Table of Contents ............................................................................................................. 2

1.1 List of Figures .................................................................................................................... 3 1.2 List of Tables ...................................................................................................................... 4

2 Acknowledgements .......................................................................................................... 5

3 Executive Summary ......................................................................................................... 6

3.1 Main Findings: ................................................................................................................... 6 3.1.1 Blue Light Hazard ................................................................................................. 6 3.1.2 Circadian Health Impact ....................................................................................... 7 3.1.3 Environmental Impact ........................................................................................... 8 3.1.4 Driving performance and safety ............................................................................ 8

4 Glossary of Terms ............................................................................................................ 9

5 List of Abbreviations ..................................................................................................... 12

6 Introduction .................................................................................................................... 13

7 Photometric concepts of blue light ............................................................................... 14

7.1 Key Photometric Concepts ............................................................................................... 14 7.1.1 Spectral Power Distribution (SPD) ..................................................................... 14 7.1.2 Correlated Colour Temperature (CCT) .............................................................. 15 7.1.3 Radiance/Irradiance ............................................................................................ 17

7.2 Standards and guidelines for photobiological impact of electric light sources ................ 18 7.3 Measuring the impact of blue light ................................................................................... 19

8 Field-based measurement methodology ....................................................................... 20

9 Blue light and retinal damage ....................................................................................... 25

9.1 Mechanisms of retinal phototoxicity ................................................................................ 27 9.2 Blue-Light Hazard (BLH) ................................................................................................ 28

9.2.1 The problem with using CCT as an indicator of Blue-Light Hazard................... 31 9.3 Blue light hazard and Macular Degeneration ................................................................... 33 9.4 Field-based measurements of LED Streetlights ............................................................... 34 9.5 Comparisons with other light sources .............................................................................. 38 9.6 Summary: LED Streetlights and retinal damage .............................................................. 40

9.6.1 Summary of literature .......................................................................................... 40 9.6.2 Summary of the field-based measurements.......................................................... 40

10 Blue light and the circadian rhythm ............................................................................ 42

10.1 Field-based measurements of LED Streetlights ............................................................... 46 10.2 Summary: LED Streetlights and circadian rhythm........................................................... 48

10.2.1 Summary of Literature ......................................................................................... 48 10.2.2 Summary of field-based data ............................................................................... 48

11 Environmental impacts of blue light ............................................................................ 49


11.1 Sky Glow .......................................................................................................................... 51 11.2 Impact of blue light on wildlife ........................................................................................ 58 11.3 Summary: Environmental impacts and Blue Light .......................................................... 62

11.3.1 Summary of literature .......................................................................................... 62

12 LED Streetlighting and driving .................................................................................... 63

13 Appendix: Field-based data .......................................................................................... 65

13.1 Raw field-based data ........................................................................................................ 65 13.2 Photometric Modelling of spill light ................................................................................ 74

14 References ....................................................................................................................... 80

1.1 List of Figures

Figure 7-1: Comparison of LED SPDs to 2700K incandescent lamp SPD (left) and 6000K daylight (right). The top two graphs show a blue-based white LED spectral output, the bottom graphs illustrate a violet-based white LED output. ........................................................................ 15

Figure 7-2: Spectral Power distributions for blackbody radiators at different temperatures [from (Ling, Sanny et al. 2018)] .............................................................................................................. 16

Figure 7-3: CIE 1931 Colour Chart showing the blackbody curve (in red) and the location of blackbody radiators at different temperatures. [from (Pro-Lite Technologies 2018)] ........................ 16

Figure 7-4:Magnified view of the blackbody curve on the CIE 1931 Colour Chart, showing the nominal CCT regions according to ANSI and MacAdam ellipse definitions. [from(ANSIC78.377 2011). .................................................................................................................................. 17

Figure 8-1: Examples of the personal light sources - TV (upper left), laptop screen (upper right), computer screen (lower left), smartphone (lower right) ..................................................... 21

Figure 8-2: Examples of the interior lighting - 3500K Linear Fluorescent (left) and 3000K LED Downlight (right) ................................................................................................................ 21

Figure 8-3 Schematic map of activity-based measures – crosses indicate measurement points at 1.5m above the ground for cyclists and pedestrians, and at ~1.0m above the ground for drivers (inside car). Coloured arrows indicate direction of travel. ................................................ 23

Figure 8-4 Schematic map of spill light assessment measures – red crosses indicate a measurement point at 1.5m above the ground and facing towards the luminaire – red crosses continue in 3m intervals until the next luminaire (up to 66m). Green crosses indicate a measurement point at 3.0m above the ground and facing towards the luminaire. ................................... 23

Figure 9-1: Illustration of the age-related change in transmission of different wavelengths as a function of age. Blue box indicates the blue light hazard range [adapted from (Turner and Mainster 2008)] .................................................................................................................................. 25

Figure 9-2: Summary of the various interactions and impact of light of different wavelengths of light on the eye ............................................................................................................................ 27

Figure 9-3: Retinal Blue Light Hazard B(λ) curve, and Retinal Thermal Hazard R(λ) curve [from (International Electrotechnical Commission 2014)] .......................................................... 29

Figure 9-4: Example spectral power distributions for 4 different light sources with identical CCT (all 3000K) showing the variation in blue light content (blue light ranging from 7.8 – 16.5% of source output) [source: (CELMA-ELC 2011)] ................................................................... 32

Figure 9-5: Graph of KB,v vs CCT for a range of light sources, demonstrating a linear trend indicating increased proportion of blue light with increasing CCT. [(International Electrotechnical Commission 2014)] ............................................................................................................. 32

Figure 10-1: Melanopic retinal ganglion cells (mRGCs) and photopic spectral response curves. ..... 42 Figure 10-2: Spectral sensitivity for assessment of Circadian Light (CLA), for warm and cool light

sources (by CCT) [from Figueiro, Gonzales et al. (2016)]. ............................................... 43


Figure 10-3: Graph of acute melatonin suppression at one-hour exposure according to Circadian Light (CLA) and Circadian Stimulus (CS) measures. Red dashed line represents the threshold below which there is little melatonin suppression [from Figueiro, Gonzales et al. (2016)]. ............................................................................................................................................ 44

Figure 10-4: Graph showing Melanopic and Circadian stimulus response curves [from Clark and Lesniak (2017)]. .................................................................................................................. 45

Figure 11-1: Scotopic sky luminance ratio variation with observer distance, ratio compared to LPS scotopic luminance [from Luginbuhl, Boley et al. (2014)]. ................................................ 52

Figure 11-2: Same data with a different metric, photopic sky luminance ratio variation with observer distance, ratio again compared to LPS photopic luminance [from Luginbuhl, Boley et al. (2014)]. ............................................................................................................................... 53

Figure 11-3: Near source observer results for modelled sky glow ratios (relative to HPS) for unweighted (left) and scotopically weighted (right) measures [from (USDOE 2017)]. ..... 55

Figure 11-4: Distant observer (40km from source) results for modelled sky glow ratios (relative to HPS) for unweighted (left) and scotopically weighted (right) measures [from (USDOE 2017)]. 56

Figure 11-5: Spectral response for a range of animal behavioural processes and visual responses, as well as photosynthesis (plants) and sky glow vs wavelength (nm) [from (Longcore, Rodríguez et al. 2018)]. ...................................................................................................... 59

Figure 11-6: Spectral Power Distributions for test light sources to evaluate wildlife impact, vs wavelength (nm) [from (Longcore, Rodríguez et al. 2018)]. .............................................. 60

Figure 11-7: Calculated impact of individual light sources on wildlife, including overall weighted ‘Wildlife Index’ based on the selected spectra. Also highlighted is source CCT by colour [from (Longcore, Rodríguez et al. 2018)]. .......................................................................... 61

Figure A 1: Spectral Power Distributions of streetlights measured in the field ................................... 68 Figure A 2: Close-up view of Spectral Power Distributions of streetlights measured in the field,

highlighting BLH region and including BLH spectral curve. ............................................. 69 Figure A 3: Map of spill light assessment field measures – vertical illuminance and luminance – red

crosses indicate a measurement point at 1.5 m above the ground and facing towards the luminaire – red crosses continue in 3 m intervals until the next luminaire (from 40.5 m to 66 m, depending on installation). Green crosses indicate a measurement point at 3.0 m above the ground and facing towards the luminaire. .................................................................... 74

1.2 List of Tables

Table 9-1: Emission limits for BLH and BLH small source from IEC 62471 Photobiological safety of lamps and lamp systems ...................................................................................................... 29

Table 9-2: Emission limits (EL) defining risk groups ........................................................................... 30 Table 9-3: Risk Groups (RG) for lamps defined in IEC 62471 ............................................................. 31 Table 9-4 Readings of blue-weighted irradiance and photopic illuminance for driving, cycling and

pedestrian activities; the highest blue-weighted irradiance measurements for each source are italicised........................................................................................................................ 35

Table 9-5: Maximum field readings of horizontal illuminance at nadir (h=1.5 m), with corresponding calculated blue-weighted irradiance, and their ratio, KB,v ................................................. 36

Table 9-6: Maximum readings of luminance in the field for each luminaire type and equivalent blue-weighted radiance, LB ......................................................................................................... 36

Table 9-7: BLH irradiance values for exterior streetlight sources and typical interior lighting scenarios ............................................................................................................................................ 39

Table 9-8: Table C2 from IEC 62778 (International Electrotechnical Commission 2012) showing the illuminance values required to give a risk group not greater than RG1 (low risk) ............ 41

Table 10-1: Maximum irradiance-based measures from street lighting scenarios (directly beneath


luminaire) and equivalent circadian measures. .................................................................. 46 Table 10-2: Maximum values measured for spill light at the property boundary (vertical illuminance at

3 m above the ground) and equivalent circadian measures. .............................................. 47 Table 11-1: Blue light measures for a range of light sources for matched lumen output (from (Kinzey

2016)] .................................................................................................................................. 50

Table A 1: Activity-based measures of blue light hazard weighted irradiance, driving, cycling and walking activities ................................................................................................................ 65

Table A 2: Maximum irradiance-based values measured for each source (horizontal illuminance directly under luminaire, nadir, at 1.5 m above the road surface) ..................................... 66

Table A 3: Maximum Radiance-based Values measured for each source in spill light assessment. .... 66 Table A 4: Irradiance based measures of BLH and circadian stimulus for common interior activities,

compared to walking activity. ............................................................................................. 67 Table A 5: Spill Light illuminance and luminance, along roadway at 1.5m above road surface ......... 70 Table A 6: Spill Light blue weighted irradiance, EB, and radiance, LB, along roadway at 1.5 m above

road surface. ....................................................................................................................... 71 Table A 7: Spill Light illuminance and luminance, along and towards property boundaries, at 3.0m

above road surface .............................................................................................................. 72 Table A 8: Spill Light blue weighted irradiance, EB, and radiance, LB, along and towards property

boundaries, at 3.0m above road surface ............................................................................. 73 Table A 9: Calculated ULOR at 0° and 5° upcast for street lighting luminaires ................................. 75 Table A 10: Modelled values of vertical illuminance facing the street light (height 3m, in lux) for 6

different luminaires types for 0° and 5° upcast. Locations for assessment are highlighted by green crosses in Figure A 3 above. ..................................................................................... 76

Table A 11: Measured values of vertical illuminance facing the street light (height 3m, in lux) for 6 different luminaires types. Locations for assessment are consistent with Table A 10 above. ............................................................................................................................................ 76

Table A 12: Measured values of vertical illuminance and luminance (height 3m) facing the street light for field study of 6 different luminaires types. Maximum values along property boundary for each lamp type are highlighted. Locations for assessment are highlighted by green crosses in Figure A 3 above. Luminances are given in kcd/m2 (×1000 for value in cd/m2) ............ 78

Table A 13: Readings of vertical illuminance facing luminaire (pole 1) and luminance of luminaire for viewing positions shown by red crosses in Figure A 3 above. Measurements taken at 1.5 m above the ground, parallel to roadway. .............................................................................. 79

2 Acknowledgements

We would like to acknowledge staff of the QUT Photometric Laboratory and Eric Isdale for

their assistance in collecting the field-based measurements. We would also like to acknowledge

Steve Coyne for his valuable advice and technical contribution.


3 Executive Summary

The purpose of this study was to provide information from a literature review and field-based

testing and modelling, regarding the level of blue light associated health and environmental

risks from LED streetlighting with different Correlated Colour Temperatures (CCT) (colour

appearances). It is important to note here, that CCT is commonly used as an indicator of blue

light content, with higher CCT generally associated with more blue light; however this is not

always true. The Spectral Power Distribution (SPD) of an LED streetlight (the amount of

radiant power at different wavelengths of light) is a more accurate representation of the amount

of blue light in a source. The blue light emitted from commonly encountered interior and

personal light sources was also considered. Blue light was analysed in terms of its potential

effect on eye health (Blue Light Hazard (BLH)), circadian health, environmental impact and

driving performance.

3.1 Main Findings:

3.1.1 Blue Light Hazard

The existing literature regarding the potential for BLH from LED sources suggests that while

blue light from these sources can cause retinal damage, this would require exposure at much

higher intensities and for more prolonged exposures than is typically encountered in the normal

road environment.

The field-based analysis discussed in this report considered the amount of BLH present for

road users and at property boundaries at a range of LED streetlight installations (with CCTs of

3200K, 4000K and 5900K) compared with other traditional streetlights (high pressure sodium,

mercury vapour and fluorescent) and commonly encountered sources of light at night (TV,

computer and smartphone). Section 9.6.2 provides detail of the BLH-related values measured

in the field study; the main findings are summarised below:

• In all of the assessed road user scenarios (driver, cyclist and pedestrian), all of the

streetlighting technologies produced blue-weighted irradiance readings that were more

than 100 times lower than the recommended exposure limits for BLH;

• Radiation in the BLH range arising from driving, cycling and pedestrian activities under

all streetlighting technologies tested was lower than for many common interior

activities and personal device use, and would be shorter in exposure than these interior


activities. (N.B. Exposure in this context refers to the period of time spent looking

directly into a light source, and is not the cumulative time spent in the vicinity of a

source without looking directly towards it);

• BLH decreases rapidly with increasing distance of the observer from the source (by

the inverse square law). The on-street and roadside measurements made in this field

study were conducted relatively near the source, so that it is expected that residences

located further from these sources will experience even lower BLH risk.

3.1.2 Circadian Health Impact

The impact of light on the circadian system is a function of the spectral power distribution of

the source and the duration of exposure. The spectral sensitivity of the circadian system peaks

at around 460nm (blue). The existing literature suggests that the potential for LED streetlight

sources to cause disruption to circadian health through suppression of the human melatonin

hormone is likely to be low, given their low intensity and relatively short exposure when

encountered in typical road environments.

The field-based analysis considered the circadian health impact for road users and at property

boundaries for LED streetlights compared with traditional streetlights and commonly

encountered sources of light at night. Section 10.2.2 provides detail on the circadian-related

values measured in the field study; the main findings are summarised below:

• All maximum exterior circadian stimulus (CS) values measured directly under the

luminaires for each technology type were well below the threshold (8 times less)

considered to induce melatonin suppression or circadian disruption;

• The CS values measured at the property boundaries (3 m above the ground) for each

technology were well below the threshold to induce melatonin suppression.

Considering that light falls off rapidly with increased distance from the source (by the

inverse square law), CS values measured at the property will be even lower than those

measured at the boundary;

• The CS values recorded for common interior night time activities were between 10 and

100 times higher than those recorded for road users at night under all street lighting

technologies.


3.1.3 Environmental Impact

Review of the existing literature emphasizes the fact that all light at night has a negative impact

on sky glow, and disrupts wildlife and plant function.

Blue light can increase the impact of sky glow through enhanced light scatter in the shorter

(blue) wavelength region, dependant on atmospheric conditions, with blue-rich sources

modelled to exhibit up to 25% increased scatter in contrast to sodium light sources. The impact

of this increased scatter on sky observation decreases with increased distance from the source.

In addition, scotopic (low-light) vision is more sensitive in the blue wavelength range, which

increases perceived sky glow at night. While some LED sources may emit higher proportions

of blue light than traditional streetlighting sources, practice-based research and case studies

have shown that design factors for LED luminaires can offset the increase in sky glow that may

arise from their increased blue light content. In particular, the reduction of upwards-directed

light, as well as light close to the horizontal, and the opportunity for dimming afforded by LED

technologies have been demonstrated, through modelling and in practice, to effectively negate

potential increases in sky glow in comparison to high pressure sodium sources. A study

exploring these factors suggests that reducing the upwards light output ratio (ULOR) of a

source is the factor with the greatest potential to reduce sky glow for distant observers.

All light at night has the potential to disrupt wildlife and plants. The spectral power distribution

of a light source is a significant factor in its environmental impact, with different biological

species displaying peak sensitivities across different spectral regions. To assess the severity of

the impact, an action spectrum for the biological mechanism of animal or plant species of

interest should be identified so light sources with appropriate spectral content can be selected.

3.1.4 Driving performance and safety

The effects of LED streetlighting on driving performance, glare and safety are not well

understood. For example, while it is well known that increasing the radiant light output of a

source will increase both discomfort and disability glare, the specific effect of a light source’s

spectral distribution on these measures of glare remains unknown and requires further

controlled experimentation.


4 Glossary of Terms

Adaptation: The adjustment of visual sensitivity of the eye to varying light intensities.

Aphakic: An individual without a naturally occurring crystalline lens, usually following

surgical removal of cataracts or a congenital condition. The prevalence of aphakia is relatively

low, as the natural lens is typically replaced with an artificial intraocular lens (IOL).

Blue Light: Short wavelength light in the blue-violet range of the visible spectrum, around

430nm – 500nm, although the specific range of wavelengths varies across reports.

Blue-Light Hazard: Short wavelength light (in the blue to violet range) that has the potential

to cause retinal damage. The peak of this spectral curve ranges from 380 – 550 nm in

wavelength, with peak values between 440 – 460nm.

Brightness: The apparent luminance of an object in the field of view that is related to the

amount of light emitted by that object.

Candela: The standard unit of light intensity, which is 1 lumen per steradian; abbreviated to

cd.

Cataract: Cloudiness and yellowing of the crystalline lens that usually occurs with increasing

age. Cataract surgery involves removal of the cloudy lens and replacement with an artificial

intraocular lens (IOL).

Cones: Photoreceptors that are specialised for daylight vision, fine visual acuity and colour

and are most densely concentrated in the centre of the retina.

Contrast: The difference in luminance between an object and its background or between the

lighter and darker sections of the same object.

Correlated Colour Temperature: The colour appearance of a white light source, in terms of

its apparent “warmth” or “coolness” expressed in Kelvin (K). The value corresponds to the

temperature at which a theoretical blackbody radiator would emit radiation of the same colour

as the source.

Crystalline lens: The crystalline lens is the natural eye lens which produces one third of the

eye's total optical power and focuses light on the retina. In individuals 45 years and below, the

lens is able to change shape in order to focus from distance to near.

Glare: Glare arises from either excessive light arising directly from a source, or indirectly


through reflection from an intermediate surface. Disability Glare refers to an objectively

measured degradation in visual performance resulting from scattered light from a glare source,

whereas, Discomfort Glare does not impair visual performance, but can be distracting and

uncomfortable.

Illuminance: The amount of light incident on a surface per unit area, measured in lux.

Illumination: A measure of the lighting conditions within the environment, that is, how objects

are struck by light photons, either directly from light sources or indirectly from reflections from

other objects. Measured in lux (lx).

Inverse Square Law: A physical law that describes the reduction of a physical quantity with

distance from its source, according to the inverse of the distance squared, i.e.

quantity measured α 1

(distance from source)2

Irradiance: The quantity of radiant flux from the source that is received on a detector surface

per unit area, and is given in W/m2.

Lighting Emitting Diode (LED): A solid-state, semiconductor device that converts electrical

energy into light.

Luminance: The light intensity per square metre of apparent area of a light source or

illuminated surface, measured in candelas per metre squared (cd/m2). This physical quantity

relates to the sensation of brightness.

Luminance ratio: The ratio of the luminance of a light source relative to the luminance of its

immediate surrounds. This measure relates to contrast.

Luminous intensity: The luminous flux in a given direction, measured in candela (cd).

Lux: Photometric unit of illuminance; abbreviated as lx.

Melanopsin retinal ganglion cells (mRGC): Specialised nerve cells in the eye that express

the photopigment melanopsin and act as a relay for dark and light signals from the rod and

cone photoreceptors to the suprachiasmatic nucleus of the hypothalamus that mediates

circadian rhythms in the body.

Mesopic: The visual conditions generally associated with low luminance levels between 0.001

and 3 cd/m2, when both rod and cone photoreceptors are active. This represents typical night

driving conditions with streetlighting.


Phakic: An individual with a naturally occurring crystalline lens in the eye.

Photometry: The measurement of light.

Photopic: The visual conditions generally associated with luminance levels greater than 3

cd/m2, when the cone photoreceptors are active and colour is perceived. This represents normal

daylight conditions, floodlit conditions or adequate artificial room lighting.

Photoreceptors: Specialised nerve cells in the eye (rods and cones) that contain

photopigments; absorption of light by these cells triggers changes in the cell’s electrical

potential.

Pseudophakia: An individual who has had an artificial lens (IOL) implanted into the eye

following cataract surgery.

Radiance: The radiant flux emitted by a source per unit solid angle, per unit projected area

(measured in W.st-1.m-2).

Rods: Photoreceptors that are specialised for night vision and provide monochromatic vision

and are located at all positions across the retina except for the very centre.

Scotopic: The visual conditions generally associated with adaptation to very low luminance

levels less than 0.001 cd/m2, when rod activity dominates vision and there is no colour

perception. This represents night-time conditions without additional light sources (eg.

starlight).

Spectral Power Distribution: The amount of radiant power a light source emits at different

wavelengths across the visible spectrum.

Upwards Light Output Ratio: The ratio of the upward luminous flux of a light source, to the

total individual luminous fluxes of the light source

Visibility: The capacity to detect the presence of an object but not necessarily the ability to

recognise the identity of that object.


5 List of Abbreviations

AMA American Medical Association

ANSI American National Standards Institute

BLH Blue Light Hazard

CCT Correlated Colour Temperature

CIE International Commission for Illumination

CLA Circadian Light

CS Circadian Stimulus

EML Equivalent Melanopic Lux

FL Fluorescent

IDA International Dark-Sky Association

IEC International Electrotechnical Commission

IESNA Illuminating Engineering Society North America

IOL Intra-ocular lens

HPS High Pressure Sodium

LED Light-Emitting Diode

LPS Low Pressure Sodium

MD Macular Degeneration

mRGC Melanopsin Retinal Ganglion Cells

MV Mercury Vapour

RPE Retinal Pigment Epithelium

SPD Spectral Power Distribution

ULOR Upwards Light Output Ratio

USDOE US Department of Energy

UV, UVR Ultraviolet , Ultraviolet radiation


6 Introduction

The introduction of new lighting technologies, particularly low energy lighting involving light-

emitting diode (LED) technology which is rich in blue light, has created significant media

awareness regarding the potential negative impact of these technologies on health and the

environment. Consideration of the negative effects of LED lighting have focused on the impact

on eye health and visual perception, including damage to the retinal structures, particularly via

the mechanism known as the blue-light hazard (Bullough 2000, Behar-Cohen, Martinsons et

al. 2011), the impact on the perception of glare (Fekete, Sik-Lanyi et al. 2010, Sweater-

Hickcox, Narendran et al. 2013), as well as the effects of this lighting technology on circadian

rhythms and sleep patterns (Rea, Bierman et al. 2008).

This report focuses on the potential impacts of blue-rich light arising from LED streetlights

with different Correlated Colour Temperature (CCT). The blue light emitted from commonly

encountered interior and personal light sources was also considered, in order to provide

additional data against which the blue-rich light emitted by LED streetlights could be

compared. The blue light emitted from LED streetlights as well as the commonly used light

sources was analysed in terms of its potential effect on eye health (Chapter 9), as well as on

circadian rhythms and sleep patterns (Chapter 10). The impact that LED streetlighting may

have on light pollution and specifically on blue light at night is discussed in Chapter 11,

covering issues that have attracted significant media attention over recent years, such as Sky

Glow and the impacts of blue light on wildlife. The potential issues of discomfort and disability

glare arising from LED lighting are covered in Chapter 12, along with consideration of the

impact of different LED streetlight CCT on driving performance measures; this section is

however, brief given the paucity of evidence in this area.


7 Photometric concepts of blue light

As a background for this discussion on the impacts of blue-rich light from LED streetlights of

different Correlated Colour Temperatures (CCT) and from commonly used light sources, this

chapter introduces a number of key photometric concepts including:

• Spectral Power Distribution

• Correlated Colour Temperature

• Radiance and Irradiance

This chapter also details the relevant published standards and guidelines for blue light

assessment for electric lighting, specifically related to the following impacts and risks:

• Blue Light Hazard for retinal injury

• Circadian Active Light, known as the Melanopic Response

7.1 Key Photometric Concepts

7.1.1 Spectral Power Distribution (SPD)

Spectral power distribution generally refers to any graphical representation of the radiant power

of a light source at each wavelength in the electromagnetic spectrum.

SPDs of white LEDs: Commercially available white LEDs for general illumination can use

two distinct means to produce white light:

• Colour addition – multiple coloured LED chips are used in combination, e.g. Red

LED + Green LED + Blue LED (+ possibly other LED) = White light, sometimes

referred to as RGB-type;

• Phosphor-based emission – single short wavelength LED pumps are used to excite

phosphor coatings in the LED package, e.g. Blue LED pump + Yellow phosphors =

White light.

Of these 2 types, the most suited for general illumination and particularly streetlighting is the

phosphor-type white LED. The majority of white LED products available use a blue LED pump

with yellow phosphors, which leads to a distinct blue peak in their SPD, as shown in Figure

7-1. However, there is increasing development in the application of violet LED pumps in place

of blue, which alters the SPD of the white LED to reduce its blue light content as shown in

Figure 7-1.


Figure 7-1: Comparison of LED SPDs to 2700K incandescent lamp SPD (left) and 6000K daylight (right). The top two graphs show a blue-based white LED spectral output, the bottom graphs illustrate a violet-based white LED output.

7.1.2 Correlated Colour Temperature (CCT)

A blackbody radiator is an idealised physical object that emits radiation across all wavelengths.

The colour appearance of a blackbody radiator depends on its temperature. The series of SPDs

for blackbody radiators shown in Figure 7-2 demonstrate how radiant power varies with

temperature, such that the wavelength of the peak of the output curve decreases (shorter

wavelengths towards the blue end of the spectrum) with increasing temperature of the

blackbody radiator. This corresponds to a cooler (bluer) white light for higher temperatures.

Conversely a lower temperature blackbody radiator will have a warmer (redder) white light.

300

2700K

No Violet

400

2700K

No UV

400

Blue

500 600 700 800 300 Wave length (nm}

No

500 600 700 800 Wavelength (nm)

Deo

No UV

No V iolet

400

400

Blue overshoot

Cyan Gap

500 600

Wavelength (nm)

500 600 Wavelength (nm)

Red Lack

700 800

700 800


Figure 7-2: Spectral Power distributions for blackbody radiators at different temperatures [from (Ling, Sanny et al. 2018)]

For blackbody radiators, the colour corresponding to body temperature can be represented by

the blackbody curve (also known as the Planckian locus) on the CIE 1931 (x, y) colour chart

as shown in Figure 7-3.

Figure 7-3: CIE 1931 Colour Chart showing the blackbody curve (in red) and the location of blackbody radiators at different temperatures. [from (Pro-Lite Technologies 2018)]

u ltraviolet Visible Infrared

• ,\maximum

0 1.0 2.0 3.0

Wavelength A (µ,m)


Correlated Colour Temperature (CCT) is the term given to the colour appearance of white light

sources that do not emit characteristic blackbody radiation (non-blackbody light sources such

as gas discharge lamps or LEDs). Such sources do not generally lie on the blackbody curve;

but when plotted on the same CIE (x, y) colour space, are still close enough to the curve to

appear similar in colour to a blackbody radiator. The CCT of such a non-blackbody light source

is equal to the actual temperature (in Kelvin) of the closest blackbody radiator in the region of

the non-blackbody light source. Figure 7-4 shows a magnified view of a section of the

blackbody curve with nominal CCT regions marked according to ANSI (black polygons) and

MacAdam ellipse (pink ovals) definitions.

Figure 7-4:Magnified view of the blackbody curve on the CIE 1931 Colour Chart, showing the nominal CCT regions according to ANSI and MacAdam ellipse definitions. [from (ANSIC78.377 2011).

It should be noted that CCT is a description of the colour appearance of a light source only,

and does not give detailed information about the spectral composition of a light source (which

requires knowledge of the SPD), nor does it provide information regarding the ability of the

source to accurately reproduce colour in illuminated objects (colour rendering).

7.1.3 Radiance/Irradiance

The quantity of electromagnetic radiation emitted by a light source can be measured by its

radiance. Radiance is defined as the radiant flux emitted by the source per unit solid angle, per

unit projected area (measured in W.sr-1.m-2). This quantity can also be expressed as a measure

of irradiance, which is the quantity of radiant flux from the source that is received on a detector

y

CIE 1931 x,y Chromaticity Diagram 0.46 .-------------------------,

0.44

0.42

0.40

0.38

0.36

0.34

0.32

0.30

3500K

Z700K

• lbrinantA

D65

Plonclaan bcu5

0.28 ------------------------0.26 0.28 0.30 0.32 D.34 0.36 0.38 0.40 D.42 0.44 0.46 D.48 0.50 D.52

X ANSI C78.377A


surface per unit area, and is given in W/m2. The two quantities are linked geometrically.

Radiance is the radiometric (broadband) equivalent of the photometric quantity of luminance.

The luminance (L) of a light source is the radiance of the source adjusted to account for the

sensitivity of the human eye (i.e. modified by the photopic relative luminous efficiency (V(λ)

curve)). Similarly, the illuminance provided by a light source is a photopic modification of the

irradiance measured at the same point.

While the luminance and illuminance measured from a light source are required to assess the

visual effectiveness of the source, radiance and irradiance are also necessary, in order to

consider non-visual effects (such as the blue light hazard).

7.2 Standards and guidelines for photobiological impact of electric light sources

IEC 62471 ((IEC) 2006): Photobiological safety of lamps and lamp systems lists the

photobiological hazards for electric lighting across all wavelengths of radiation emitted by

lamps (from ultraviolet through to infrared), including:

• Actinic UV hazard to skin and the eye (due to radiation from 200nm - 400nm)

• Near UV hazard to the eye (315nm – 400nm)

• Retinal blue light hazard (300nm – 700nm; peak range 380nm – 550 nm)

• Retinal thermal hazard (380nm – 1400nm)

• Infrared radiation hazard to the eye (780nm – 3000nm)

• Thermal exposure hazard to the skin (380nm – 3000nm)

Of these, the only hazards that are relevant to the light in the blue wavelength range are the

retinal blue light and thermal hazards. However, the retinal thermal hazard is not a realistic

concern in a streetlighting scenario, given that the only general light source that can trigger

retinal thermal damage is a xenon arc focused directly onto the retina (Bullough 2000);

therefore this hazard is not considered further. Furthermore, while the wavelength range for

thermal exposure hazard to the skin contains blue wavelengths, in the context of streetlighting,

the thermal exposure to the skin, which includes tissue burns due to the proximity of high levels

of radiation, is not a relevant or realistic concern (and according to the standard cannot be

evaluated in this type of open environment).

There are no additional standards that provide exposure limits for blue light emitted from

electric light sources. However, there are emerging recommendations that relate to blue light


as a circadian stimulus. Conflicting opinions exist regarding the best way to quantify the

potential impact of a light source on the human circadian cycle and which threshold values are

appropriate. The impact of artificial light sources on human circadian health is a growing

concern in health and lighting fields, and so the existing work on quantifying circadian light

stimulus will be considered.

7.3 Measuring the impact of blue light

Blue light is broadly considered as electromagnetic radiation in the range of 430 to 500 nm.

However, more technical definitions are required if the impact of blue light on health and

environmental conditions is to be appropriately assessed.

The potential hazard due to blue light from a light source is mediated by two key characteristics

of the light source: the source’s spectral power distribution, and its spectral radiance (or

irradiance). However, the actual impact of the potential hazard on a person, animal or the

environment is dependent on the action spectrum (or spectral sensitivity) of the biological or

environmental effect, the threshold level for the effect to be realized and the duration of

exposure.

The general procedure for quantitatively evaluating a light source for its impact on health and

the environment is as follows:

1. Measure and record the spectral radiance of the light source;

2. Identify the action spectrum defined for the targeted biological or environmental

effect;

3. Calculate the power the light source emits, weighted by the action spectrum for the

effect;

4. Estimate the duration of exposure;

5. Compare the resulting hazard-weighted energy emitted by the source with known

effect threshold values to assess the impact of the light source.


8 Field-based measurement methodology

Field measurements of six (6) different street lighting installations were made 12th –14th June

2018 at several locations in Brisbane. Measurements coincided with the new moon phase (13th

June) to minimise the effects of ambient moonlight, and were taken under both clear and

overcast skies. Measurements included spectral irradiance, illuminance and luminance to

assess light at the eye from the perspective of drivers, pedestrians and cyclists and also to assess

spill light surrounding the streetlights.

The street light technologies included in this report are:

• 3200K LED

• 4000K LED

• 5900K LED

• 70W High Pressure Sodium

• 50W Mercury Vapour

• 32W Compact Fluorescent

Measurements were also made of spectral irradiance and illuminance at the eye for the

following common night lighting scenarios and personal light sources, as shown in Figure 8-1

and Figure 8-2 :

• Watching television (no other lighting) from 3.5 m viewing distance

• Using a computer screen (no other light source) from 0.7 m viewing distance

• Using a laptop screen (no other light source) from 0.5 m viewing distance

• Using a smartphone device (no other light source) from 0.3 m viewing distance

• Interior lighting, kitchen – 3500K Linear Fluorescent

• Interior lighting, kitchen – 3000K LED Downlight

• Interior lighting, utility – 4000K LED Downlight


Figure 8-1: Examples of the personal light sources - TV (upper left), laptop screen (upper right), computer screen (lower left), smartphone (lower right)

Figure 8-2: Examples of the interior lighting - 3500K Linear Fluorescent (left) and 3000K LED Downlight (right)

Note that we were unable to obtain accurate measures of the spectral properties of in-vehicle

dashboard lighting due to their very low light levels, and these data are therefore not reported.

Measurement of illuminance and spectral irradiance for the exterior, street lighting assessments

were made with a Konica Minolta CL-500A Spectrophotometer. The illuminance measurement

range for this instrument is 0.1 – 100,000 lux and the spectral range is 360 – 780 nm, with a

bandwidth of 10 nm. The instrumental accuracy reported for this meter is ± 2% ± 1 digit of


displayed value. Measurement of illuminance and spectral irradiance for the interior lighting

scenarios and personal light sources were made with an Asensetek Light Passport Pro

Spectrophotometer. The illuminance range for this meter is 5 – 50,000 lux, the spectral range

is 380 – 780 nm and the spectral bandwidth (optical resolution) is 8 nm. The instrumental

accuracy for this meter is reported as ± 3% for illuminance.

Luminance measurements were made with a Konica Minolta LS-100, which has a

measurement range from 0.001 – 299,900 cd/m2 using a 1° acceptance angle. The instrumental

accuracy reported for this meter is ± 2% ± 1 digit of displayed value for measurements greater

than 1 cd/m2.

Data recorded in the field in the field can be broadly grouped into activity-based measures

(measuring light at the eye for drivers, pedestrians and cyclists using the road) or spill light

assessment (measurements made with respect to residential boundaries).

Figure 8-3 shows a schematic diagram (not to scale) of spectral irradiance measurements for

the different simulated road uses, with measurements made at 10 m intervals starting from

directly underneath one pole (start position distance 0 m) up to a distance of 50 m away from

that pole, while facing the direction of travel (shown by coloured arrows) from a typical vertical

height for the road use: 1.5 m for pedestrians and cyclists and at ~1.0 m inside a car for drivers.

The pedestrians’ perspective was measured walking along the footpath closest to the pole

(except for Lilley Rd Bardon location with 3000K LED, where the closest footpath was

obstructed and so the opposite path was used). Cyclists were measured at the road edge closest

to the pole, and drivers in the lane closest to the pole (except for Innes Rd, Geebung location

with Mercury Vapour, where the opposite lane was used). Where the pole spacing was less

than 50 m, readings were made to 40 m only.

Figure 8-4 shows locations of measurement for the spill light assessment. Readings of

luminance and spectral irradiance were taken along the roadway at 1.5 m above the ground

level (highlighted in red), every 3 m between adjacent luminaire poles (up to 66 m for 3200K

LED). The green crosses show luminance and spectral irradiance readings made at a height of

3.0 m, facing the luminaire. These readings were taken in line with the luminaire perpendicular

to the roadway (in 2 m intervals towards the nearest residential boundary, and in 4 m intervals

towards the opposite residential boundary), and along the property boundary parallel to the

roadway.

Page 23 of 84: Blue Light and Public LED

Lighting

Figure 8-3 Schematic m

ap of activity-based measures – crosses indicate m

easurement points

at 1.5m above the ground for cyclists and pedestrians, and at ~1.0m

above the ground for drivers (inside car). C

oloured arrows indicate direction of travel.

Figure 8-4 Schematic m

ap of spill light assessment m

easures – red crosses indicate a m

easurement point at 1.5m

above the ground and facing towards the lum

inaire – red crosses continue in 3m

intervals until the next luminaire (up to 66m

). Green crosses indicate a

measurem

ent point at 3.0m above the ground and facing tow

ards the luminaire.

Roadw

ay

Footpath

Driver

Cyclist

Pedestrian

Luminaire

Pole

0 m

10 m

20 m

30 m

40 m

50 m

Readings continue

every 3 m to next

luminaire

Luminaire

Pole

+15 m

+

30 m

+4 m

+3 m

+12 m

Near residential boundary

Opposite residential boundary

+8 m

+6 m

-4 m

-2 m

Roadw

ay

® ® ® 0 • I &• 0

0 10 0 ®

® 0 I @ 0

® I ® I ®

® I

0 I@ 0

®

® 0 0 0

® I ® I ® i i i 0 0 0

I

®

®

..


The results and analysis of the field measurements are presented in this report in their relevant

sections, according to the BLH for retinal injury (Chapter 9), and circadian impact (Chapter

10). All data collected in the field and interior assessments and photometric modelling of spill

light are tabulated in Appendix 1.


9 Blue light and retinal damage

While light can penetrate deep into the layers of the eye, the ocular tissues absorb or transmit

selected wavelengths before they reach the retina (Sliney 2002, Sliney 2005). Knowing which

wavelengths are absorbed and those that are transmitted through the various ocular tissues is

important in understanding the potential negative impact of light, particularly the effects of

blue-rich light on eye health.

The cornea protects the ocular structures, including the retina, from ultraviolet radiation (UVR)

of wavelengths less than 300nm, while the crystalline lens blocks UVR between 300 – 400nm

(Boettner and Wolter 1962). With increasing age, the crystalline lens becomes progressively

more yellow and opaque, resulting in further reduction in the transmission of short wavelength

visible light (Norren and Vos 1974, Mellerio 1987, van Norren and van de Kraats 2007), as

shown in Figure 9-1.

Figure 9-1: Illustration of the age-related change in transmission of different wavelengths as a function of age. Blue box indicates the blue light hazard range [adapted from (Turner and Mainster 2008)]

Thus, UVR light does not contribute to the blue light hazard in eyes with a normal crystalline

lens (phakic), as wavelengths shorter than 400nm are absorbed by the crystalline lens of the

eye. Aphakic individuals, in whom the crystalline lens has been removed as result of cataract

C 0

".j::i Cl]

C

Ew ::J >-

<l)

- "O Cl] -c::: 0

'.j::i ,.._ Cl) ct! ,.._ <l)

>-~o u..-Cl) ct! Q. 0 r.n ...

"O <l) Cl) > ..... ·-..c ... C) .!2 ·- <l)

Cl) ... ~-..!. ·o.. ::J

Cl..

1.0

0,8

0.6

0.4

0,2

0

350 400 450 500 550 600 650

Wavelength (nm)

10 years old 15 years old

25 years old

35 years old

45 years old 55 years old

65 years old

75 years old 85 years old 95 years old

700


surgery, would be vulnerable to the effects of UVR. However, true aphakia is rare, as the

majority of individuals undergoing cataract surgery have an artificial intraocular lens (IOL)

implanted to replace the power of the crystalline lens; all modern IOLs have UV blocking filter

properties (Mainster 2005). Blue-light filtering, also termed ‘blue-blocking’, IOLs, as well as

blue-blocking ophthalmic spectacle lenses have been introduced over the last few decades and

are designed to selectively attenuate the transmission of UV radiation and short-wavelength

visible light entering the eye (Mainster 2006). Blue-blocking IOLs typically attenuate the

transmission of about half of incident short-wavelength light, depending upon their dioptric

power (Mainster 2006). However, a recent Cochrane review (Downie, Busija et al. 2018) failed

to show any clear evidence that blue-blocking IOLs preserve the health of the retina,

specifically, damage to the central region of the retina, known as Macular Degeneration (MD),

nor did they find any improvements in visual function, despite several papers claiming to the

contrary. A recent systematic review also reported a lack of high quality evidence supporting

the suggestion that visual performance or sleep quality were improved through the wear of

blue-blocking spectacle lenses (Lawrenson, Hull et al. 2017).

Another important optical filtering mechanism in the eye is the presence of yellow carotenoids

(lutein, zeaxanthin, and meso-zeaxanthin) in the inner retinal layers of the macular region,

collectively known as the macular pigment (Arunkumar, Calvo et al. 2018). These pigments

have a peak absorption at 460nm and thus absorb the short-wavelength light which protects the

photoreceptors from photo-oxidative damage. As stated by Bullough, Bierman et al. (2017),

the presence of macula pigment reduces the risk of damage by the blue-light hazard by a factor

of two compared to peripheral regions where the pigment is absent.

Figure 9-2 provides an overview of how different wavelengths of light are absorbed/scattered

by different ocular tissues and how they potentially impact on the retina and circadian-related

responses.


Figure 9-2: Summary of the various interactions and impact of light of different wavelengths of light on the eye

9.1 Mechanisms of retinal phototoxicity

Research in animal models (Noell, Walker et al. 1966, Ham, Mueller et al. 1982) and cell

culture studies (Davies, Elliott et al. 2001, Sparrow, Miller et al. 2004) has shown that blue

light can induce phototoxic damage in the retina. Two mechanisms of phototoxic damage have

been identified. These mechanisms have slightly different absorption or action spectra (both

peak in the blue), and have been termed “Noell damage” (Type I) and “Ham damage” (Type

II) after the researchers who originally identified these effects.

Type I photochemical retinal damage occurs during extended and less intense exposure to white

light. This damage involves cellular disruption following prolonged bleaching of rhodopsin,

Wav elength (nm) 32,0 34 0 3 60 380 400 420 440 460 480 500 52.0 540 560 580 600 620 64 0 660 680 7 00

ULTRAVIOLET VISIBLE LIGHT

Iris (Absorption/heat dissipation)

Crystalline Lens Interactions of (Absorption/heat dissipation)

light ·with ocular

t issue1

Retinal Photoreceptors (Absorption)

Retinal Pigment Epithelium (RPE) (Absorption/heat dissipation)

Impact of light on .._ ______ _

the eye1 ,3

I: Beh ar-Coh en et al (2011), 2: ICNIRP (2013), 3: Clark and Lesniak (2017)


and occurs initially in the outer segment of the photoreceptors, followed by changes in the

retinal pigment epithelium (RPE). The cone photoreceptors are more vulnerable to damage

than the rods, with the short wavelength or blue cones appearing to be most sensitive to damage

(Noell, Walker et al. 1966, Sperling, Johnson et al. 1980, Wu, Seregard et al. 2006).

Conversely, Type II damage occurs following relatively short and high intensity exposures

(between 10 secs and 2 hours), principally in the wavelength range of 380 – 550nm; with the

range extending into the UV sector of the spectrum 300 – 550 nm in aphakic eyes which have

no UV absorbing crystalline or intraocular lenses (Ham, Mueller et al. 1976, Lund, Stuck et al.

2006). Shorter wavelengths (around a peak action spectra of 440 – 460nm) result in more

intense damage, with the primary locus being in the RPE (Ham, Ruffolo et al. 1978). This

damage is believed to lipofuscin-mediated at the level of the photoreceptors, thus affecting the

ability of the RPE to provide nutrients to the photoreceptors, thus affecting their viability

(Steinberg 1994), and is known as the Blue Light Hazard (BLH) as described in the following

section.

9.2 Blue-Light Hazard (BLH)

The Blue-Light Hazard (BLH) is defined by the potential of a source to cause retinal damage

due to short wavelength light (in the blue to violet range). The action spectrum of the BLH is

defined by a characteristic function, B(λ), as shown in Figure 9-3. The spectral curve ranges

from 380 – 550 nm in wavelength, with peak values between 440 – 460nm.

This action spectrum can be applied as a weighting function to the spectral power distribution

of a light source to yield a measure of the BLH potential of the source.

The BLH weighted radiance of a light source, LB, is given (in units of W.m-2.sr-1) by:

Where Lλ is the spectral radiance, B(λ) is the blue light hazard action spectrum, and Δλ is the

bandwidth in nm.

Within IEC 62471:2006, exposure limits are recommended for LB, with the intent that

‘individuals in the vicinity of lamps and lamp systems shall not be exposed to levels exceeding

the limits’. Exposure limits are set according to the expected duration of source exposure, t, to

prevent retinal damage for an observer. These limits are summarised in Table 9-1.

700

LB= ~),i · 8(2) · LU 300


Figure 9-3: Retinal Blue Light Hazard B(λ) curve, and Retinal Thermal Hazard R(λ) curve [from (International Electrotechnical Commission 2014)]

Table 9-1: Emission limits for BLH and BLH small source from IEC 62471 Photobiological safety of lamps and lamp systems

Hazard Name Relevant EquationWavelength range (nm)

Exposure duration (s)

Emissions Limit in terms of constant

irradiance (W.m-2)< 100 100/t

> 100 1.0

Hazard Name Relevant EquationWavelength range (nm)

Exposure duration (s)

Emissions Limit in terms of constant

radiance (W.m-2.sr-1)0.25 - 10 106/t10 - 100 106/t

100 - 10000 106/t> 10000 100

300 - 700Blue Light

Emission Limits - irradiance based values

Blue Light, small source

(<0.011 radian)300 - 700

Emission Limits - radiance based values

>, CJ Ill CJ :E Q)

iu ,_ -CJ Q) Q.

(/)

1E+01

1E+OO

1 E-01

1E-03

:l t ---r

a

J:

~-~-~~~---~---~---~---~----~---~----• I I I I

I I t I I I I I

-- I -- - -- --~ - -- ~ - -- -!- ---~ ---~ ---~ --\- ~ --- ~ ----t

--~---·----~---~---~---~--------~----• I I I I I I I

B I I I I I I ~-----------------------·---~----

1 E-04 .......... ~ ........... ~ ........... ~-~.....+-,~.....+-~ ........... ~ ......... ~ .......... ~.,....,.....~~.........,....,....,.....

300 400 500 600 700 800 900 1000 1100 1200 1300 1400

Wavelength [nm]

700

Ea= LE,i -B(l)·M -

300

700

La= L/,i -B(J)-M 300


For light sources that are sufficiently small (subtending an angle < 0.011 radians at the eye), a

similar limit can be set based on the spectral irradiance measured at the eye, Eλ. This

measurement is used to calculate the BLH weighted irradiance of a source, EB, which is given

(in units of W.m-2) by

There are similarities between the Luminance, L, of a source and its LB value. Both quantities

are determined from the spectral radiance Lλ values; however in the case of L, the spectral

radiance is multiplied by the photopic curve, V(λ), instead of B(λ). For any given light source

spectrum, there will be a constant relationship between the values of L and LB. Similarly,

illuminance, E, is based on spectral irradiance measurements (modified by the photopic curve)

so that for any given spectrum, E will be proportional to EB. These proportionality factors are

the same for any single spectrum, and called the blue light hazard efficacy of luminous

radiation, and denoted by the symbol KB,v. This quantity is a useful measure of the blue light

content of a light source, as it indicates the quantity of BLH radiation per unit visible radiation.

The emission limits shown in Table 9-1 above address the threshold quantities for risk of hazard

and are defined for individuals in the vicinity of light sources; assessment below these values

suggest the situation presents no or low risk of BLH injury. As a general method of classifying

lamps, IEC 62471:2006 defines risk categories for lamp systems as shown in Table 9-2 below.

The measurement of radiance or irradiance from lamp systems for BLH risk classification is

generally conducted at a distance of 200 mm from the source, or at a distance such that

illuminance is 500 lux.

Table 9-2: Emission limits (EL) defining risk groups

Exempt Low riskModerate

riskBlue Light L B 100 10000 4000000 W.m-2.sr-1

Blue Light, small source E B 1.0* 1.0 400 W.m-2

* Small source defined as angular subtense < 0.011 radians.

RiskAction

Spectrum SymbolEmission limit

Units

B(λ)

700

E8 = LE..t -B(A) -M 300


The risk groups and names and hazard level descriptors are shown in Table 9-3 below.

Table 9-3: Risk Groups (RG) for lamps defined in IEC 62471

Risk group name Risk group number Philosophical basis

Exempt RG0 No photobiological hazard

Low risk RG1 No photobiological hazard under normal behavioural limitations

Moderate risk RG2 Does not pose a hazard due to aversion response to bright light or thermal discomfort

High risk RG3 Hazardous even for momentary exposure

9.2.1 The problem with using CCT as an indicator of Blue-Light Hazard

Figure 9-4 below shows the SPDs for four different examples of light sources, all with a CCT

of 3000K. The proportion of light output in the BLH range by each source ranges from 7.8%

(incandescent) to 16.5% (warm white fluorescent) of the total light power. This example

highlights the failing of CCT as a proxy measure for the blue light content of light sources.

In this specific example, two different light source technologies with the same CCT are shown

to differ in their percentage of blue light hazard content by a factor of two. In this case, the

colour appearance of the sources would be the same, but importantly, the proportion of blue

light is not.

In general, as CCT increases for an LED light source, the proportion of blue light in the SPD

also increases. In IEC Technical report IEC/TR 62778 regarding the assessment of the blue

light hazard (International Electrotechnical Commission 2014), a chart of the blue light hazard

efficacy of luminous radiation, KB,v, vs CCT for a range of light sources (reproduced in Figure

9-5 below) shows this general trend.


Figure 9-4: Example spectral power distributions for 4 different light sources with identical CCT (all 3000K) showing the variation in blue light content (blue light ranging from 7.8 – 16.5% of source output) [source: (CELMA-ELC 2011)]

Figure 9-5: Graph of KB,v vs CCT for a range of light sources, demonstrating a linear trend indicating increased proportion of blue light with increasing CCT. [(International Electrotechnical Commission 2014)]

100

.; 80 J 0 a. g 80

{ rn 40 Cl) > ;:,

; 20

100

.; 80 J 0 a. g 60

~ Q. rn 40 Cl) > ;, .. ; 20

Incandescent 3000K

300 400 500 800 700 BOO Wavelength (nm)

Ceramic Metal Halide 3000K

300 400 500 800 700 800 Wavelength (nm)

0,001 6

0,001 4

0,001 2

E 0,001 0

• [ 0,000 8

> .ff. 0,000 6

0,000 4

0,000 2

0

100

.; 80 J 0 a. ~ 80

8. rn 40 Cl) > ;:,

& 20

100

.; 80 J 0 a. g 60

~ Q. rn 40 Cl) > i & 20

Fluorescent (Tri) Warm White 3000K

LED Warm White 3000K

300 400 500 800 Wave length (nm)

■

• Halogen and incandescent

■ Fluorescent

HID

LED

Daylight

700

0 2 000 4 000 6 000 8 000 10 000 12 000 14 000 16 000

Correlated colour temperature (Kl

800


Therefore, it is safe to assume that an LED chip with a CCT of 6500K would emit more blue

light than a chip with CCT of 3000K (if matched for luminous output). However, it is possible

that a 4000K LED chip might emit less blue light than a 3500K chip matched for luminous

output, given variations in their relative SPD due to the mix of phosphors used.

9.3 Blue light hazard and Macular Degeneration

As described in a previous section, animal and laboratory studies have shown that blue light

may damage the retina. It has been suggested that the BLH may be a factor in the pathogenesis

of Macular Degeneration (MD) which occurs in the macular region of the retina and affects

central visual function (fine detail and colour perception). Epidemiological evidence shows

that higher levels of lifetime sunlight exposure are associated with higher prevalence of MD

(Taylor, Munoz et al. 1990, Algvere, Marshall et al. 2006, Sui, Liu et al. 2013). However, other

studies have failed to find such associations (Darzins, Mitchell et al. 1997) and only one study

has specifically reported an association between higher blue light exposure (from sunlight) and

neovascular MD in humans, and this was only evident for adults with low levels of antioxidants

(Fletcher, Bentham et al. 2008). Determining the extent of the causative nature of these

associations is, however, problematic given the challenge in determining an individual’s

cumulative life-time overall sunlight exposure, as well as blue light-specific sunlight exposure.

Further, MD is a complex multifactorial disorder involving genetics, inflammation, health

behaviours (e.g. smoking), atherosclerosis and numerous other factors (McCarty, Mukesh et

al. 2001).

The mechanism through which any long-term high-intensity light exposure induces

photoreceptor damage is also less well known. Several studies have indicated that lipofuscin

(peak absorption around 450nm) is a possible mediator of the risk associated with long-term

exposure to blue light induced retinal damage (Wolf 2003, Sparrow and Boulton 2005). When

lipofuscin absorbs blue light, free radicals are produced which are responsible for the oxidative

damage that occurs in the retina. The number of free radicals produced by lipofuscin is

wavelength dependent, and decreases from 400 to 490 nm (Pawlak, Rozanowska et al. 2002).

The accumulation of lipofuscin in the RPE has been linked to death of photoreceptors and

development of MD (Wolf 2003). In addition, the amount of lipofuscin present in the RPE

increases with age, thus the potential for blue light to damage the retina may also increases

with age (Delori, Goger et al. 2001), although this may be offset by increasing absorption of

shorter wavelengths by an ageing crystalline lens, as discussed previously.


Importantly, the risk from artificial light sources (either broad spectrum or ‘blue-rich’), as

opposed to sunlight exposure in humans and the development of MD, is much more difficult

to determine. Nevertheless, some organisations, including the American Medical Association

(AMA) have expressed concerns that LED sources may increase the risk of impaired vision

from factors such as the BLH and have provided a series of recommendations including those

relating to the installation of LED streetlighting (CSAPH 2016). However, Bullough, Bierman

et al. (2017) highlighted that an increase in the relative magnitude of blue light in an artificial

light source does not necessarily translate to any meaningful difference in its BLH risk. Indeed,

from their analysis of the spectral radiant power characteristics of incandescent, fluorescent,

LED and daylight sources, in terms of the BLH for phakic, aphakic and pseudophakic eyes,

they concluded that LEDs do not provide greater risk for the blue-light hazard than other

sources (e.g., incandescent) and for aphakic eyes they present lower risk given their limited

production of light in the UV range (Bullough, Bierman et al. 2017). Under normal

circumstances, the photophobic response of the human eye limits exposure to bright light

sources, as does pupil constriction in response to bright lights and the normal involuntary eye

movements that naturally occur to maintain normal visual perception, spread retinal irradiance,

which reduces the maximum localized irradiance incident on the retina over time (CELMA-

ELC 2011, Energy 2013). However, caution should be exercised in relation to the BLH in the

following situations (CSAPH 2016, Bullough, Bierman et al. 2017): where photophobic and

aversion responses are lacking (eye surgery patients or infants), for those for whom intentional

light exposure might be anticipated (actors staring into bright lights) and for those with lupus

or certain eye conditions, such as those with aphakia, although this condition is very rare.

9.4 Field-based measurements of LED Streetlights

For activity-based measures, measurements of irradiance at the road user’s eye were made from

positions simulating driving, cycling and walking tasks under the six different light source

types: 3200K, 4000K and 5900K LED, High Pressure Sodium (HPS), Mercury Vapour (MV)

and Fluorescent (FL).

Spectral irradiance measurements (W.m-2.nm-1) were collected by the Konica Minolta CL-

500A at 1 nm intervals, and these values were multiplied by the BLH curve (at 1 nm intervals

using a linear interpolation of the 5 nm data provided in IEC 62471) to calculate the blue-

weighted irradiance EB. Table 9-4 details the photopic illuminance, Ev and blue-weighted

irradiance measured at the eye for each activity type across the range of measurement locations.


Table 9-4 Readings of blue-weighted irradiance and photopic illuminance for driving, cycling and pedestrian activities; the highest blue-weighted irradiance measurements for each source are italicised.

The irradiance values recorded for these activities are generally maximal at approximately 10

m from the base of the luminaire, which is around the closest distance at which the luminaire

appears in the field of view of the road user. Cycling has the highest values recorded, they are

Distance (m) 3200K LED 4000K LED 5900K LED HPS MV FL

0 0.00000 0.0001 0.00004 0.00003 0.00001 0.0000110 0.00058 0.0017 0.00073 0.00020 0.00028 0.0009820 0.00019 0.0007 0.00077 0.00010 0.00018 0.0003830 0.00007 0.0002 0.00027 0.00002 0.00009 0.0001040 0.00001 0.0001 0.00013 0.00001 0.00004 0.0000550 0.00000 0.00008 0.00000 0.000030 0.03 0.16 0.09 0.50 0.03 0.05

10 1.97 3.82 0.98 2.12 0.57 2.0520 0.81 1.69 1.13 1.26 0.38 0.8030 0.37 0.53 0.44 0.39 0.18 0.2440 0.13 0.32 0.21 0.21 0.10 0.1250 0.01 0.12 0.15 0.06

Driving - values measured at the driver's eye

E B - Blue light hazard weighted irradiance

(W/m-2)

E v - Illuminance (lux)


0 0.00009 0.0002 0.00011 0.00009 0.00006 0.0003310 0.00086 0.0025 0.00192 0.00108 0.00064 0.0015820 0.00024 0.0010 0.00088 0.00024 0.00036 0.0005730 0.00010 0.0004 0.00043 0.00007 0.00018 0.0002140 0.00012 0.0002 0.00025 0.00003 0.00012 0.0001250 0.00003 0.00018 0.00003 0.000120 0.33 0.39 0.22 1.04 0.16 0.71

10 2.88 5.59 2.41 9.64 1.18 3.1720 0.92 2.35 1.22 2.13 0.67 1.1230 0.39 0.91 0.69 0.66 0.35 0.4140 0.30 0.64 0.45 0.36 0.25 0.2650 0.10 0.31 0.36 0.28

Cycling - values measured at the cyclist's eye


(W/m-2)



0 0.00001 0.0001 0.00009 0.00015 0.00005 0.0000710 0.00011 0.0012 0.00221 0.00073 0.00063 0.0008020 0.00023 0.0007 0.00100 0.00019 0.00026 0.0003430 0.00011 0.0003 0.00043 0.00005 0.00013 0.0001540 0.00009 0.0002 0.00023 0.00002 0.00009 0.0000950 0.00004 0.00015 0.00002 0.000130 0.03 0.28 0.20 1.59 0.17 0.18

10 0.39 3.34 3.27 6.81 1.14 1.5720 0.90 1.89 1.46 1.79 0.51 0.6730 0.48 0.81 0.68 0.53 0.28 0.3240 0.33 0.53 0.42 0.23 0.24 0.2050 0.18 0.31 0.28 0.35

Walking - values measured at the pedestrian's eye


(W/m-2)



closer to the luminaires than pedestrians and have no loss of light through the car windscreen

as is the case for drivers.

The maximum illuminance and spectral irradiance readings for each luminaire collected in the

field were the horizontal illuminances measured directly beneath the luminaire (nadir) at a

height of 1.5 m above the ground. This reading is significant as it is (geometrically) the least

influenced by external or ambient (other) light sources and for this reason, along with the strong

signal at this position permitting relatively low instrument noise, it is considered the most

accurate reading of the spectral irradiance of each luminaire in the field. These results are

presented in Table 9-5 below, and the spectral data is used to calculate the quantity KB,v - the

blue light hazard efficacy of luminous radiation for each source. These KB,v values are

consistent with values found for other sources with similar CCTs as shown in Figure 9-5.

Table 9-5: Maximum field readings of horizontal illuminance at nadir (h=1.5 m), with corresponding calculated blue-weighted irradiance, and their ratio, KB,v

Of note, the KB,v value measured for this specific 4000K LED luminaire (0.00054) is similar to

that of the selected Mercury Vapour and Fluorescent luminaires (0.00056 and 0.00051

respectively), indicating that they have similar proportions of BLH content in their SPDs.

Although spectral radiance was not directly measured, luminance values were recorded in the

field for the spill light measurement locations (shown in Figure 8-4) Using the luminance

values measured on site, L, and the ratio, KB,v, the blue-weighted radiance, LB, can be calculated

(since the ratio of blue-weighting to photopic-weighting is the same for irradiance and

radiance). The maximum L and LB values measured in the field are shown in Table 9-6.

Table 9-6: Maximum readings of luminance in the field for each luminaire type and equivalent blue-weighted radiance, LB

3200K LED 4000K LED 5900K LED HPS MV FLIlluminance (lux) 9.4 13.2 5.1 22.9 4.6 9.5

Blue Light weighted irradiance E B (W/m-2)

0.0032 0.0072 0.0045 0.0026 0.0026 0.0048

K B,v 0.00034 0.00054 0.00088 0.00011 0.00056 0.00051

3200K LED 4000K LED 5900K LED HPS MV FLMax Luminance (cd.m-2) 25000 41000 24000 56400 7500 17000

Max L B (W.sr-1.m-2) 8.4 22.2 21.2 6.3 4.2 8.6


Overall, these illuminance and luminance values are typical of street lighting scenarios in

residential areas of Brisbane. It should be noted that the illuminance values at nadir are not

matched across each luminaire type, with the HPS value being substantially higher (~2 times

greater) than the mid-range values of 3200K LED, 4000K LED and FL; and the 5900K LED

and MV values approximately half the mid-range values.

The evaluation of a light source’s blue light hazard for retinal injury is generally done by blue-

weighted radiance, LB; however, if a source is sufficiently small (i.e. subtends an angle less

than 0.011 radians) then blue-weighted irradiance, EB, may be used. As highlighted in IEC

62778 - Application of IEC62471 for the assessment of blue light hazard to light source and

luminaires (International Electrotechnical Commission 2014), the small source regime

represents a “worse case” in terms of source luminance.

The emissions limits that are consistent with low or no risk for BLH at maximum exposure

durations are:

• EB 1.0 W.m-2 (maximum exposure duration > 100 seconds)

• LB 100 W.m-2.sr-1 (max. exposure duration > 10,000 seconds)

Practically, these maximum exposure durations far exceed the realistic exposure times for

anyone using the road (particularly if they are in motion) or viewing the luminaires from nearby

locations. Reducing the exposure duration acts to increase the exposure limit, for example

reducing the duration to 10 second increases the LB limit value to 10,000 W.m-2.sr-1 and the EB

value to 10 W.m-2.

Using the small source assumption for the activity-based measures of road users, it is evident

that the blue-weighted irradiances are all at least 100 times less than the exposure limit of 1.0

W.m-2. The maximum value of EB = 0.0025 W.m-2 (400 times less than the exposure limit) is

found for cyclist activity at 10 m from the luminaire for the 4000K LED scenario. It should be

noted that this limit is set higher than necessary by considering the maximum exposure duration

of more than 100 seconds; the actual exposure duration of a luminaire to a cyclist travelling at

normal speed would be brief.

Similarly, taking the maximum blue-weighted radiances recorded in the field for the spill light

assessments, 22.2 and 21.2 W.m-2.sr-1 for 4000K and 5900K LED luminaires respectively,

measurements are also below the exposure limits for BLH, even when evaluated for an

exposure duration of more than 2.5 hours (>10,000 seconds).


9.5 Comparisons with other light sources

The widespread use of computers and mobile technologies has also raised concerns regarding

the emission spectra of devices such as smart phones and tablets and their potential for causing

blue light damage to the eye. It has been shown that exposure of cultured retinal pigment

epithelium cells to light equivalent to that emitted from mobile display devices causes free

radical production and reduced cell viability (Moon, Yun et al. 2017). Interestingly, the

negative effects of blue light were greater at shorter wavelengths (especially 450nm) while

damage was reduced for blue light with peak wavelength of 470nm. These results raised

concerns regarding the potential effects that cumulative exposure to blue light from these

devices might have in terms of retinal toxicity and the risk of macular degeneration (Tosini,

Ferguson et al. 2016).

Importantly, a recent study (O'Hagan, Khazova et al. 2016) evaluated the potential risk of a

range of these devices by comparing their spectral output against the blue light hazard exposure

limit (International Commission on Non-Ionizing Radiation 2013) and the exposure likely to

be received from staring at a blue sky. The results of this study indicated that even under

extended viewing, none of the sources approached the exposure limits and were lower than

those associated with viewing natural sources such as a blue sky. Interestingly, concerns

regarding the long-term safety effects of the self-luminous LED displays in modern electronic

devices on eye health was also raised in a recent systematic review on the effects of blue-

blocking spectacle lenses (Lawrenson, Hull et al. 2017). However, as stated by these authors,

the weighted spectral irradiance from these devices does not reach international exposure

limits.

To place the blue light level readings from streetlight sources in context, Table 9-7 shows the

BLH quantified for a range of typical interior lighting scenarios and personal lighting devices.

Although many of the screen-based sources are not ‘small sources’ by the angular definition,

this table illustrates the order of magnitude of the difference in impact some of these alternate

sources have, compared to street lighting source scenarios in the BLH wavelength range.


Table 9-7: BLH irradiance values for exterior streetlight sources and typical interior lighting scenarios

Illuminance (lux)

Blue Light weighted

irradiance EB (W/m-2)

KB,v

Exterior Sources Streetlighting, Walking Activity

3200K LED 18W 0.9 0.0002 0.0003

4000K LED 20W 3.3 0.0012 0.0004

5900K LED 17W 3.3 0.0022 0.0007

HPS 70W 6.8 0.0007 0.0001

MV 50W 1.1 0.0006 0.0006

FL 32W 1.6 0.0008 0.0005

Interior Sources Room Lighting, Standing Activity

Fluorescent 3500K 36W 78 0.030 0.0004

LED 4000K 9W 150 0.070 0.0005

LED 3000K 9W 85 0.025 0.0003

Technology Devices, Seated Activity

TV @ 3.5m 5 0.005 0.0010

Computer @ 0.7m 49 0.041 0.0008

Laptop, Max Brightness @ 0.5m 58 0.043 0.0007

Smartphone, Max [email protected] 45 0.047 0.0010 All quantities measured at the eye


9.6 Summary: LED Streetlights and retinal damage

9.6.1 Summary of literature

The existing literature on the potential for BLH from LED sources suggests that blue light can

cause retinal damage, yet this would require exposure at much high intensities and for more

prolonged exposures than is typically encountered in the normal road environment under LED

streetlights.

9.6.2 Summary of the field-based measurements

The field-based analysis considered the amount of BLH present in LED streetlights for road

users and at property boundaries compared with other streetlights and commonly encountered

light sources at night. The absolute quantity of BLH radiation in any scenario depends not only

on the SPD of the source, but also the overall radiant power of the luminaire, and the spatial

distribution of light (e.g. the presence of ‘hot spots’). The six scenarios assessed were

representative in terms of their KB,v values, but were not matched for overall radiant power.

Furthermore, some LED luminaires displayed evident hot spots, while other luminaires did not.

These factors can influence the generalisability of the results to all LED street lighting.

Summarising the key data related to BLH for retinal damage:

• All of the assessed road user scenarios (driver, cyclist and pedestrian) produced blue-

weighted irradiance readings that were more than 100 times lower than the

recommended exposure limits for small sources, even when based on the maximum

exposure duration of >100 seconds;

• The maximum blue-weighted radiance measured for each light source is more than 4

times lower than the exposure limit set for the maximum exposure duration of

>10,000 seconds (more than 2.5 hours);

• The proportion of BLH content in the SPDs measured for the sample luminaires

(represented by the value KB,v) were consistent with expected values. Of note, the

4000K LED streetlights had similar KB,v values to mercury vapour and compact

fluorescent streetlights;

• Radiation in the BLH range arising from normal activities under streetlighting was

lower than for many common interior activities, this is particularly relevant

considering the longer exposure duration of some screen-based activities compared to

activities that involve road users moving through the exterior environment.


In summary, these findings place the BLH risk of LED streetlighting for road users and

residents in a similar category to that of traditional white streetlighting sources (50W Mercury

Vapour and 32W 4000K Compact Fluorescent), and in a lower risk category than interior

lighting and personal electronic devices.

Table 9-8 below, from IEC 62778 (International Electrotechnical Commission 2014) shows

that a 4000K LED streetlight would need to provide illuminance values greater than 850 lux at

the eye to exceed the low blue light hazard risk category (RG1), whereas, maximum

illuminance values for the LED streetlights recorded in this field-based study were below 6 lux

at the eye, across the driving, cycling and pedestrian activities.

Table 9-8: Table C2 from IEC 62778 (International Electrotechnical Commission 2012) showing the illuminance values required to give a risk group not greater than RG1 (low risk)

Rated CCT llluminance E (Ix)

CCT ::; 2 350 K 4 000

2 350 K < CCT ::; 2 850 K 1 850

2 850 K < CCT ::; 3 250 K 1 450

3 250 K < CCT ::; 3 750 K 1 100

3 750 K < CCT $ 4 500 K 850

4 500 K < CCT ::; 5 750 K 650

5 750 K < CCT ::; 8 000 K 500


10 Blue light and the circadian rhythm

The human biological cycle follows the 24-h dark/light cycle, known as the circadian rhythm.

Light exposure plays an important role in setting the circadian regulation of the human

endocrine system (Turner and Mainster 2008). The hormone melatonin is released from the

pineal gland under dim lighting conditions and is involved in the physiological control of sleep.

The release of melatonin is controlled via the pathway driven by the responses of the

intrinsically-photosensitive or melanopsin retinal ganglion cells (mRGCs). The spectral

sensitivity of the mRGCs are used as an action spectrum for the circadian response to yield a

metric known as the Equivalent Melanopic Lux (EML). The mRGC action spectrum is shown

below, along with the photopic curve V(λ) for reference, in Figure 10-1.

Figure 10-1: Melanopic retinal ganglion cells (mRGCs) and photopic spectral response curves.

The EML was introduced in 2011 (Enezi, Revell et al. 2011) and developed by Lucas, Peirson

et al. (2014) as a way of measuring the biological impact of light on processes such as circadian

entrainment. EML is measured on the vertical plane at eye level. This quantity has been applied

in the Well Building Standard (International Well Building Institute 2016), where it forms the

basis of recommendations on the minimum threshold for daytime light intensity for building

interiors (commercial, residential and education areas). The minimum values for these areas,

based on the stimulating properties of light during daylight hours, range from 125 to 250

equivalent melanopic lux. There is no recommendation on maximum values for night time

1.00

\

\ \

I

0.75

~ \

"> \ ·.;:: "cii \ C: I QJ

CJ) 0.50 - Melanopic Response QJ

- Photopic Response > ~ \

ai I Ct'.

0.25

0.00

400 500 600 700

Wavelength (nm)


hours in interior or exterior areas, where the goal may be to minimise inappropriate melatonin

suppression.

The melatonin suppressive effects of light exposure are wavelength and dose-dependent,

greater suppression will occur with increasing light levels and emission of more short-

wavelength (blue) light. For example, low illuminances (below 80 lux) of a white light

(overhead cool white fluorescent lamps) evoked little change in plasma melatonin

concentrations, whereas a room with higher illuminances (greater than 200 lux) completely

suppressed plasma melatonin (Zeitzer, Dijk et al. 2000).

Other researchers have suggested that EML, being based on the response of only one single

photoreceptor, does not provide a comprehensive picture of the mechanisms involved in

melatonin suppression in the body, and therefore cannot accurately gauge the impact of light

on the circadian system (Rea, Figueiro et al. 2005, Rea, Figueiro et al. 2010, Figueiro, Gonzales

et al. 2016). These researchers have suggested two novel measures: Circadian Light and

Circadian Stimulus:

Circadian Light (CLA): is a measure of spectral irradiance of a source, weighted to reflect the

spectral sensitivity of the circadian system assessed by acute melatonin suppression after one

hour of exposure. This spectral sensitivity is shown in Figure 10-2 below.

Figure 10-2: Spectral sensitivity for assessment of Circadian Light (CLA), for warm and cool light sources (by CCT) [from Figueiro, Gonzales et al. (2016)].

1

-- Cool 1 - QJ QJ \ - - Warm u u

\ :i :i 0.75 0 0 \ Vl Vl \ ..... .....

\ 0.75 ..c .c g g \

0 0.5 \ E 0 \ ro u 0.5 ~ \ >,

\ >, u C u QJ \ C

·u 0.25 \ QJ

0.25 ·u !E \ !E QJ \ QJ Vl

' Vl :::, 0 '

:::, C , ... 0

E 0 0 C

:::, .E

_J :::, _J

- 0.25 - 0.25

400 450 500 550 600 650 700

Wavelength (nm)


Circadian Stimulus (CS): is the effectiveness of a light source at suppressing melatonin from

a threshold value of 0.1 (below which there is little suppression) up to a saturation value of 0.7

(beyond which no further suppression is achieved).

Figure 10-3 below shows these researchers’ assessment of the relationship between CLA, CS

and melatonin suppression. This method of quantifying circadian stimulus is useful to include

as it incorporates some clearly defined impacts on the circadian system (i.e. acute melatonin

suppression).

Figure 10-3: Graph of acute melatonin suppression at one-hour exposure according to Circadian Light (CLA) and Circadian Stimulus (CS) measures. Red dashed line represents the threshold below which there is little melatonin suppression [from Figueiro, Gonzales et al. (2016)]. Both measures are applied in practical assessment of circadian light. Figure 10-4 below shows

the relative spectral response curves of both metrics, alongside the SPD for an example tunable

LED white light source at a range of CCT values. This figure highlights potential disagreement

between the metrics in the magnitude of impact that changing CCT might have on circadian

stimulus, dependant on the location of the violet/blue peak in the source SPD. Therefore, both

metrics are reported in the assessment of lighting conditions in this report. Concerns have been

raised regarding the impact of LED lighting on circadian rhythms, as suggested by a recent fact

sheet (USDOE 2017). This publication suggests that chronic disruption of circadian

rhythmicity by LED sources, has the potential to yield serious long term health consequences,

although there are no examples in any literature of such circadian disruption arising from the

installation of LED public lighting.

80% 0.8 indoors indoors outdoors daytime

70% night office 0.7

60% 1000 lux ----- 0.6 vi C

0 1000 lux ~ ·v; 50% 0.5 <I\ V> QJ

Q_ ::)

40% 300 lux ----- 0.4 "S a. E ::) l/) ·..::;

30% 0.3 l/)

C ·c C fl)

0 '6 ..., 20% 0.2 fl) fl)

cij ~ ~

10% 0.1 0

0 --- daylight

0 - incandescent

-10% - 0.1 1 10 100 1000 10,000 100,000

Circadian Light (CL. )


Figure 10-4: Graph showing Melanopic and Circadian stimulus response curves [from Clark and Lesniak (2017)].

Exposure to blue light has been purported to increase alertness, however, a recent Cochrane

review (Pachito, Eckeli et al. 2018) indicated that there was only very low quality evidence to

support this suggestion. One study found improvements in self-reported alertness and mood

states following individually administered blue-enriched light (Bragard and Coucke 2013),

however, the level of evidence was rated as very low, due to potential bias, lack of masking,

and small sample size. In another study that involved evaluation of changes in office lighting

(Viola, James et al. 2008), participants reported higher levels of alertness for the higher CCT

lighting (17,000K vs 4000K), however, a high risk of bias was identified in terms of lack of

allocation concealment, masking, risk of carry-over effects and lack of randomisation. Blue

light has also been used to treat individuals with circadian and sleep dysfunctions (Tosini,

Ferguson et al. 2016). LED studies have shown that blue light in the range 460 – 480nm is

more effective relative to monochromatic light of 555nm in phase shifting the human circadian

clock (Lockley, Brainard et al. 2003).

Given the role of blue light in the timing of the circadian system, the effect of blue-blocking

spectacle lenses on sleep quality has been investigated with conflicting results. While one study

reported a small improvement in sleep quality in individuals with self-reported sleep

difficulties after wearing high blue blocking lenses compared to low blue blocking lenses

(Burkhart and Phelps 2009), another study failed to find any effect on subjective assessment of

sleep quality in normal participants when wearing blue blocking lenses (Leung, Li et al. 2017).

•. ,

Circadian Stimulus Response Curve Rea, MS et al /2005)

Melanopic Lux Response Curve lucaseral(W14)

..,

wave lengtro (nm)

Circadia n Actlve Spectra ---------- Clrcadian Inactive Spectra

--2200K -- 2500K -- 2700K 3050K 3500K 4000K 45001( -- SOOOK --SSOOK -- 6000K

Spectral Distribution Curves for Specific Correlated Color Temperatures of a Tunable White Fixture


10.1 Field-based measurements of LED Streetlights

The circadian stimulating capacity of each exterior street lighting scenario was evaluated based

on irradiance measurements at the eye. These irradiance measurements were weighted using

the photopic response curve, circadian stimulus response curve and mRGC response curve; to

provide the illuminance, circadian light (CLA), circadian stimulus (CS) and equivalent

melanopic lux (EML), respectively. The calculations were made at 1 nm intervals, using the

value-calculators provided by the research groups that developed each of the circadian metrics:

• Melanopic Lux Toolbox (excel-based) from Lucas Group at University of Manchester

http://lucasgroup.lab.manchester.ac.uk/measuringmelanopicilluminance/

• Circadian Stimulus Calculator (web- and excel-based) from Lighting Research Center

at Rensselaer Polytechnic institute https://www.lrc.rpi.edu/programs/lightHealth/

The data presented in Table 10-1 below are based on maximum readings taken in the field,

which were readings of horizontal irradiance taken directly under the luminaire (nadir) at a

height of 1.5 m above the road surface, as if an individual stood directly under the streetlight

and looked up at it. Although this is not a typical position for any road user interacting in the

environment, it is a useful indicator of the ‘worst case’ for irradiance-based measures in this

scenario.

Table 10-1: Maximum irradiance-based measures from street lighting scenarios (directly beneath luminaire) and equivalent circadian measures.

The circadian stimulus (CS) is a useful indicator of the impact of the light at this position: the

CS value represents the percentage of effective melatonin suppression at one-hour exposure,

with none of these lighting scenarios meeting the level that the metric’s authors consider to be

the threshold for melatonin suppression (CS threshold of 0.1) (see Figure 10-3 above). The

values in Table 10-1 are consistent with ‘outdoors night’ (Figueiro, Gonzales et al. 2016) and

Figure 10-3. Similarly, this threshold was not met in any of the assessments of light at

residential boundaries adjacent to the streetlighting installations. The maximum vertical

illuminance readings of spill light in every exterior scenario were measured 4 m behind the

luminaire, towards the nearest residential boundary at a height of 3 m above the ground. These

readings are considered indicative of light travelling towards the windows of residences nearest


Circadian Light (Cla) 8.9 8.6 5.9 8.1 1.6 4.4Circadian Stimulus (CS) 0.0118 0.0114 0.0075 0.0106 0.0018 0.0055

Equivalent Melanopic Lux 4.9 8.9 3.9 4.5 1.8 5.6

http://lucasgroup.lab.manchester.ac.uk/measuringmelanopicilluminance/

https://www.lrc.rpi.edu/programs/lightHealth/


to the streetlights; however it should be noted that illuminance decreases with distance from a

light source of this type according to the inverse square law. This means that illuminance

readings taken at the property boundary due to streetlights will be consistently higher than those

at the residence itself. Table 10-2 summarises these maximum spill (obtrusive) light

illuminances measured at the property boundaries and their equivalent circadian measures. To

provide further context on the impact of these street lighting scenarios on circadian health, the

calculated exterior circadian stimulus values for pedestrian activity were compared with values

for common interior activities.

Table 10-2: Maximum values measured for spill light at the property boundary (vertical illuminance at 3 m above the ground) and equivalent circadian measures.

Illuminance (lux)

Circadian Light (CLA)

Circadian Stimulus

(CS)

Equivalent Melanopic Lux (EML)

Exterior Sources Streetlighting, max reading at residential boundary @ 3 m height

3200K LED 18W 3.0 2.8 0.0034 1.6 4000K LED 20W 3.4 2.2 0.0026 2.3 5900K LED 17W 3.6 4.1 0.0051 2.8 HPS 70W 4.8 1.7 0.0019 0.9 MV 50W 2.6 0.9 0.0009 1.0 FL 32W 1.2 0.6 0.0006 0.7

Interior Sources Room Lighting, Standing Activity

Fluorescent 3500K 36W 78 80.3 0.114 42 LED 4000K 9W 150 87.5 0.123 100 LED 3000K 9W 85 82.5 0.116 45

Technology Devices, Seated Activity

TV @ 3.5m 5 8.4 0.011 5.1 Computer @ 0.7m 49 87.5 0.112 47 Laptop, Max Brightness @ 0.5m 58 61.2 0.088 49 Smartphone, Max Brightness @0.3m 45 87.5 0.112 44

All quantities measured at the eye

Again the CS values do not meet the threshold level of 0.1 (10% suppression upon one-hour

exposure) and are at least 10 times less than this value.

Reviewing this comparison, all of the interior activities recorded have CS values that are

significantly higher than those for light measured at the property boundary under all street


lighting conditions. With respect to interior electric lighting at night time, CS values are just

above the 0.1 threshold (values ranging from 0.114 – 0.123 for LED and fluorescent lighting)

which is consistent with ‘indoors night’ level reported by (Figueiro, Gonzales et al. 2016) and

Figure 10-3 above. For personal light sources and luminous activities these CS values range

from 0.011 (watching TV) to 0.112 (smartphone and computer use). All of these interior

activities have CS values from 2 to more than 150 times higher than the exterior street light

activity.

10.2 Summary: LED Streetlights and circadian rhythm

10.2.1 Summary of Literature

Light exposure, particularly to the blue wavelengths that stimulate the melanopsin retinal

ganglion cells, plays an important role in setting circadian rhythms. However, the existing

literature suggests that the potential for LED streetlight sources to cause disruption of the

circadian rhythms is low, given their low intensity and relatively short exposure when typically

encountered in the normal road environment.

10.2.2 Summary of field-based data

Summarising the key data related to circadian health impact:

• All exterior circadian stimulus (CS) values measured directly under the luminaires for

each technology type were less than 0.012, which was consistent with values reported

in the literature for exterior night circadian light levels and well below the threshold

levels required to suppress melatonin;

• The highest CS values based on vertical irradiance at the property boundaries (3 m

above the ground) was found for the 5900K LED (0.0051), followed by the 3200K

LED (0.0034) and 4000K LED (0.0026), then the High Pressure Sodium, Mercury

Vapour and Fluorescent sources (0.0019, 0.0009 and 0.0006 respectively). These

values are well below the threshold to suppress melatonin;

• The CS values recorded for common interior night time activities were 1 to 2 orders

of magnitude higher than those recorded for road users at night under all street

lighting technologies.


11 Environmental impacts of blue light

The increasing application of LED sources for public lighting has raised concern about the

impact these sources will have on light pollution and specifically on blue light at night. The

International Dark-Sky Association (IDA) issued a white paper in 2010, “Visibility,

Environmental, and Astronomical Issues Associated with Blue-Rich White Outdoor Lighting”

((IDA) 2010) which brought attention to the then-emerging research on the effect of light

source spectrum on a range of topics, including most relevant to this section: sky glow and the

impacts of blue light on wildlife. In this white paper, the authors regarded any white light source

as ‘blue-rich’, including fluorescent, white LED (all CCT) and metal halide; in contrast to

‘yellow-rich’ sources (Sodium lamps). The paper recognises that CCT is a ‘blunt’ metric to

model the range of impacts of short wavelength radiation and calls for a more refined metric

to assess the magnitude of potential detrimental effects. However, in the absence of such a

metric, the IDA has recommended a 3000K upper limit on CCT for LED streetlighting as part

of its Fixture Seal of Approval (FSA) program since 2014 (International Dark-Sky Association

2014).

In 2016, the American Medical Association issued a report titled, “Human and Environmental

Effects of Light Emitting Diode Community Lighting” (CSAPH 2016), which discussed the

impact of 4000K LED (referring to it as ‘the first generation of LED outdoor lighting’) and

encouraged the use of 3000K or lower CCT for roadway LED lighting. The report

unequivocally states that 29% of the spectrum of 4000K LED lighting is emitted as blue light,

while 3000K LEDs only have 21% of emissions in the ‘blue-appearing part of the spectrum’.

Importantly, the source of these statements is not clear and blue light is not defined by

wavelength range or otherwise. The document also states that an LED’s light emissions are

characterised by their CCT (with no mention of SPD), and makes unsubstantiated claims about

3000K LED light being ‘more pleasing to humans’ and ‘much better received by citizens in

general’. Houser (2017) responded directly to this report in an editorial for the IESNA

publication with similar criticisms, acknowledging that while the intentions of the AMA report

are sound – to reduce the detrimental effects of lighting on humans and animals – the

recommendation for 3000K or less is based on ‘misunderstandings and false or unsubstantiated

assertions’.

The editorial by Houser (2017) is one of many formal responses made to the AMA’s report

that provide a more complete picture of how to mitigate the hazards of blue light at night. The


Lighting Research Center’s response by Rea and Figueiro (2016) highlights that CCT is a

simplification of a light source’s SPD, and using it as a metric ignores most of the important

factors associated with light exposure (quantity, exposure duration and timing). It asserts that

CCT should never be used to characterise light as a stimulus for BLH or similar spectrally

selective impacts, describing its use in this context as ‘erroneous’ and ‘misleading’. Bruce

Kinzey, director of the Municipal Solid-State Street Lighting Consortium at Pacific Northwest

National Laboratory issued a response through an MSSLC newsletter that included the

following table of data (Kinzey 2016) (Table 11-1), showing the blue light characteristics of

market available outdoor lighting sources. The data shows the overlap in metrics for quantities

like scotopic and melanopic content of sources with different CCTs, for example, the highest

values for relative scotopic and melanopic content for available 3500K LED light sources (2.73

and 3.57 times HPS values respectively) are greater than for 4000K sources (at 2.65 and 3.40

respectively).

Table 11-1: Blue light measures for a range of light sources for matched lumen output (from (Kinzey 2016)]

Table 6 Selected blue light characteristics of various outdoor lighting sources al equivalent light output.

luminous Relative Scotopic Relative Melanopic

Row light source Flux(lm) CCT(KI % Blue• Content Content••

A PC Wh ite LED 1000 2700 17% - 20% 1.77 - 2.20 1.90 - 2.68

B PC White LED 1000 3000 18% • 25% 1.89 • 2.39 2.10 • 2.99

C PC Wh ite LED 1000 3500 22% - 27% 2.04 - 2.73 2.34 - 3.57

D PC Wh ite LED 1000 4000 27% - 32% 2.10 - 2.65 2.35 -3 .40 PC White LED 1000 4500 31% - 35% 2.35 - 2.85 2.75 - 3.81

F PC Wh ite LED 1000 5000 34% - 39% 2.60 - 2.89 3.18 - 3.74

G PC White LED 1000 5700 39% - 43% 2.77 - 3.31 3.44 - 4.52

H PC Wh ite LED 1000 6500 43% -48% 3.27 - 3.96 4.38 - 5.84

Narrowband Amber LED 1000 1606 0% 0.36 0.12 Low Pressure Sodium 1000 1718 0% 0.34 0.10

K PC Amber LED 1000 1872 1% 0.70 0.42 L High Pressure Sodium 1000 1959 9% 0.89 0.86 M High Pressure Sodium 1000 2041 10% 1.00 1.00

N Mercury Vapor 1000 6924 36% 2.33 2.47 0 Mercury Vapor 1000 4037 35% 2.13 2.51 p Meta l Halide 1000 3145 24% 2.16 2.56

Q Metal Halide 1000 4002 33% 2.53 3.16 R Meta l Halide 1000 4041 35% 2.84 3.75

s Moonlight 1000 4681t 29% 3.33 4.56 T Incandescent 1010 2836 12% 2.23 2.73

u Halogen 1000 2934 13% 2.28 2.81 V F32T8/830 Fluorescent 1000 2940 20% 2.02 2.29 w F32T8/83S Fluorescent 1000 3480 26% 2.37 2.87 X F32T8/841 Fluorescent 1000 3969 30% 2.58 3.18

Percent blue calculated according to LS PDD: Light Spectral Power Distribution Database, http://ga lileo.graphycs.cegepsherbrooke.ge.CA/app/en/home. The speci fic ca lculation, developed for eva luating the potential for affect ing sky glow, divides the radiant power contained in the wavelengths between 405 and 530 nm by the total radiant power contained from 380 to 730 nm, for each l ight source.

•• Mela no pie content calcu lated according to CIE lrradiance Toolbox, http://f iles.cie.co.at/784 TN003 Too lbox.xis, 2015 as derived from Lucas et al., 2014.

Key: PC -- phosphor converted LED - light emitt ing diode

t Moonlight CCT measured and provided by Telelumen, LLC.


As for the consideration of the health impacts of blue light, the environmental impacts must be

considered in the context of the SPD of the light source, along with absolute quantities and

exposure durations.

11.1 Sky Glow

Light that increases the diffuse luminance of the sky at night is commonly referred to as sky

glow. Sky glow can arise from natural sources (e.g. starlight) but most often refers to the effect

of direct scattered and reflected emissions from electric light sources. Through increasing the

luminance of the night sky, sky glow can have negative impacts on wildlife and is detrimental

to visibility for astronomical observation.

Intrinsic lighting factors (related to light sources) that influence sky glow include:

• Spectral Power Distribution (SPD) of the luminaire;

• Total lumen output;

• Spatial distribution of light output, particularly Upwards Light Output Ratio (ULOR).

As noted in a 2017 US Department of Energy report, ’An investigation of LED Street

Lighting’s Impact on Sky Glow’ (USDOE 2017), while LED SPDs may have more violet/blue

wavelength content than other streetlighting sources (particularly sodium lamps), LED

luminaires also have other characteristics that can “reduce or completely offset” potential

negative effects. The report notes that with careful attention to the total lumen output and

distribution of light, the increase in blue-rich light content in LED SPDs can be mitigated to

offset the sky glow impacts from blue light.

The sky glow arising from exterior lighting is not only dependant on intrinsic lighting factors,

but also depends upon environmental factors. Some of the key external factors that influence

the observed sky glow are:

• Atmospheric conditions;

• Observer’s location to the light source (e.g. distance from city centre);

• Size of the source (e.g. city size and light density);

• Pole heights;

• Surrounding geographical features, including the reflectance of the ground surface.

Consequently, models used to predict sky glow must account for atmospheric conditions and

observer distance in their calculations. The impact that blue light has on increasing sky glow


effects, and the mechanisms for this increase are well documented ((IDA) 2010, Bierman 2012,

Luginbuhl, Boley et al. 2014). The scattering processes that light undergoes in the atmosphere

are wavelength dependant (particularly Rayleigh scattering from smaller molecules), with

shorter-wavelength light being more strongly scattered. The enhanced scattering of shorter

wavelengths increases sky glow for observers near the source, but also increases the extinction

(fall off) of glow at longer observer distances. However, this effect is relatively small compared

to the effect that an observer’s enhanced blue-light visual sensitivity at low light levels has on

increasing the perceived sky glow (Bierman 2012).

Several researchers have sought to evaluate the impact of increased blue light (often in the

context of increasing use of LED in public lighting) on sky glow. Luginbuhl, Boley et al. (2014)

suggested that blue-rich light sources would produce 15 – 20% more radiant (broad spectrum)

sky glow than yellow-rich light sources, namely sodium lamps. This effect is magnified when

the authors consider the visual observation of sky glow at low-light, scotopic levels, since

scotopic sensitivity is weighted towards blue light. The authors model sky glow for six lamp

types matched for photopic lumen output: low pressure sodium (LPS), high pressure sodium

(HPS), 2400K LED, 5100K LED, metal halide (MH), and a filtered (500nm) 3000K FLED.

The 5100K LED and MH lamps produced 6-8 times the scotopic luminance of the LPS, and

three times that from HPS close to the sky glow source (shown in Figure 11-1). This factor

decreases with observation distance from the source.

Figure 11-1: Scotopic sky luminance ratio variation with observer distance, ratio compared to LPS scotopic luminance [from Luginbuhl, Boley et al. (2014)].

9.00 ~

8.00 ...J

0 ... 7.00 .2 ....

I'll 6 .. 00 a::

8 5.00 C

\

' \. I'll C 4.00 E .3 3 .. 00

-~ .

u 2.00 ·-D.

....... - -0

8 LOO --U')

0.00 I

0 50 I

100 I

150

Distan,ce (km),

I

200

-

I

250

- LED510OK MIH4100K

LED2400KFLED HPS LPS

300


Comparing these results with a photopic consideration of sky glow gives a differently scaled

graph of lamp luminance ratios, as shown in Figure 11-2 below.

Figure 11-2: Same data with a different metric, photopic sky luminance ratio variation with observer distance, ratio again compared to LPS photopic luminance [from Luginbuhl, Boley et al. (2014)].

Bierman (2012) assessed the increase in atmospheric scatter arising from a 6500K LED source

compared with a HPS source, accounting for aerosol content derived from aerosol optical depth

measurements (which reduces the wavelength dependent effect of Rayleigh scatter). The

findings suggest that the 6500K LED source contributed 10 – 20% more scattered light to sky

glow. The author points out that if the light sources are photopically matched for output, but

evaluated for scotopic measures of sky glow, then blue-rich sources will perform worse.

Conversely, they found that matching sources for equal scotopic light output eliminates much

of the difference.

Rea and Bierman (2015) further acknowledge that sky glow magnitude depends on the

operating characteristics of the detector, and highlight that while scotopic luminance is relevant

for night sky observation, photopic vision is used for foveal viewing even in the dark.

In the US DOE report (USDOE 2017), the sky glow arising from seven different LED products

was modelled (using actual available product SPDs with CCT ranging from 1872K – 6101K).

For each luminaire, four levels of ULOR were compared (0%, 2%, 5% and 10%) and two levels

of lumen output (100% and 50%). The results presented in that report are unique and valuable

V, Q. ...I 0 ... o 1 ·-'la a:: ~ ; 0.9 C ·-E ~ u 0.8 ·a 0 0 .c Q.

0 50 100 150

Distance {km}

200 250

LIPS HPS

LIED240OK FLED

MH41001K

LIEDSlOOK

300


to this report in their assessment of realistic scenarios based on actual instances for street

lighting conversions in the U.S.

The US DOE report (USDOE 2017) showed comparisons of LED streetlighting source

contribution to sky glow compared to High Pressure Sodium (HPS) as a baseline condition. In

particular, it highlights what is deemed a typical result of conversions from High Pressure

Sodium (HPS) to LED in the U.S: an incumbent HPS cobra head product with 2% ULOR being

replaced by an LED product at 50% lumen output and 0% ULOR. The key findings are shown

in Figure 11-3 and Figure 11-4 for near observers, and observers 40 km from the city

(respectively). The dashed red-orange line shows the baseline sky glow impact arising from

the incumbent HPS scenario. For a range of seven LED SPDs shown, three conversion

scenarios are modelled: 0% ULOR and lumen output 100% for both HPS and LED (top); 0%

ULOR for both, LED lumen output reduced to 50% (middle); and typical conversion 2%

ULOR for HPS, 0% ULOR and lumen output reduced to 50% for LED (bottom). Spectrally

unweighted sky glow impact is a useful measure for the impact on astronomical observation,

while scotopically weighted sky glow is a better indication of visible sky glow (for human

visual observation at very low light levels). ATM 1-4 are clear skies conditions with increasing

turbidity, and ATM5 represents complete cloud covered skies (which lead to increased light

scatter, particularly of blue light). As a reference, SPD 6 and 7 have CCT ~3000K, SPD 8 and

9 have CCT ~4000K.


Lighting

Figure 11-3: Near source observer results for m

odelled sky glow ratios (relative to H

PS) for unw

eighted (left) and scotopically weighted (right) m

easures [from (U

SDO

E 2017)].

C:

~ "' ~ :i: ;;t 0

~

~ .,, ~ r ~ .. ::E "' ~ ~ "' 0

Impact of SPD, Lumen Output, and Uplight LE D: 0% up, SO~ out vs HPS: 2% up, 100% out

Sky Glow mpact :Rola1!11• to lnrumbontSe•nario 0 0 Q ~ -" b .._ OQI i-.- o,

I SfDS

SP06

SPD7

s,poa I SP09 I

SPDlO ■ SPDn ■

SPDS I S P□

SP07

SPD8 I SPD9 1

~ ... ~ a ::,,. cc "' "" ~ C"I

,. ~ .... l,,

~

== "'

I sP01o l ! 1~ 5PD11 j : ~

l ~ I I SPDS 1 ~ 1~ SPD6 i 3: ;;;

I ~ :,0

SPO7 : .I,

SPD8 .

>POS

s,Po&

SP07

! ■ SP09

! ■ SPDl0

i - SPDll I I

I SPDS

I SP09

I SPDlO

I SPDll

~ "'

C: z :E m ili ::c ~ 0

.., ,., !a 0 .,. ~ ~ :::E .., iii :c _, ,., 0

lmp,act of SPD and Lumen Output LED : 0% up, 50% out vs HPS: 0% up, 100% out

Sky Glow Impact Relative to lncumbentScenario 0 0 Cl - - ~ ~ ~ "' "' b -... .. ;., .,., b .. " ;, ..

sros I ,. SPD6 ~ .... SPD7 l,,

SPOO I SP□9 I

SPDlO I SPOll I

SPDS ~

SPD6 == '" SPD7

SPDS:

SP09

SP010 I z "' )>

SPDll I X,

0

"' I sros "' )> "' .., X,

SP06 == ~ HD7

'.i: X,

1 sros I SPD9

I SPDlO

I SPDll

SPDS ,. _, SPD6 :i:

"' SP07

I SPOO I I SPD9 I

: ' I SPDlD I I

' SPDll '

~2J??.l ~~ ??

Impact of SPD Only LED: 0% up, 100% out vs HPS: 0% up, 100% out

Sky Glow Impact R•la!lwuo lncumbontSUnarlo 0 Q 0 - ~ ... ~ .. "' 0 -... 00 ;.; .,., b -... " " ""

SPDS I l> _, SP06 :i: SPD7 i:

I S'P08

I spog

c:: ■ SPDlO z ::E • SPDll ,., c\ ::c: SPDS ~ _, "' SPD6 :i: 0 .,,

SP07

$PD8

SP09

I SPOlO z li I SPD11 "' 0 "'

sP□s I "' I> m

SPD6 i ~ SP07 ! "'

"' I SPOS

8 I SPD9

cl SPD10 I .., n SPOU ■ )>

~ :E SPDS l>

"' _,

c\ SPD6 :i: :c "' _,

SPD7 m 0 I SPDS

I 5P09

I SPDlO

I SPDll


Lighting

Figure 11-4: Distant observer (40km

from source) results for m

odelled sky glow ratios (relative

to HPS) for unw

eighted (left) and scotopically weighted (right) m

easures [from (U

SDO

E 2017)].

Impact of SPD, Lumen Output, and UpHght Impact of SPD and Lumen Outpu~ Impact of SPD Only LED: 0% up, 50% out vs HPS: 2% up, 100% out LED: 0% up, 50% out vs HPS: 0% up, 100% out LED: 0% up, 100% out vs HPS: 0% ,up, 100% out

Sky GI ow Imp ad Re l"ll voto lnrumb""t s ... narl o Sky Glow Impact Rel11tiveto lr,a,mbenlS<:enario Sky Glow I mp a Gt Relati~e to In <u mben tSc,enarl o 0 0 Q ..., .., ..,

"' "' 0 Q 0 u .. .. .., w .... 0 0 0 ~ ~ " ., .. "' "' a -.. "' ;,,, "' 0 -.. "' " "' a -.. .. .... "' a -.. .. .... "' :, ... .. ~ "' :, -.. .. .... ;,,,

SPDS l> .... , SP06 ~ .. SPD7 1,.

SPD S I :: SPD6 ~ .. SPD7 1,.

• SPOS . l> I .... i SP06 ~ ... ' S~07 I,. I I

SP08 SPDB I !I SPD8

SPD9 SPOO I ! ■ sPD9 C SPOIO z ~ SPO U

iii ::c: SP05 l> -I ... .... , 0 SPD6 ~

"''

C: SPOIO I I z SPDll ■ :E

m i:i :c SPDS l> ;;l .... 0 SPD6 ~

"'

C: • SPOlO z :E 1 • SPD1 1 .., ci I

I

~ 1 SPDS l> ....

0 SPD6 ~ "' SP07 SP07 SP07

SPD8 SPOB SP08

SPD9 0 SPD9 0 SP09 0

SPD10 .;:; -I l>

SPD11 z -I

SPOlO ~ SPDU ~

SPOlO ~ I SPDU

l> z -I

0 0 0 SPDS l> e: .... ... SPD6 ~ "' .... <: SPD7 1,. ;:;

I SPOS "' l> "' .... "' SP0 6 ~ ~ ... .., SP07 I,. "'

sros I ... l> .,, .... "'

SP0 6 ~ "' ... <: SP07

1,. g;

.,. I SP0 8 n I SP09 s 0 I SPD10 ~

E I SPOll

~ ~ SPOS l> ... .... , i:i SP06 ~ :!i

.,,, ... SP0 7 0

SP08

.,, SP08 n I SPD9 0 ....

■ SPOlO 0 -a n ■ SPDll l>

~ :E SPD5 :: m i;i SPD6 31: :c "' .... SP07 m Q

SP08

~ I SPD8

~ - SP09

0 - SP010 .., £ SPOll -

'< SPO S I :E l> .., ....

ci SPD6 ~ :c "' .... SP07 ,., C I SPD8

SPD9 SP09 SPD9

SPDlO I SPD lO SPDlO

SP011 I SPDll SfDll I

?J?:, , 1~~ p~


The key findings of the US DOE report for typical LED street lighting conversions (a 2%

ULOR HPS installation converted to a 0% ULOR LED at 50% original lumen output) are as

follows:

• For residents near a city, on an unweighted basis, the contribution to sky glow from a

typical conversion across all CCTs (SPDs) modelled should be reduced relative to the

baseline HPS condition. Even when results were scotopically weighted, there were

still several LED products (ranging from Amber LED at 1874K to white LED at

3940K in this product group) that achieved a modelled reduction in sky glow;

• For residents further from the city (40 km away), the sky glow from typical

conversions was calculated to be negligible for all SPDs modelled, for both human

(scotopically weighted) and astronomical (unweighted) observations

It should be noted that where SPD is the only variable considered (i.e. no change to ULOR and

no reduction in lumen output), then all the modelled HPS to white LED conversions resulted

in an increase in sky glow.

The study looked at the contributions of the intrinsic lighting and external factors in sky glow.

The key observations relevant to this report were:

• SPD is the factor with greatest potential to increase sky glow across the entire range

of LED and High Intensity Discharge (HID) lighting products evaluated. However,

when a comparison is made between two products relatively close in characteristics

(the example given in the report is between two products with nominal CCTs of

3000K vs 4000K) the range of impacts is much narrower, regardless of whether

unweighted or scotopically weighted consideration is used;

• ULOR is the factor with the greatest potential to decrease sky glow for the distant

observer (for all SPDs and lumen output levels);

• Sky glow directly scales with reduction in lumen output for all observer positions,

which suggests that scheduled dimming where appropriate would decrease sky glow

contributions effectively.

A recent case study (Barentine, Walker et al. 2018) offers data on sky glow changes from

before and after streetlighting conversion from 445 million fully shielded HPS/LPS to 142

million fully shielded 3000K white LED luminaires. The study took place in Tucsson, Arizona

near astronomical observatories, so minimizing sky glow by design was a priority. Simulated

sky glow models predicted a reduction of 10-20% in sky glow. The city’s upwards-directed


radiance detected from Earth’s orbit decreased by approximately 7% after the transition.

11.2 Impact of blue light on wildlife

Literature on the impact of artificial light at night on fauna shows it can affect behaviour,

foraging, communication, breeding cycles and the habitat of species, including insects, bats,

birds, turtles and other animals (Rich and Longcore 2006, Longcore, Rodríguez et al. 2018).

There is also evidence to indicate that light at night affects plant processes and healthy growth

(Rich and Longcore 2006, Bennie, Davies et al. 2016, Somers-Yeates, Bennie et al. 2016).

In a recent publication on this topic, Longcore, Rodríguez et al. (2018) recognises the challenge

for conservationists in assessing the impact of an increasing number of lamp types developed

with different SPDs on wildlife in experimental or field situations. The authors were driven by

this challenge to propose a method for assessing the suitability of lamps to different light-

sensitive environments based on calculation alone. They note that switching to LEDs can be

negative or positive for different species depending on the spectrum, duration, intensity and

directionality of the new source. The impact of blue light specifically on wildlife depends on

the action spectrum of the animal process in question, so the first step in this method is to

establish the spectral response. To demonstrate the calculation process, the spectra shown in

Figure 11-5 for different animal processes (some behavioural, some visual responses) were

used. Figure 11-6 shows the SPDs for light sources used, while Figure 11-7 ranks the light

sources by average impact on different wildlife types.


Figure 11-5: Spectral response for a range of animal behavioural processes and visual responses, as well as photosynthesis (plants) and sky glow vs wavelength (nm) [from (Longcore, Rodríguez et al. 2018)].

(I) VJ C 0 0. VJ (I) [[

~ ti (I) 0.

en

1.00

0.80

0.60 0.40

0.20

0.00

1.00

0.80

0.60 0.40

0.20

0.00

1.00

0.80

0.60 0.40

0.20

0.00

1.00

0.80

0.60 0.40

0.20

Bee Attraction Green Turtle Adult Green Turtle Hatchling Human Melanopic

Human Photopic Human Scotopic Insect Attraction Juvenile Salmon

Logg:erhead Hatchling Moth Attraction !Newell's Shearwater Photosynthesis

Skyglow Mie Star Light Index

0.00 __.,.,~~~~ ....... ,..,.,...,...~-,-~~m---r--.',.,.,..T"T"',...,..,,...._-1"""1""~m~~ .,..,...,..,.....,...~~~~m~~~m~ 350 450 550 650 750 350 450 550 650 750 350 450 550 650 750 350 450 550 650 750

Wavelengih


Figure 11-6: Spectral Power Distributions for test light sources to evaluate wildlife impact, vs wavelength (nm) [from (Longcore, Rodríguez et al. 2018)].

Kerosene Oil

Full Moon

Sil0380 ◄.20itft0500560560ft2G&e0700

AEL 75W

340 380 420 cc soo 540 580 e20 eeo roo no 11110

SORAAVivid

CFL Greenlite 13W

Iwasaki 60W

150W HPS

: _1_8_W_L_P_S __ ~j~----34o 380 ◄20 460 S1X1 5,40 580 GO eeo 100 140 no


Figure 11-7: Calculated impact of individual light sources on wildlife, including overall weighted ‘Wildlife Index’ based on the selected spectra. Also highlighted is source CCT by colour [from (Longcore, Rodríguez et al. 2018)].

This work highlights that while CCT is correlated positively to the impact on wildlife

(increasing CCT generally increases source impact on wildlife relative to Daylight), there are

still examples where CCT does not consistently predict impact relative to other light sources.

Furthermore, it is evident that some sources that have low impact for some wildlife may have

relatively high impact on others, for example, the lower CCT lamps have minimal average

impact on insects, but large impact relative to other sources with higher CCT on the Newell

shearwater (a seabird).

One publication explored the impact of spectrum and CCT in attracting insects by using traps

to count a range of insect species drawn to six LED luminaires with CCTs of 2700K, 3000K,

3500K, 4000K, 5000K and 6500K compared with a baseline of HPS (Pawson and Bader 2014).

Insect Average Sea Turtle Average

Newell 's Shearwater

Juvenile Salmon

Wildlife Index CCT

6500

5583

4667

3750

2833

00000~00000~00000~00000~00000~ oN~mooooN~mooooN~mooooN~mooooN~mooo I I I I I I I I I I I I I I I I I I

LPS 18 W

74WWCW10

74WWCW7

1917 150WHPS

1000 AEL 75W

PC Amber Cree

Anna 's Light

LLT Telescope Light

LSG Good Night 2016

Philips AmbientLED

CFL Greenlite 13 W

Cosmopolis 60W

3000K LED

Ceramic Metal Halide 70 W

Los Angeles LED

OCTRON32W

LE□way Streetl ight CW 54W

A

Yard Blaster

TL950

Full moon

Iwasaki 60W

Kerosene Oil

SORAA Vivid

065

-E -■ ■

= •

-


The LED luminaires attracted 48% more insects than the HPS. Interestingly, for most of the

insect taxa there was no significant difference between the count for different CCTs, which

may suggest the presence of a threshold effect (where some minimum quantity of blue light

causes the increased attraction, but increasing the quantity does not increase the effect). The

authors call for further work into LED technology developments to minimise potential

ecological effects.

The selection of an appropriate lamp for environmentally sensitive areas is clearly influenced

by the relative match between the organism’s spectral response and the source SPD. Intensity

of the light source is also recognised as important, although threshold levels for impacts are

not always clear. Some species may respond to spectra differently depending on the intensity

of the light (Stone, Harris et al. 2015). The precautionary principle suggests that even where

spectral tuning is used to reduce the impact of exterior lighting, intensity levels should be kept

to a minimum.

11.3 Summary: Environmental impacts and Blue Light

11.3.1 Summary of literature

Light at night has an undeniable negative impact on sky glow, and disrupts wildlife and plant

function. The impacts of blue light on sky glow are:

• Enhanced light scatter in the blue wavelength region (up to 25%) dependant on

atmospheric conditions – this effect decreases considerably with distance from the

source;

• Scotopic vision is more sensitive in the blue wavelength range, which serves to

increase perceived sky glow at night.

Practice-based research and case studies have shown that design factors for LED luminaires

can offset the increase in sky glow expected due to increased blue light content.

All light at night has potential to disrupt wildlife and plants. In order to assess the severity of

the impact, an action spectrum for the biological mechanism should be identified so that light

sources with appropriate spectral content can be selected.


12 LED Streetlighting and driving

It is important to also consider the impact of LED streetlighting on driving performance and

safety. The positive benefits of streetlighting on night-time crash risk is well established, as

evidenced in a review by Beyer and Ker (2009), potentially through improvement of drivers’

visual capabilities. However, the majority of studies included in this review were published

prior to 1990, were rated as prone to bias and confounding factors, and did not consider the

safety benefits of more recently introduced streetlighting, such as LEDs. Two recent meta-

analyses of crash data and streetlighting levels in the US (Gibbons, Guo et al. 2014) and UK

(Steinbach, Perkins et al. 2015), confirmed the benefits of traditional streetlighting for reducing

night-time crash risk, and also demonstrated that reducing traditional streetlighting levels may

not adversely reduce safety. To date, there has been no research undertaken into differences in

crash risk under LED streetlighting compared to conventional sources, although one

government report showed that recognition distances for pedestrians wearing non-reflective

clothing were not significantly different under normal levels of LED (6000K) and HPS

(2100K) streetlighting, although the effects of streetlight dimming on pedestrian distances was

more marked for the LED compared to HPS streetlight dimming (Gibbons, Meyer et al. 2015).

Interestingly, this report also presented evidence to suggest that some aspects of performance

were improved under LED streetlighting, including off-axis pedestrian distances and target

colour recognition (red and green).

The AMA report (CSAPH 2016) suggests that LED lighting can lead to increased glare

compared to conventional lighting, particularly for older drivers. However, the effects of glare

on road safety overall, and specifically from streetlight or headlight sources on road safety are

poorly understood (Owsley and McGwin 2010). It is also important to differentiate between

the effects of Disability Glare and Discomfort Glare, as they are different metrics of glare.

Disability glare is an objectively measured degradation in visual performance resulting from

scattered light from a glare source and is commonly measured by assessing the reduction in

visual function, such as contrast sensitivity, in the presence and absence of a glare source

(Abrahamsson and Sjostrand 1986). While drivers often complain about the ‘blueness’ of

oncoming HID headlights, particularly older drivers (Mainster and Timberlake 2003),

disability glare is not significantly affected by the wavelength of the light source (Whitaker,

Steen et al. 1993). Indeed, it has been shown that the problems resulting from oncoming HID

car headlights are related to their increased brightness rather than the relative amount of blue


light (Mainster and Timberlake 2003). To date there is no research evidence to suggest that

LED streetlights cause any additional disability glare compared to conventional sources.

Discomfort glare on the other hand, does not impair visual function, but it can be startling and

distracting to a driver and is wavelength dependent, being greater at shorter wavelengths

(Bullough 2009). Importantly, the physiological and psychological origins of discomfort glare

are not well known (Howarth, Heron et al. 1993). One small scale study examined the

discomfort glare for a range of LED streetlights with different CCT, using the de Boer scale

(Lin, Liu et al. 2014). The authors reported that a higher CCT resulted in more discomfort

glare, however, their results are confounded by differences in radiant output, contrast and size

of the different LED streetlight sources. Further work is needed to establish whether there is

any relationship between the CCT of an LED streetlight, when matched for radiant output and

other characteristics, and its potential to result in discomfort glare.

In summary, the effects of LED streetlighting on driving performance, glare and safety are not

well understood. For example, while it is well known that increasing the radiant light output of

a source will increase both discomfort and disability glare, the specific effect of a light source’s

spectral distribution on these measures of glare remain unknown and requires further controlled

experimentation. The concerns raised by the AMA report are therefore not supported by

evidence. As stated by Bullough and Radetsky (2013) there is still a critical lack of evidence

regarding the impact of LED streetlight configurations, dimming and adaptive capacity on both

visual performance and road safety.


13 Appendix: Field-based data

13.1 Raw field-based data

Table A 1: Activity-based measures of blue light hazard weighted irradiance, driving, cycling and walking activities


0 0.00000 0.0001 0.00004 0.00003 0.00001 0.0000110 0.00058 0.0017 0.00073 0.00020 0.00028 0.0009820 0.00019 0.0007 0.00077 0.00010 0.00018 0.0003830 0.00007 0.0002 0.00027 0.00002 0.00009 0.0001040 0.00001 0.0001 0.00013 0.00001 0.00004 0.0000550 0.00000 0.00008 0.00000 0.000030 0.03 0.16 0.09 0.50 0.03 0.0510 1.97 3.82 0.98 2.12 0.57 2.0520 0.81 1.69 1.13 1.26 0.38 0.8030 0.37 0.53 0.44 0.39 0.18 0.2440 0.13 0.32 0.21 0.21 0.10 0.1250 0.01 0.12 0.15 0.060 0.02 0.25 0.13 0.23 0.03 0.0710 2.64 6.05 1.74 1.27 0.63 3.2420 1.02 2.66 1.93 0.70 0.41 1.2730 0.43 0.86 0.73 0.19 0.19 0.3740 0.13 0.50 0.35 0.10 0.10 0.1950 0.01 0.21 0.07 0.07

E(s) - Scotopic Illuminance, i.e. Scotopic weighted

irradiance (scotopic lux)


Driving - values measured at the driver's eye


(W/m-2)


0 0.00009 0.0002 0.00011 0.00009 0.00006 0.0003310 0.00086 0.0025 0.00192 0.00108 0.00064 0.0015820 0.00024 0.0010 0.00088 0.00024 0.00036 0.0005730 0.00010 0.0004 0.00043 0.00007 0.00018 0.0002140 0.00012 0.0002 0.00025 0.00003 0.00012 0.0001250 0.00003 0.00018 0.00003 0.000120 0.33 0.39 0.22 1.04 0.16 0.7110 2.88 5.59 2.41 9.64 1.18 3.1720 0.92 2.35 1.22 2.13 0.67 1.1230 0.39 0.91 0.69 0.66 0.35 0.4140 0.30 0.64 0.45 0.36 0.25 0.2650 0.10 0.31 0.36 0.280 0.42 0.56 0.33 0.55 0.17 1.08

10 3.73 8.69 4.26 5.84 1.35 4.9020 1.15 3.59 2.08 1.27 0.76 1.7430 0.49 1.41 1.13 0.39 0.41 0.6540 0.45 0.96 0.70 0.19 0.29 0.3950 0.12 0.48 0.21 0.31




Cycling - values measured at the cyclist's eye


(W/m-2)


Table A 2: Maximum irradiance-based values measured for each source (horizontal illuminance directly under luminaire, nadir, at 1.5 m above the road surface)

Table A 3: Maximum Radiance-based Values measured for each source in spill light assessment.


0 0.00001 0.0001 0.00009 0.00015 0.00005 0.0000710 0.00011 0.0012 0.00221 0.00073 0.00063 0.0008020 0.00023 0.0007 0.00100 0.00019 0.00026 0.0003430 0.00011 0.0003 0.00043 0.00005 0.00013 0.0001540 0.00009 0.0002 0.00023 0.00002 0.00009 0.0000950 0.00004 0.00015 0.00002 0.000130 0.03 0.28 0.20 1.59 0.17 0.18

10 0.39 3.34 3.27 6.81 1.14 1.5720 0.90 1.89 1.46 1.79 0.51 0.6730 0.48 0.81 0.68 0.53 0.28 0.3240 0.33 0.53 0.42 0.23 0.24 0.2050 0.18 0.31 0.28 0.350 0.03297158 0.36 0.30 0.89 0.20 0.26

10 0.48943689 4.93 5.44 4.06 1.32 2.4220 1.11429536 2.77 2.45 1.04 0.59 1.0430 0.5830211 1.20 1.11 0.28 0.33 0.5040 0.42831236 0.75 0.66 0.11 0.26 0.3050 0.2152063 0.47 0.15 0.38




Walking - values measured at the pedestrian's eye


(W/m-2)


Blue Light weighted irradiance E B (W/m-2)

0.0032 0.0072 0.0045 0.0026 0.0026 0.0048

K B,v 0.00034 0.00054 0.00088 0.00011 0.00056 0.00051Circadian Light (Cla) 8.9 8.6 5.9 8.1 1.6 4.4

Circadian Stimulus (CS) 0.0118 0.0114 0.0075 0.0106 0.0018 0.0055Equivalent Melanopic Lux 4.9 8.9 3.9 4.5 1.8 5.6

3200K LED 4000K LED 5900K LED HPS MV FLMax Luminance (cd.m-2) 25000 41000 24000 56400 7500 17000

Max L B (W.sr-1.m-2) 8.4 22.2 21.2 6.3 4.2 8.6

r


Table A 4: Irradiance based measures of BLH and circadian stimulus for common interior activities, compared to walking activity.

3200K LED

4000K LED

5900K LED HPS MV FL

Fluore-scent

3500K

LED 4000K

LED 3000K

TV @ 3.5m

Computer @ 0.7m

Laptop Max

Brightness @ 0.5m

Smartphone Max

Brightness @ 0.3m

Illuminance (lux) 0.9 3.3 3.3 6.8 1.1 1.6 78 150 85 5 49 58 45Blue Light weighted

irradiance E(B) (W/m-2)0.0002 0.0012 0.0022 0.0007 0.0006 0.0008 0.030 0.070 0.025 0.005 0.041 0.043 0.047

Scotopic illuminance 1.1 4.9 5.4 4.1 1.3 2.4 109.2 250.7 119 11.5 112.7 116 103.5Circadian Light (Cla) 0.9 2.2 3.8 2.4 0.4 0.7 80.3 87.5 82.5 8.4 87.5 61.2 87.5

Circadian Stimulus (CS) 0.001 0.003 0.005 0.003 0.000 0.001 0.114 0.123 0.116 0.011 0.112 0.088 0.112Equivalent Melanopic

Lux 0.47 2.25 2.50 1.34 0.45 0.92 42 100 45 5.1 47 49 44

All quantities measured at the eye

Walking Activity, Streetlighting Exterior Sources Walking Activity, Room Seated Activity, Technology Sources


Figure A 1: Spectral Power Distributions of streetlights measured in the field

1.000E-03

9.000E-04

8.000E-04

7.000E-04

~ < E C, 6.000E-04 N

--3200KLED

< E

i --4000KLED

a, 5.000E-04 u C --5600KLED

.!!! "O

~ --HPS ~

4.000E-04 ... u a, a. --MV Vl

3.000E-04 --FL

2.000E-04

1.000E-04

0.0OOE+00

350 400 450 500 550 600 650 700 750

Wavelength (nm)


Figure A 2: Close-up view of Spectral Power Distributions of streetlights measured in the field, highlighting BLH region and including BLH spectral curve.

5.000E-04 "" - - ...... ' ,, ' ' ,,

' ,, .... ,I ' ,I \

I \ --3200KLED

' I ' 4.000E-04 I \ --4000KLED I \

\

-.=.-- \

< E

\ --5600KLED \

C \

C;' < E 3.000E-04 i

\ \ --HPS

' \ "' \ u C \ --MV .!!! \ "O

~

] u 2.000E-04 "' C.

\

I \

I ' --FL

' \ \

V, \ - - - BLH curve \ (dimensionless) \ \

I I

1.000E-04

-:~ - . 0.000E+00 ---

380 400 420 440 460 480 500 520

Wavelength (nm)

L


Table A 5: Spill Light illuminance and luminance, along roadway at 1.5m above road surface

0 3 6 9 12 15 18 21 24 27 30 33Illuminance (lux) 1.6 4.1 4.2 3.5 2.8 1.8 1.5 1.1 1.1 0.6 0.5 0.3Luminance (kcd.m -2 ) 21.6 28.8 18.2 14.8 12.3 8 5.1 3.6 2.8 2.2 1.8 1.4Illuminance (lux) 0.4 6.3 6.3 5.3 5 4.1 2.9 2 1.6 1.2 0.9 0.7Luminance (kcd.m -2 ) 33.8 41 30.1 23 20.8 16.6 11.7 7.6 5.8 4.6 3.4 2.2Illuminance (lux) 0.2 2.2 2.5 3.1 2.8 2.1 1.6 1.2 0.9 0.7 0.7 0.6Luminance (kcd.m -2 ) 7.2 8.2 12.1 14.2 13.6 9.5 7 4.9 3.5 2.7 2.3 1.9Illuminance (lux) 1 9.6 13 10.1 6.9 4.8 3 1.8 1.1 0.7 0.6 0.5Luminance (kcd.m -2 ) 30.5 38.3 56.4 46.2 29.7 19.8 11.6 6.3 3.68 2.72 1.95 1.48Illuminance (lux) 0.2 1.9 1.8 1.7 1.3 1.1 0.8 0.6 0.5 0.4 0.3 0.3Luminance (kcd.m -2 ) 5 5.4 6.1 6.2 5.2 4 2.8 2 1.5 1.1 0.84 0.66Illuminance (lux) 0.3 4 4.7 3.4 2.5 0.9 1.4 1 0.7 0.5 0.4 0.3Luminance (kcd.m -2 ) 6.6 9.4 17 13.1 10.2 7.7 5.5 3.86 2.6 1.82 1.34 0.99

36 39 42 45 48 51 54 57 60 63 66Illuminance (lux) 0.3 0.2 0.2 0.1 0.1 0.1 0.1 0 0.1 0.2 0.2Luminance (kcd.m -2 ) 1.1 0.83 0.63 0.5 0.4 0.3 0.25 0.18 0.15 0.13 0.11Illuminance (lux) 0.8 0.8 0.7 0.9 3.5 (46.5m)Luminance (kcd.m -2 ) 2.1 1.7 1.45 1.25 1Illuminance (lux) 0.5 0.5 0.4 0.4 0.3 0.3 0.3 0.3 0.3 (58.4m)Luminance (kcd.m -2 ) 1.6 1.3 1.1 0.8 0.7 0.6 0.5 0.4 0.4Illuminance (lux) 0.4 0.4 0.3 0.3 0.3 0.4 1Luminance (kcd.m -2 ) 1.18 0.93 0.78 0.65 0.54 0.46 0.4Illuminance (lux) 0.3 0.3 0.3 0.3 0.4 0.4Luminance (kcd.m -2 ) 0.54 0.43 0.37 0.28 0.26 0.21Illuminance (lux) 0.3 0.3 0.4 (40.5m)Luminance (kcd.m -2 ) 0.75 0.58 0.53

M50

FL32

Distance from below pole 1 LED

3200KLED

4000KLED

5900K

S70

M50

FL32

Distance from below pole 1 LED

3200KLED

4000KLED

5900K

S70


Table A 6: Spill Light blue weighted irradiance, EB, and radiance, LB, along roadway at 1.5 m above road surface.

0 3 6 9 12 15 18 21 24 27 30 33E B (W.m-2) 0.0005 0.0014 0.0014 0.0012 0.0009 0.0006 0.0005 0.0004 0.0004 0.0002 0.0002 0.0001LB (W.sr -1 .m -2 ) 7.3 9.7 6.1 5.0 4.1 2.7 1.7 1.2 0.9 0.7 0.6 0.5

E B (W.m-2) 0.0002 0.0034 0.0034 0.0029 0.0027 0.0022 0.0016 0.0011 0.0009 0.0007 0.0005 0.0004LB (W.sr -1 .m -2 ) 18.3 22.2 16.3 12.5 11.3 9.0 6.3 4.1 3.1 2.5 1.8 1.2

E B (W.m-2) 0.0002 0.0019 0.0022 0.0027 0.0025 0.0019 0.0014 0.0011 0.0008 0.0006 0.0006 0.0005LB (W.sr -1 .m -2 ) 6.4 7.2 10.7 12.5 12.0 8.4 6.2 4.3 3.1 2.4 2.0 1.7

E B (W.m-2) 0.0001 0.0011 0.0015 0.0011 0.0008 0.0005 0.0003 0.0002 0.0001 0.0001 0.0001 0.0001LB (W.sr -1 .m -2 ) 3.4 4.3 6.3 5.2 3.3 2.2 1.3 0.7 0.4 0.3 0.2 0.2

E B (W.m-2) 0.0001 0.0011 0.0010 0.0010 0.0007 0.0006 0.0004 0.0003 0.0003 0.0002 0.0002 0.0002LB (W.sr -1 .m -2 ) 2.8 3.0 3.4 3.5 2.9 2.2 1.6 1.1 0.8 0.6 0.5 0.4

E B (W.m-2) 0.0002 0.0020 0.0024 0.0017 0.0013 0.0005 0.0007 0.0005 0.0004 0.0003 0.0002 0.0002LB (W.sr -1 .m -2 ) 3.4 4.8 8.6 6.7 5.2 3.9 2.8 2.0 1.3 0.9 0.7 0.5

36 39 42 45 48 51 54 57 60 63 66E B (W.m-2) 0.0001 0.0001 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0001 0.0001LB (W.sr -1 .m -2 ) 0.4 0.3 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.0 0.0

E B (W.m-2) 0.0004 0.0004 0.0004 0.0005 0.0019LB (W.sr -1 .m -2 ) 1.1 0.9 0.8 0.7 0.5

E B (W.m-2) 0.0004 0.0004 0.0004 0.0004 0.0003 0.0003 0.0003 0.0003 0.0003LB (W.sr -1 .m -2 ) 1.4 1.1 1.0 0.7 0.6 0.5 0.4 0.4 0.4

E B (W.m-2) 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0001LB (W.sr -1 .m -2 ) 0.1 0.1 0.1 0.1 0.1 0.1 0.0

E B (W.m-2) 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002LB (W.sr -1 .m -2 ) 0.3 0.2 0.2 0.2 0.1 0.1

E B (W.m-2) 0.0002 0.0002 0.0002LB (W.sr -1 .m -2 ) 0.4 0.3 0.3

LED 5900K

S70

M50

FL32

M50

FL32

Distance from below pole 1

LED 3200K

LED 4000K

Distance from below pole 1

LED 3200K

LED 4000K

LED 5900K

S70


Table A 7: Spill Light illuminance and luminance, along and towards property boundaries, at 3.0m above road surface

015300153002404812

Illuminance (lux)3.00.5Obstructed0.30.40.20.34.53.00.38.41.80.4Luminance (kcd.m

-2)8.905.50Obstructed1.071.801.1319.5013.608.9021.0025.003.701.07

Illuminance (lux)3.40.70.51.01.10.50.65.33.40.59.04.31.0Luminance (kcd.m

-2)10.505.501.072.353.601.3235.3416.9010.5033.6029.4011.302.35


-2)11.304.901.369.701.300.937.506.3011.307.4010.505.301.00


-2)8.809.181.609.373.951.5329.6511.008.8028.2128.9419.049.37


-2)7.501.620.351.531.170.604.129.007.505.003.502.951.78


-2)1.302.070.352.102.100.846.401.861.306.3010.405.202.10

M50

FL32

4m towards nearest residential boundary12m towards opposite residential

boundary

Near properties boundaries, parallel to roadway

LED 3200KLED

4000KLED

5900K

S70

From directly below light to nearest residential boundary

From directly below light to opposite residential boundary

Perpendicular to Roadway

Distance from plane of pole 1 (m)

l ,


Table A 8: Spill Light blue weighted irradiance, EB, and radiance, LB, along and towards property boundaries, at 3.0m above road surface

0 15 30 0 15 30 0 2 4 0 4 8 12EB (W.m-2) 0.00101 0.00017 Obstructed 0.00010 0.00013 0.00007 0.00010 0.00152 0.00101 0.00010 0.00283 0.00061 0.00013L B (W.sr -1 .m -2 ) 3.00 1.85 Obstructed 0.36 0.61 0.38 6.57 4.58 3.00 7.08 8.43 1.25 0.36

EB (W.m-2) 0.00185 0.00038 0.00027 0.00054 0.00060 0.00027 0.00033 0.00288 0.00185 0.00027 0.00488 0.00233 0.00054L B (W.sr -1 .m -2 ) 5.70 2.98 0.58 1.28 1.95 0.72 19.18 9.17 5.70 18.23 15.95 6.13 1.28

EB (W.m-2) 0.00318 0.00062 0.00018 0.00053 0.00053 0.00026 0.00026 0.00185 0.00318 0.00018 0.00353 0.00203 0.00053L B (W.sr -1 .m -2 ) 9.98 4.33 1.20 8.57 1.15 0.82 6.62 5.56 9.98 6.53 9.27 4.68 0.88

EB (W.m-2) 0.00054 0.00008 0.00002 0.00040 0.00012 0.00004 0.00004 0.00082 0.00054 0.00006 0.00174 0.00084 0.00040L B (W.sr -1 .m -2 ) 0.99 1.03 0.18 1.05 0.44 0.17 3.33 1.24 0.99 3.17 3.25 2.14 1.05

EB (W.m-2) 0.00146 0.00017 0.00011 0.00039 0.00022 0.00028 0.00006 0.00134 0.00146 0.00011 0.00134 0.00073 0.00034L B (W.sr -1 .m -2 ) 4.20 0.91 0.20 0.86 0.66 0.34 2.31 5.04 4.20 2.80 1.96 1.65 1.00

EB (W.m-2) 0.00061 0.00020 0.00051 0.00061 0.00036 0.00036 0.00015 0.00147 0.00061 0.00010 0.00371 0.00137 0.00061L B (W.sr -1 .m -2 ) 0.66 1.05 0.18 1.07 1.07 0.43 3.25 0.94 0.66 3.20 5.28 2.64 1.07

M50

FL32

Distance from plane of pole 1

LED 3200K

LED 4000K

LED 5900K

S70

Perpendicular to Roadway4m towards nearest residential boundary 12m towards opposite residential From directly below light to nearest From directly below light to opposite residential

Near properties boundaries, parallel to roadwayr t


13.2 Photometric Modelling of spill light

For modelling of LED sources, the following (supplied) files were used:

• 214201 StreetLED Eco.ies • 214213.ies • ll17025.ies

Representative example photometric data were sourced for modelling of the traditional

streetlighting sources:

• Urban S70D – 201145.ies • Urban M50D – 201076.ies • Light-Street-SLA Suburban ECO-CFL32.ies

Calculations were made of vertical illuminance (facing the source) based on the same

measurement geometry as employed for spill light assessment in the field. This geometry is

shown in Figure A3.

Figure A 3: Map of spill light assessment field measures – vertical illuminance and luminance – red crosses indicate a measurement point at 1.5 m above the ground and facing towards the luminaire – red crosses continue in 3 m intervals until the next luminaire (from 40.5 m to 66 m, depending on installation). Green crosses indicate a measurement point at 3.0 m above the ground and facing towards the luminaire.

Readings continue every 3 m to next luminaire

Luminaire

Pole

+15 m +30 m

+4 m

+3 m

+12 m

Near residential boundary

Opposite residential boundary

+8 m

+6 m

-4 m

-2 m


The results of spill light assessment are presented here alongside modelled results for reference;

but they are not intended to be used as validation of the modelling. Several factors lead to

differences in the modelled vs measured values, one of the most significant of these is the

presence of ambient light and reflected light from all exterior surfaces (not modelled). This is

particularly an issue for streetlighting assessment where ambient light is of the same order of

magnitude as – or greater than - some of the lowest readings of spill light due to the modelled

source. Ambient light was minimised in field measurements by scheduling site visits at the

time of the new moon (minimising moonlight). Furthermore, the uncertainties of: source

photometric data, luminaire manufacturing tolerances, luminaire installation, field

measurement instruments, and field measurement position; all combine to create variation

between modelled and measured data.

The photometric data was first used to calculate the ULOR values for each luminaire: for the

LED luminaires, this was calculated at 0° upcast and 5° upcast. For the traditional luminaires,

this was only calculated at 5° upcast. These results are presented in Table A 9.

Table A 9: Calculated ULOR at 0° and 5° upcast for street lighting luminaires

The results show that selecting luminaires with minimal ULOR and reducing the upcast both

act to significantly reduce the upwards lumens. In this particular example, the LED luminaires

are designed to direct 100 to 10 times fewer lumens upwards (depending on upcast) than the

HPS at 5° upcast. It is noted that the total lumen output for the HPS lamp modelled here is 2.4

– 3.8 times greater than the other lamp types.

It should be also be noted that direct (upwards) light is not the only contributor to sky glow,

with reflected light also a major factor in light pollution. The modelled vertical illuminances

for spill light are summarised in Table A 10, followed by actual field measurements of

illuminance for comparison in Table A 11.

0° upcast 5° upcast 0° upcast 5° upcastLED 3200K 18W 0.2% 1.8% 2100 3.8 39LED 4000K 20W 0.02% 0.2% 1725 0.4 4.1LED 5900K 17W 0.2% 2.1% 1517 3.7 33

HPS 70W 6.9% 5800 401MV 50W 4.6% 1800 83CFL 32W 4.1% 2400 97

ULORLamp

lumens (lm)

Upwards lumens (lm)


Table A 10: Modelled values of vertical illuminance facing the street light (height 3m, in lux) for 6 different luminaires types for 0° and 5° upcast. Locations for assessment are highlighted by green crosses in Figure A 3 above.

Table A 11: Measured values of vertical illuminance facing the street light (height 3m, in lux) for 6 different luminaires types. Locations for assessment are consistent with Table A 10 above.

0 15 30 0 15 30 0 2 4 0 4 8 120 degree upcast 3.8 0.87 0.07 0.6 0.06 0.049 - 7.5 3.8 - 7.3 1.6 0.65 degree upcast 3.2 0.54 0.004 0.7 0.26 0.15 - 8.2 3.2 - 8.2 1.9 0.70 degree upcast 2.8 0.89 0.011 0.3 0.07 0.025 - 5.4 2.8 - 4.5 1.7 0.35 degree upcast 2.3 0.48 0.0002 0.8 0.37 0.18 - 5.7 2.3 - 4.8 2.5 0.80 degree upcast 2.0 0.76 0.03 0.2 0.07 0.032 - 4.0 2.0 - 3.5 1.3 0.25 degree upcast 1.6 0.48 0.003 0.5 0.24 0.14 - 4.1 1.6 - 3.5 1.8 0.5

HPS 70W 2.5 0.04 0.001 0.2 0.06 0.013 - 4.4 2.5 - 2.5 0.6 0.2MV 50W 2.8 0.03 0.0004 0.2 0.07 0.015 - 5.0 2.8 - 2.7 0.7 0.2CFL 32W 0.8 0.04 0.0003 0.3 0.07 0.015 - 3.0 0.8 - 2.6 0.9 0.3

From directly below light to opposite residential boundary

From directly below light to nearest residential boundary

5 degree upcast

LED 3200K 18W

LED 4000K 20W

LED 5900K 17W

4m towards nearest residential boundary

12m towards opposite residential boundary


Near properties boundaries, parallel to roadway Perpendicular to Roadway

0 15 30 0 15 30 0 2 4 0 4 8 12LED 3200K 18W Illuminance (lux) 3.0 0.5 - 0.3 0.4 0.2 0.3 4.5 3.0 0.3 8.4 1.8 0.4LED 4000K 20W Illuminance (lux) 3.4 0.7 0.5 1.0 1.1 0.5 0.6 5.3 3.4 0.5 9.0 4.3 1.0LED 5900K 17W Illuminance (lux) 3.6 0.7 0.2 0.6 0.6 0.3 0.3 2.1 3.6 0.2 4.0 2.3 0.6

HPS 70W Illuminance (lux) 4.8 0.7 0.2 3.6 1.1 0.4 0.4 7.3 4.8 0.5 15.5 7.5 3.6MV 50W Illuminance (lux) 2.6 0.3 0.2 0.7 0.4 0.5 0.1 2.4 2.6 0.2 2.4 1.3 0.6CFL 32W Illuminance (lux) 1.2 0.4 1.0 1.2 0.7 0.7 0.3 2.9 1.2 0.2 7.3 2.7 1.2


Near properties boundaries, parallel to roadway Perpendicular to Roadway4m towards nearest residential 12m towards opposite residential From directly below light to From directly below light to opposite


The modelled spill light evaluation at 0° upcast shows that light is expected to be extinguished

to below 1 lux at 15m away from the source (assessed 3m above the ground along the property

boundary), and below 0.1 at 30m away from the source. The values for the opposite boundary

increase, up to 5 times greater, when the upcast is increased to 5°; however, these values all

remain relatively low, below 0.5 lux. Spill light values decrease slightly (~20%) along the near

property boundary (as expected) when upcast is increased.

Modelled values are generally within the correct order of magnitude for the actual measured

values where values are greater than 0.5 lux. For very low modelled values, the impact of

background (ambient) light on comparison with actual readings is greater. This is evidenced

by the non-zero readings made in the field at 0m (directly below) the luminaire. In the absence

of ambient or reflected light, these vertical illuminances are modelled to have zero value – in

reality, they have measured values ranging from 0.1 – 0.6 lux.

Table A 12 and Table A 13 give the complete measurements made in the field for the spill light

assessment.


Table A 12: Measured values of vertical illuminance and luminance (height 3m) facing the street light for field study of 6 different luminaires types. Maximum values along property boundary for each lamp type are highlighted. Locations for assessment are highlighted by green crosses in Figure A 3 above. Luminances are given in kcd/m2 (×1000 for value in cd/m2)

0 15 30 0 15 30 0 2 4 0 4 8 12Illuminance (lux) 3.0 0.5 Obstructed 0.3 0.4 0.2 0.3 4.5 3.0 0.3 8.4 1.8 0.4Luminance (kcd.m -2 ) 8.90 5.50 Obstructed 1.07 1.80 1.13 19.50 13.60 8.90 21.00 25.00 3.70 1.07Illuminance (lux) 3.4 0.7 0.5 1.0 1.1 0.5 0.6 5.3 3.4 0.5 9.0 4.3 1.0Luminance (kcd.m -2 ) 10.50 5.50 1.07 2.35 3.60 1.32 35.34 16.90 10.50 33.60 29.40 11.30 2.35Illuminance (lux) 3.6 0.7 0.2 0.6 0.6 0.3 0.3 2.1 3.6 0.2 4.0 2.3 0.6Luminance (kcd.m -2 ) 11.30 4.90 1.36 9.70 1.30 0.93 7.50 6.30 11.30 7.40 10.50 5.30 1.00Illuminance (lux) 4.8 0.7 0.2 3.6 1.1 0.4 0.4 7.3 4.8 0.5 15.5 7.5 3.6Luminance (kcd.m -2 ) 8.80 9.18 1.60 9.37 3.95 1.53 29.65 11.00 8.80 28.21 28.94 19.04 9.37Illuminance (lux) 2.6 0.3 0.2 0.7 0.4 0.5 0.1 2.4 2.6 0.2 2.4 1.3 0.6Luminance (kcd.m -2 ) 7.50 1.62 0.35 1.53 1.17 0.60 4.12 9.00 7.50 5.00 3.50 2.95 1.78Illuminance (lux) 1.2 0.4 1.0 1.2 0.7 0.7 0.3 2.9 1.2 0.2 7.3 2.7 1.2Luminance (kcd.m -2 ) 1.30 2.07 0.35 2.10 2.10 0.84 6.40 1.86 1.30 6.30 10.40 5.20 2.10

CFL 32w


LED 3200K 18W

LED 4000K 20W

HPS 70W

MV 50W

Perpendicular to Roadway4m towards nearest residential 12m towards opposite residential From directly below light to From directly below light to opposite

LED 5900K 17W

Near properties boundaries, parallel to roadway


Table A 13: Readings of vertical illuminance facing luminaire (pole 1) and luminance of luminaire for viewing positions shown by red crosses in Figure A 3 above. Measurements taken at 1.5 m above the ground, parallel to roadway.

0 3 6 9 12 15 18 21 24 27 30 33Illuminance (lux) 1.6 4.1 4.2 3.5 2.8 1.8 1.5 1.1 1.1 0.6 0.5 0.3Luminance (kcd.m -2 ) 21.6 28.8 18.2 14.8 12.3 8 5.1 3.6 2.8 2.2 1.8 1.4Illuminance (lux) 0.4 6.3 6.3 5.3 5 4.1 2.9 2 1.6 1.2 0.9 0.7Luminance (kcd.m -2 ) 33.8 41 30.1 23 20.8 16.6 11.7 7.6 5.8 4.6 3.4 2.2Illuminance (lux) 0.2 2.2 2.5 3.1 2.8 2.1 1.6 1.2 0.9 0.7 0.7 0.6Luminance (kcd.m -2 ) 7.2 8.2 12.1 14.2 13.6 9.5 7 4.9 3.5 2.7 2.3 1.9Illuminance (lux) 1 9.6 13 10.1 6.9 4.8 3 1.8 1.1 0.7 0.6 0.5Luminance (kcd.m -2 ) 30.5 38.3 56.4 46.2 29.7 19.8 11.6 6.3 3.68 2.72 1.95 1.48Illuminance (lux) 0.2 1.9 1.8 1.7 1.3 1.1 0.8 0.6 0.5 0.4 0.3 0.3Luminance (kcd.m -2 ) 5 5.4 6.1 6.2 5.2 4 2.8 2 1.5 1.1 0.84 0.66Illuminance (lux) 0.3 4 4.7 3.4 2.5 0.9 1.4 1 0.7 0.5 0.4 0.3Luminance (kcd.m -2 ) 6.6 9.4 17 13.1 10.2 7.7 5.5 3.86 2.6 1.82 1.34 0.99

36 39 42 45 48 51 54 57 60 63 66Illuminance (lux) 0.3 0.2 0.2 0.1 0.1 0.1 0.1 0 0.1 0.2 0.2Luminance (kcd.m -2 ) 1.1 0.83 0.63 0.5 0.4 0.3 0.25 0.18 0.15 0.13 0.11Illuminance (lux) 0.8 0.8 0.7 0.9 3.5 (46.5m)Luminance (kcd.m -2 ) 2.1 1.7 1.45 1.25 1Illuminance (lux) 0.5 0.5 0.4 0.4 0.3 0.3 0.3 0.3 0.3 (58.4m)Luminance (kcd.m -2 ) 1.6 1.3 1.1 0.8 0.7 0.6 0.5 0.4 0.4Illuminance (lux) 0.4 0.4 0.3 0.3 0.3 0.4 1Luminance (kcd.m -2 ) 1.18 0.93 0.78 0.65 0.54 0.46 0.4Illuminance (lux) 0.3 0.3 0.3 0.3 0.4 0.4Luminance (kcd.m -2 ) 0.54 0.43 0.37 0.28 0.26 0.21Illuminance (lux) 0.3 0.3 0.4 (40.5m)Luminance (kcd.m -2 ) 0.75 0.58 0.53

MV 50W

CFL 32w

Distance from below pole 1 (m)

LED 3200K 18W

LED 4000K 20W

LED 5900K 17W

HPS 70W

MV 50W

CFL 32w

Distance from below pole 1 (m)

LED 3200K 18W

LED 4000K 20W

LED 5900K 17W

HPS 70W


14 References

Abrahamsson, M. and J. Sjostrand (1986). "Impairment of contrast sensitivity function (CSF) as a measure of disability glare." Invest Ophthalmol Vis Sci 27(7): 1131-1136.

Algvere, P. V., J. Marshall and S. Seregard (2006). "Age-related maculopathy and the impact of blue light hazard." Acta Ophthalmol Scand 84(1): 4-15.

ANSIC78.377 (2011). Specifications for the Chromaticity of Solid State Lighting Products (2008), and its revision C78.377-2011. U.S.A., American National Standard Lighting Group.

Arunkumar, R., C. M. Calvo, C. D. Conrady and P. S. Bernstein (2018). "What do we know about the macular pigment in AMD: the past, the present, and the future." Eye (Lond) 32(5): 992-1004.

Barentine, J. C., C. E. Walker, M. Kocifaj, F. Kundracik, A. Juan, J. Kanemoto and C. K. Monrad (2018). "Skyglow changes over Tucson, Arizona, resulting from a municipal LED street lighting conversion." J Quant Spect Rad Trans 212: 10-23.

Behar-Cohen, F., C. Martinsons, F. Vienot, G. Zissis, A. Barlier-Salsi, J. P. Cesarini, O. Enouf, M. Garcia, S. Picaud and D. Attia (2011). "Light-emitting diodes (LED) for domestic lighting: any risks for the eye?" Prog Retin Eye Res 30(4): 239-257.

Bennie, J., T. W. Davies, D. Cruse and K. J. Gaston (2016). "Ecological effects of artificial light at night on wild plants." J Ecol 104(3): 611-620.

Beyer, F. R. and K. Ker (2009). "Street lighting for preventing road traffic injuries." Cochrane Database Systematic Rev (1): CD004728.

Bierman, A. (2012). "Will switching to LED outdoor lighting increase sky glow?" Light Res Technol 44(4): 449-458.

Boettner, E. A. and J. R. Wolter (1962). "Transmission of the ocualr media." Invest Ophthalmol Vis Sci 1: 776-783.

Bragard, I. and P. A. Coucke (2013). "[Impact of the use of Luminette(R) on well-being at work in a radiotherapy department]." Cancer Radiother 17(8): 731-735.

Bullough, J. D. (2000). "The blue-light hazard: A review." J Illumin Eng Soc 29: 6-14.

Bullough, J. D. (2009). "Spectral sensitivity for extrafoveal discomfort glare." J Mod Optics 56: 1518-1522.

Bullough, J. D., A. Bierman and M. S. Rea (2017). "Evaluating the blue-light hazard from solid state lighting." Int J Occup Saf Ergon: 1-10.

Bullough, J. D. and L. C. Radetsky (2013). Analysis of new Highway Lighting Technologies - Report Prepared for the National Cooperative Highway Research Program, Transportation Research Board of the National Academies, NY, Lighting Research Centre, Rensselaer Polytechnic institute, Troy.

Burkhart, K. and J. R. Phelps (2009). "Amber lenses to block blue light and improve sleep: a randomized trial." Chronobiol Int 26(8): 1602-1612.

CELMA-ELC (2011). Optical safety of LED lighting. Federation of National Manufacturers Association for Luminaires and Electrotechnical Components for Luminaires in the European Union.

Clark, E. and N. Lesniak. (2017). "Circadian Lighting Solutions Are Real and Important—Why Aren’t They Being Used?" Retrieved July 2018, from https://www.metropolismag.com/design/circadian-lighting-survey/.

CSAPH, A.M.A. (2016). "Human and Envoronmental Effects of Light Emitting Diode (LED) Community Lighting." CSAPH Report 2-A-16: 1-9.

Darzins, P., P. Mitchell and R. F. Heller (1997). "Sun exposure and age-related macular degeneration. An Australian case-control study." Ophthalmol 104(5): 770-776.

https://www.metropolismag.com/design/circadian-lighting-survey/

https://www.metropolismag.com/design/circadian-lighting-survey/


Davies, S., M. H. Elliott, E. Floor, T. G. Truscott, M. Zareba, T. Sarna, F. A. Shamsi and M. E. Boulton (2001). "Photocytotoxicity of lipofuscin in human retinal pigment epithelial cells." Free Radic Biol Med 31(2): 256-265.

Delori, F. C., D. G. Goger and C. K. Dorey (2001). "Age-related accumulation and spatial distribution of lipofuscin in RPE of normal subjects." Invest Ophthalmol Vis Sci 42(8): 1855-1866.

Downie, L. E., L. Busija and P. R. Keller (2018). "Blue-light filtering intraocular lenses (IOLs) for protecting macular health." Cochrane Database Syst Rev 5: CD011977.

Enezi, J., V. Revell, T. Brown, J. Wynne, L. J. Schlangen and R. Lucas (2011). "A “Melanopic” Spectral Efficiency Function Predicts the Sensitivity of Melanopsin Photoreceptors to Polychromatic Lights." J Biol Rhythms 26(4): 314-323.

Fekete, J., C. Sik-Lanyi and J. Schanda (2010). "Spectral discomfort glare sensitivity investigations." Ophthalmic Physiol Opt 30(2): 182-187.

Figueiro, M. G., K. Gonzales and D. Pedler (2016). "Designing with circadian stimulus." Light Des Appl 8: 30-34.

Fletcher, A. E., G. C. Bentham, M. Agnew, I. S. Young, C. Augood, U. Chakravarthy, P. T. de Jong, M. Rahu, J. Seland, G. Soubrane, L. Tomazzoli, F. Topouzis, J. R. Vingerling and J. Vioque (2008). "Sunlight exposure, antioxidants, and age-related macular degeneration." Arch Ophthalmol 126(10): 1396-1403.

Gibbons, R., F. Guo, A. Medina, T. Terry, J. Du, P. Lutkevich and Q. Li (2014). Design Criteria for Adaptive Roadway Lighting. US Department of Transportation, Federal Highway Administration. FHWA-HRT-14-051.

Gibbons, R., J. Meyer, T. Terry, R. Bhagavathula, A. Lewis, M. Flanagan and C. Connell (2015). Evaluation of the impact of spectral power distribution on driver performance. US Department of Transportation, Federal Highway Administration. FHWA-HRT-15-047.

Ham, W. T., Jr., H. A. Mueller, J. J. Ruffolo, Jr., D. Guerry, 3rd and R. K. Guerry (1982). "Action spectrum for retinal injury from near-ultraviolet radiation in the aphakic monkey." Am J Ophthalmol 93(3): 299-306.

Ham, W. T., Jr., H. A. Mueller and D. H. Sliney (1976). "Retinal sensitivity to damage from short wavelength light." Nature 260(5547): 153-155.

Ham, W. T., Jr., J. J. Ruffolo, Jr., H. A. Mueller, A. M. Clarke and M. E. Moon (1978). "Histologic analysis of photochemical lesions produced in rhesus retina by short-wave-length light." Invest Ophthalmol Vis Sci 17(10): 1029-1035.

Houser, K. (2017). "The AMA’s Misguided Report on Human and Environmental Effects of LED Lighting." LEUKOS 13(1): 1-2.

Howarth, P. A., G. Heron, D. S. Greenhouse, I. L. Bailey and S. M. Berman (1993). "Discomfort from glare: the role of pupillary hippus." Light Res & Technol 25: 37-44.

International Commission on Non-Ionizing Radiation, P. (2013). "ICNIRP guidelines on limits of exposure to incoherent visible and infrared radiation." Health Physics 105: 74-96.

International Dark-Sky Association. (2010). "Visibility, environmental, and astronomical issues associated with blue-rich white outdoor lighting." Tucson-Washington, DC.

International Dark-Sky Association. (2014). "Fixture Seal of Approval." Retrieved August 2018.

International Electrotechnical Commission (2006). "IEC 62471: 2006." Photobiological safety of lamps and lamp systems. Geneva: IEC.

International Electrotechnical Commission (2014). IEC/TR 62778, Application of IEC 62471 for the Assessment of Blue Light Hazard to Light Sources and Luminaires. Geneva, International Electrotechnical Commission.


International Well Building Institute (2016). WELL Building Standard (v1). New York, International Well Building Institute.

Kinzey, B. (2016). The Light Post, Official MSSLC e-newsletter. Washington, Municipal Solid-State Street Lighting Consortium.

Lawrenson, J. G., C. C. Hull and L. E. Downie (2017). "The effect of blue-light blocking spectacle lenses on visual performance, macular health and the sleep-wake cycle: a systematic review of the literature." Ophthalmic Physiol Opt 37(6): 644-654.

Leung, T. W., R. W. Li and C. S. Kee (2017). "Blue-Light Filtering Spectacle Lenses: Optical and Clinical Performances." PLoS One 12(1): e0169114.

Lin, Y., Y. Liu, Y. Sun, X. Zhu, J. Lai and I. Heynderickx (2014). "Model predicting discomfort glare caused by LED road lights." Opt Express 22(15): 18056-18071.

Ling, S. J., J. Sanny and B. Moebs. (2018). "6.1: Blackbody Radiation." Retrieved August 2018, from https://phys.libretexts.org/TextBooks_and_TextMaps/University_Physics/Book%3A_University_Physics_(OpenStax)/Map%3A_University_Physics_III_Optics_and_Modern_Physics_(OpenStax)/6%3A_Photons_and_Matter_Waves/6.1%3A_Blackbody_Radiation.

Lockley, S. W., G. C. Brainard and C. A. Czeisler (2003). "High sensitivity of the human circadian melatonin rhythm to resetting by short wavelength light." J Clin Endocrinol Metab 88(9): 4502-4505.

Longcore, T., A. Rodríguez, B. Witherington, J. F. Penniman, L. Herf and M. Herf (2018). "Rapid assessment of lamp spectrum to quantify ecological effects of light at night." J Exp Zool A Ecol Integr Physiol 329: 511-521.

Lucas, R. J., S. N. Peirson, D. M. Berson, T. M. Brown, H. M. Cooper, C. A. Czeisler, M. G. Figueiro, P. D. Gamlin, S. W. Lockley, J. B. O’Hagan, L. L. A. Price, I. Provencio, D. J. Skene and G. C. Brainard (2014). "Measuring and using light in the melanopsin age." Trends in Neurosci 37(1): 1-9.

Luginbuhl, C. B., P. A. Boley and D. R. Davis (2014). "The impact of light source spectral power distribution on sky glow." J Quant Spect Rad Trans 139: 21-26.

Lund, D. J., B. E. Stuck and P. Edsall (2006). "Retinal injury thresholds for blue wavelength lasers." Health Phys 90(5): 477-484.

Mainster, M. A. (2005). "Intraocular lenses should block UV radiation and violet but not blue light." Arch Ophthalmol 123(4): 550-555.

Mainster, M. A. (2006). "Violet and blue light blocking intraocular lenses: photoprotection versus photoreception." Br J Ophthalmol 90(6): 784-792.

Mainster, M. A. and G. T. Timberlake (2003). "Why HID headlights bother older drivers." Br J Ophthalmol 87(1): 113-117.

McCarty, C. A., B. N. Mukesh, C. L. Fu, P. Mitchell, J. J. Wang and H. R. Taylor (2001). "Risk factors for age-related maculopathy: the Visual Impairment Project." Arch Ophthalmol 119(10): 1455-1462.

Mellerio, J. (1987). "Yellowing of the human lens: nuclear and cortical contributions." Vision Res 27(9): 1581-1587.

Moon, J., J. Yun, Y. D. Yoon, S. I. Park, Y. J. Seo, W. S. Park, H. Y. Chu, K. H. Park, M. Y. Lee, C. W. Lee, S. J. Oh, Y. S. Kwak, Y. P. Jang and J. S. Kang (2017). "Blue light effect on retinal pigment epithelial cells by display devices." Integr Biol (Camb) 9(5): 436-443.

Noell, W. K., V. S. Walker, B. S. Kang and S. Berman (1966). "Retinal damage by light in rats." Invest Ophthalmol Vis Sci 5(5): 450-473.

Norren, D. V. and J. J. Vos (1974). "Spectral transmission of the human ocular media." Vision Res 14(11): 1237-1244.

O'Hagan, J. B., M. Khazova and L. L. Price (2016). "Low-energy light bulbs, computers, tablets and the blue light hazard." Eye (Lond) 30(2): 230-233.


Owsley, C. and G. McGwin, Jr. (2010). "Vision and driving." Vision Res 50(23): 2348-2361.

Pachito, D. V., A. L. Eckeli, A. S. Desouky, M. A. Corbett, T. Partonen, S. M. Rajaratnam and R. Riera (2018). "Workplace lighting for improving alertness and mood in daytime workers." Cochrane Database Syst Rev 3: CD012243.

Pawlak, A., M. Rozanowska, M. Zareba, L. E. Lamb, J. D. Simon and T. Sarna (2002). "Action spectra for the photoconsumption of oxygen by human ocular lipofuscin and lipofuscin extracts." Arch Biochem Biophys 403(1): 59-62.

Pawson, S. and M.-F. Bader (2014). "LED lighting increases the ecological impact of light pollution irrespective of color temperature." Ecol Appl 24(7): 1561-1568.

Pro-Lite Technologies. (2018). "LED Colour Temperatures –Are Your White Giving You the Blues? " Retrieved July 2018, from http://www.pro-lite.co.uk/File/colour_temp.php.

Rea, M. and A. Bierman (2015). "Spectral considerations for outdoor lighting: Consequences for sky glow." Light Res Technol 47(8): 920-930.

Rea, M. S., A. Bierman, M. G. Figueiro and J. D. Bullough (2008). "A new approach to understanding the impact of circadian disruption on human health." J Circ Rhythms May 29; 6: 7.

Rea, M. S., M. G. Figueiro, A. Bierman and J. D. Bullough (2010). "Circadian light." J Circ Rhythms 8(1): 2.

Rea, M. S., M. G. Figueiro, J. D. Bullough and A. Bierman (2005). "A model of phototransduction by the human circadian system." Brain Res Brain Res Rev 50(2): 213-228.

Rea, M. S. and G. M. Figueiro (2016). Response to the 2016 AMA Report on LED Lighting. Troy, Lighting Research Center.

Rich, C. and T. Longcore (2006). Ecological consequences of artificial night lighting. Washington, DC, Island Press.

Sliney, D. H. (2002). "How light reaches the eye and its components." Int J Toxicol 21(6): 501-509.

Sliney, D. H. (2005). "Exposure geometry and spectral environment determine photobiological effects on the human eye." Photochem Photobiol 81(3): 483-489.

Somers-Yeates, R., J. Bennie, T. Economou, D. Hodgson, A. Spalding and P. K. McGregor (2016). "Light pollution is associated with earlier tree budburst across the United Kingdom." Proc. R. Soc. B 283(1833): 20160813.

Sparrow, J. R. and M. Boulton (2005). "RPE lipofuscin and its role in retinal pathobiology." Exp Eye Res 80(5): 595-606.

Sparrow, J. R., A. S. Miller and J. Zhou (2004). "Blue light-absorbing intraocular lens and retinal pigment epithelium protection in vitro." J Cataract Refract Surg 30(4): 873-878.

Sperling, H. G., C. Johnson and R. S. Harwerth (1980). "Differential spectral photic damage to primate cones." Vision Res 20(12): 1117-1125.

Steinbach, R., C. Perkins, L. Tompson, S. Johnson, B. Armstrong, J. Green, C. Grundy, P. Wilkinson and P. Edwards (2015). "The effect of reduced street lighting on road casualties and crime in England and Wales: controlled interrupted time series analysis." J Epidemiol Com Health 69(11): 1118-1124.

Steinberg, R. H. (1994). "Survival factors in retinal degenerations." Curr Opin Neurobiol 4(4): 515-524.

Stone, E., S. Harris and G. Jones (2015). "Impacts of artificial lighting on bats: a review of challenges and solutions." Mammalian Biology-Zeitschrift für Säugetierkunde 80(3): 213-219.

Sui, G. Y., G. C. Liu, G. Y. Liu, Y. Y. Gao, Y. Deng, W. Y. Wang, S. H. Tong and L. Wang (2013). "Is sunlight exposure a risk factor for age-related macular degeneration? A systematic review and meta-analysis." Br J Ophthalmol 97(4): 389-394.

Sweater-Hickcox, K., N. Narendran, J. D. Bullough and J. P. Freyssinier (2013). "Effect of different

http://www.pro-lite.co.uk/File/colour_temp.php


coloured luminous surrounds on LED discomfort glare perception." Light Res Technol 45: 464-475.

Taylor, H. R., B. Munoz, S. West, N. M. Bressler, S. B. Bressler and F. S. Rosenthal (1990). "Visible light and risk of age-related macular degeneration." Trans Am Ophthalmol Soc 88: 163-173; Discussion 173-168.

Tosini, G., I. Ferguson and K. Tsubota (2016). "Effects of blue light on the circadian system and eye physiology." Mol Vis 22: 61-72.

Turner, P. L. and M. A. Mainster (2008). "Circadian photoreception: ageing and the eye's important role in systemic health." Br J Ophthalmol 92(11): 1439-1444.

USDOE (2013). Optical Safety of LEDs: Building Technologies Office Solid-Sate Lighting Technology Fact Sheet. US Department of Energy.

USDOE (2017). An Investigation of LED Street Lighting's Impact on Sky Glow, US Department of Energy; Office of Energy Efficiency and Renewable Energy.

USDOE (2017). Street Light and Blue Light - Frequently Asked Questions, US Department of Energy; Office of Energy Efficiency and Renewable Energy.

van Norren, D. and J. van de Kraats (2007). "Spectral transmission of intraocular lenses expressed as a virtual age." Br J Ophthalmol 91(10): 1374-1375.

Viola, A. U., L. M. James, L. J. Schlangen and D. J. Dijk (2008). "Blue-enriched white light in the workplace improves self-reported alertness, performance and sleep quality." Scand J Work Environ Health 34(4): 297-306.

Whitaker, D., R. Steen and D. B. Elliott (1993). "Light scatter in the normal young, elderly, and cataractous eye demonstrates little wavelength dependency." Optom Vis Sci 70(11): 963-968.

Wolf, G. (2003). "Lipofuscin and macular degeneration." Nutr Rev 61(10): 342-346.

Wu, J., S. Seregard and P. V. Algvere (2006). "Photochemical damage of the retina." Surv Ophthalmol 51(5): 461-481.

Zeitzer, J. M., D. J. Dijk, R. Kronauer, E. Brown and C. Czeisler (2000). "Sensitivity of the human circadian pacemaker to nocturnal light: melatonin phase resetting and suppression." J Physiol 526: 695-702.

Documents

Assessment of blue light hazards and correlated colour