light & engineering

Volume 20, Number 1, 2012

LIGHT & ENGINEERING

Znack Publishing House, Moscow

ISSN 0236-2945

Volu

me

20,

Num

ber

1, 2

012

LIG

HT &

EN

GIN

EER

ING

Editor-in-Chief: Julian B. Aizenberg

Associate editor: Sergey G. Ashurkov

Editorial Board: Lyudmila V. Abramova Leonid B. PrikupetsArtyom E. Ataev Vladimir M. PyatigorskyVictor V. Barmin Anna G. ShakhparunyantsVladimir P. Budak Alexei K. SolovyovAndrey A. Grigoryev Raisa I. StolyarevskayaAlexei A. Korobko Alexander I. TereshkinDmitry O. Nalogin Konstantin A. TomskyAlexander T. Ovcharov Leonid P. Varfolomeev

Moscow, 2012

Foreign Editorial Advisory Board:

Lou Bedocs, Thorn Lighting Limited, United KingdomWout van Bommel, Philips Lighting, the NetherlandsPeter R. Boyce, Lighting Research Center, the USALars Bylund, Bergen’s School of Architecture, NorwayStanislav Darula, Academy Institute of Construction and Architecture, Bratislava, Slovakia Peter Dehoff, Zumtobel Lighting, Dornbirn, AustriaMarc Fontoynont, Ecole Nationale des Travaux Publics de l’Etat (ENTPE), FranceFranz Hengstberger, National Metrology Institute of South AfricaWarren G. Julian, University of Sydney, AustraliaZeya Krasko, OSRAM Sylvania, USAEvan Mills, Lawrence Berkeley Laboratory, USALucia R. Ronchi, Higher School of Specialization for Optics, University of Florence, ItalyJanos Schanda, University of Veszprem, HungaryNicolay Vasilev, Sofi a Technical University, BulgariaJennifer Veitch, National Research Council of Canada

LIGHT & ENGINEERING(Svetotekhnika)

NEW WEB SITE OF JOURNALS«SVETOTEKHNIKA»

and «LIGHT & ENGINEERING»http://www. sveto-tekhnika.ru

At the end of 2011, the new web site www.sveto-tekhnika.ru for Journals “Svetotekhnika” – Russian version and “Light & Engineering”- English version had been opened.

Now it is possible to download from this site full version of the Journals for the last years for free.

In nearest future the following paid services will be organized:

- Online sale of defi nite papers from Journals;

- Online subscription to Journals.

Also will be organized the paper submission by the site accounts for authors.

There is a forum on the web site, where users can ask and discuss with Editorial Board and with each other some questions in scientifi c and technical aspects also as well as in organizing sphere.

3

Low BedocsLou Bedocs is a freelance Lighting Applications Advisor to Thorn Lighting dealing with external affairs and standards. He is a Chartered electrical and lighting engineer worked for Thorn for over 50 years in various technical and marketing roles in the lighting business and in 2010 retired from his post as Lighting Applications Director. Lou continues to chair and serve on many National, European and International standards and Learned Institutes committees. He is Honorary President of the Lighting Industry Federation (LIF) and provides technical links with CELMA and relevant UK and EU government departments, visiting lecturer to a number of UK Universities and published several technical papers

Wout Van Bommel Prof. Wout van Bommel worked for more than 35 years with Philips Lighting in different lighting application functions. For the period 2003–2007 he has been President of the International Lighting Commission, CIE. From 1988 to 2008 he was the Dutch representative of the European Lighting Normalisation Committee CEN, TC 169. He is a board member of the Dutch “Light & Health Research Foundation”, SOLG. Wout van Bommel was in 2004 appointed Consulting Professor at the Fudan University of Shanghai and in 2008 External Examiner of the Master Course “Light and Lighting” at the University College of London (UCL- Bartlett Institute). He has published more than 150 papers in national and international lighting journals in different languages. He is the author of the book “Road Lighting”. After his retirement from Philips Lighting, he advices, as an independent Lighting Consultant

Peter R. Boyce Peter Boyce is Professor Emeritus at Rensselaer Polytechnic Institute in Troy, New York, USA. From1966–1990 he was a Research Offi cer at the Electricity Council Research Centre in England. From 1990–2004 he was Head of Human Factors at the Lighting Research Center at Rensselaer Polytechnic Institute. Since 2008 he has been the Technical Editor of the journal “Lighting Research and Technology”. He is a Fellow of both the Society of Light and Lighting and of the Illuminating Engineering Society of North America and has received awards from both bodies for his work. He is a recognized authority on the interaction of people and lighting, being the author of the classic text “Human Factors in Lighting”, as well as numerous book chapters, papers and articles

Lars BylundLars Bylund is a professor at Bergen Arkitekthögskole, Bergen NorwayHollistic Building, Light and Energy, at present he is Invited Professor, Faculty of Architecture, University of Ljubljana, Slovenia and lighting designer.He awarded by "Lighting Design of the Year" in England, Finland, Norway and Sweden. Prof. Bylund is working with a number of leading European lighting fi xture manufacturers and he is member of CIE and IEA working committees for daylighting and lighting

Light & Engineering Foreign Editorial Advisory Board

2012

4

Stanislav Darula, Ph.D., Head of the Building Physics Department in Institute of Construction and Architecture of Slovak Academy of Sciences in Bratislava. Stanislav is a national member of the CIE Div. 3, member of CIE TC 3.25, TC 3.49, TC 3–51. Also, he is a member of the Presidium and chairmen of the Slovak National CIE TC 8 Daylighting, member of the ISES (International Solar Energy Society), member of the SKSI (Slovak Chamber of Civil Engineers), member of the SZSI (The Slovak Union of Building Engineers), member of the SSTS (Slovak Illuminating Engineering Society). His fi eld of science is research in building physics, daylighting, sun energy utilisation, energy performance of buildings, daylighting design in the buildings. He is an external lecturer at Slovak Technological University in Bratislava and a consultant for daylighting design

Peter Dehoff Dipl. Ing. Peter Dehoff is the Director of Professional Associations and Standardisation, Lighting Application Management in Zumtobel Lighting, Dornbirn, Austria. He is the President of Austria CIE National Committe, Representative from Austria in Div. 3 CIE, Chair of TC 3–49, member of numerous associations and standardisation organisations (CIE, CEN, ASI, DIN, ZVEI, AK licht.de, LiTG, LTG, CELMA, FEEI). Peter Dehoff graduated from Karlsruhe TU and since 1987 with Zumtobel Department of Strategic Applications, responsible for fi nding and developing trends in modern lighting applications and proving them to the market. Special topics: quality of light, lighting and the wellbeing of people, energy effi ciency, aspects of dynamic lighting

Marc Fontoynont Marc Fontoynont is professor at the ENTPE National Engineering School of State Public Works in Lyon, France and Head of the Building Sciences Laboratory. He is also the Vice-President of the International Lighting Commission (Vienna, Austria), First Vice-President of the French Lighting Association and the Vice-President and co-founder of the Cluster Lumière (Lyon, France). He has been working for about 30 years in lighting and daylighting optimization, starting in Lawrence Berkeley National Laboratory, California. He focuses on energy effi cient schemes, with integrating energy concern, life cycle cost reduction, improved user satisfaction. He is recipient of the Fresnel Medal and the Alfred Monnier Lighting Award. He is now Operating Agent of the Solid state Lighting Annex of the IEA (International Energy Agency), 4 E Implementing Agreement, aiming at improving reliability of tests of quality and performance of SSL worldwide

Franz Hengstberger Dr. Franz Hengstberger is the head of National Physical Research Laboratory in Council for Scientifi c and Industrial Research at National Metrology Institute of South Africa, (CSIR NML), Pretoria. In period of time 2007–2011 he was the President of CIE. Dr. F. Hengstberger is the member of the International Committee for Weights and Measures represented CSIR NML, President of Consultative Committee in Photometry and Radiometry (CCPR) and wide world well known specialist in fi eld of photometry, radiometry and metrology

5

Warren G. JulianDr. Warren Julian is an Associate Professor, coordinator of Building Services Program and Illumination Program at the Faculty of Architecture, Design and Planning of the University of Sydney, Editor of the Journal Lighting. He has been recipient of the next awards: Building Science Prize 1970 for History of Building Science; AP Turnbull Award, 1978 Illuminating Engineering Society for an outstanding contribution to lighting education; Life Fellowship, 1988 Illuminating Engineering Society for an outstanding contribution to the art and science of lighting in Australia. His major honorary appointments are: President of National Council of Illuminating Engineering Society of Australia and New Zealand (IESANZ); Vice-President CIE (Publications), Vice-President CIE (Technical); Member, (for Australia) CIE General Assembly, CIE Chair, Divisional Director, CIE Board of Administration Member, Lux Pacifi ca Organisation Chair, CIE Australian Member of Division VII, Chair of Australian National Committee on Illumination (now CIE Australia). He is an author of 180 books, book chapters, scientifi c papers and reports

Zeya Krasko Zeya Krasko, Ph.D. in technical sciences, graduated from Moscow Power Institute in 1960. She worked for almost 40 years in research and development of metal halide lamps fi rst at VNISI (Moscow, Russia), then at GTE/SYLVANIA and OSRAM SYLVANIA (Boston, USA). She published more than 30 technical papers, received six Russian and 15 US patents, and two major Sylvania company technical awards. From 2001 she has been participating in technical editing and translation papers from Russian into English for the Light & Engineering Journal

Evan MillsDr. Evan Mills graduated from Swedish University of Lund in 1991 by specialty in ecology and energy effi ciency systems, and he is in stuff of Lawrence Berkley National Laboratory for a long time

Lucia R. Ronci Prof. Lucia R. Ronchi (Florence, Italy) is a physics graduated from the University of Florence, in 1948. She had got her Ph.D. in Physiological Optics, in 1955. Since 1949, experimenting at the INO (National Institute of Optics, Florence) in Visual Science, psychophysics, electrophysiology, theoretical backgrounds. From 1983 to 1991, Dr. Ronchi was the Director of CIE Division 6 “Photochemistry and Photobiology”, which has a deal with visual and non visual effects of optical radiation. From 1993 to 1997, she is the President of AIC (International Colour Association). In 2011, JUDD/AUC awarded. At present she is working in advanced Visual Science and its Applications

6

Janos Schanda Dr. János Schanda is Professor Emeritus at the University of Pannonia, Hungary. He heads at present the “Virtual Environment and Imaging Technologies Laboratory”. The Hungarian Academy of Sciences granted him the degree of “Doctor of Technical Sciences” for his thesis work on colour rendering. During the 1980 –1990 he worked for the International Commission on Illumination (CIE). Between 2007 and 2011 he was the Vice President Technical of the Commission. Dr. Schanda is member of the Optical Society of America, of The Society for Imaging Science and Technology and of several Hungarian Societies in the fi elds of light and lighting and optical measurements. He served also on the Board of the International Colour Association (AIC) as its vice-president. He is on the editorial / international advisory board of Color Res. & Appl., USA, Light & Engineering, Russia, Lighting Research & Technology, UK, and Journal of Light & Visual Environment, Japan. Since 2010 he is member of the Advisory Board of the Colour & Imaging Institute, Art & Science Research Centre, Tsinghua University, China; since 2011 – of the Centre for Colour Culture and Informatics (C3 I) of Taiwan. In 2010 the British Colour Group awarded him with the Newton Medal, in 2011 CIE presented him the “De Boer Pin”. He is author of over 600 technical papers and conference lectures

Nicolay Vasilev Dr. Vasilev has come a 57-years long “lighting” way in the Sofi a Technical University! Along with teaching, 30 years ago he has founded “Scientifi c research lighting laboratory” with the practical orientation of investigations in next directions: rationalization of lighting systems operating modes, optimization of luminous distribution and of the arrangement of street and tunnel luminaries, investigation of road surface refl ection properties, adaptive street and tunnel lighting, revaluation of street lighting in the mesopic vision, discomfort glare and others. The Laboratory turned into a school of lighting, and under his guidance 16 research associates defended doctors hips, two of them have already become professors and 7 – associated professors. During these years, he has published 5 books, a guidebook and teaching materials in the fi eld of lighting techniques and around 300 articles and presentations. Particularly fruitful proved to be in collaboration with CIE and with most lighting centers and scientists all over Europe. There should be mentioned 45-years long continual co-operation with the Moscow Power Engineering Institute, VNISI, and Journal Svetotekhnika

Jennifer Veitch Dr. Veitch is a Senior Research Offi cer at the National Research Council of Canada, where she leads research into lighting effects on health and behaviour. She joined NRC in 1992 following the completion of her Ph.D. in environmental psychology at the University of Victoria, British Columbia, Canada. She is best known for her research on lighting quality, which has infl uenced lighting design recommendations in North America through the Illuminating Engineering Society of North America Lighting Handbook (IES) and the IES design guide Light + Design: A Guide to Designing Quality Lighting for People and Buildings. She is a Fellow of the Canadian Psychological Association, the American Psychological Association, and the Illuminating Engineering Society of North America. In 2011 she received the Waldram Gold Pin Award for Outstanding Contributions to Applied Illuminating Engineering from the CIE. Among various volunteer roles in both the lighting and psychology, she currently serves as Director of CIE Division 3

7

CONTENTS

VOLUME 20 NUMBER 1 2012

LIGHT & ENGINEERING(SVETOTEKHNIKA)

Foreign Editorial Advisory Board 2012 3

Nicolai N. Usov Prospective Use of Organic Light Emission Diodes for Information Displays and for Illumination 9

Carl GardnerTackling Unwanted Light: An International Perspective 24

Alexei K. Solovyov Hollow Tubular Light Guides: Their Application for Natural Illumination of Buildings and Energy Saving 40

Roger NarboniThe Old City of Jerusalem Lighting Master Plan 50

Julia B. Babanova and Vadim A. LunchevEnergy Saving Capabilities When Using Control Systems for Interior Illumination 58

Bryan King Development of a Road and Urban Lighting Holistic Assessment Model 66

Pål J. Larsen Use of LED for Road Lighting 75

Annu Haapakangas, Jukka Keränen, Marko Nyman, and Valtteri HongistoLighting Improvement and Subjective Working Conditions in an Industrial Workplace 86

Matthias Lindemann, Robert Maass, and Georg Sauter A Brief History of Traceable Goniophotometry at PTB 97

Werner Halbritter, Werner Horak, and Werner JordanSimplifi ed Approach for Classifi cation the Potential Photobiological Hazards of LEDs According to CIE S009 113

Paola Iacomussi, Giuseppe Rossi, and Laura RossiA Comparison Between Different Light Sources Induced Glare on Perceived Contrast 121

Wouter R.A. Ryckaert, Inge A.A. Roelandts, Mieke Van Gils, Guy Durinck, Stefaan Forment, Jan Audenaert, and Peter Hanselaer Performance of LED Linear Replacement Lamps 129

Contents #2 140

8

Scientifi c EditorsSergey G. Ashurkov Raisa I. Stolyarevskaya

Style EditorMarsha Vinogradova

Art and CAD EditorAndrey M. Bogdanov

Editorial Offi ce:VNISI, Rooms 327 and 334 106 Prospekt Mira, Moscow 129626, Russia Tel: +7.495.682.26.54 Tel./Fax: +7.495.682.58.46 E-mail: [email protected] http://www.sveto-tekhnika.ru

Znack Publishing House P.O. Box 648, Moscow, 101000, Russia Tel./Fax: +7.495.361.93.77

© Svetotekhnika, 2009© Znack Publishing House, 2009

Moscow Power Engineering Institute Press

9

Light & Engineering SvetotekhnikaVol. 20, No. 1, pp. 9-23, 2011 No. 5, 2011, pp. 4-14

of OLEDs is considered to be in the sphere of na-notechnology. Due to a wide variety of organic ma-terials (low-molecular, polymeric, dendrymeric), in which electroluminescence takes place, one can form functional layers using both the vacuum ther-mal evaporation method, and deposition from so-lutions. The latter one for the present has not been widely used because of some diffi culties of forming great numbers of different functional layers from solutions. However, it will open wide possibilities hereafter to manufacture devices with OLEDs using inexpensive methods of printer or offset printing.

As opposed to inorganic light emitting diodes, OLEDs make it possible to manufacture lighting panels uniformly, with a high quality of colour rendi-tion comfortable for the human eye. Moreover, pro-duction of transparent and fl exible OLEDs, as well as devices with them is possible.

The progress of OLED displays is based on an earlier developed technology of LC displays, and the fi rst ones passed through the same stages of devel-opment: from passive matrix devices (PMOLED) to displays with active addressing (AMOLED). From year to year, size of substrates being processed in-creases, and similar to LC display technology, the processing of 2500×3000 mm size substrates can be expected. OLED production technology inten-sively develops, and the main attention is given to the development of new approaches which increase electroluminescence effi ciency, to the development of new more effective materials and types of device structure, design, production technology and to the increase of OLED display’s service life [5–7].

ABSTRACT

Application of thin-fi lm organic light emission diodes (OLED) is prospective for development of a new generation of information display systems and energy saving illumination. Progress of OLED dis-plays is based on the success of the earlier deve loped LC display technology, and like the latter ones, they passed the same stages of development: from passive matrix devices to displays with active addressing. In recent years, development of power effi cient light sources based on OLEDs was conducted. OLEDs continue to develop, and in doing so, the main at-tention is given to the search for new methods of in-creased electroluminescence effi ciency, to the search for new materials, types of device structure and de-sign, as well as an increased durability of devices with OLEDs.

In the article, wide possibilities of OLED appli-cation in systems of information display and of illu-mination are presented.

Keywords: organic light emission diode, OLED, display, illumination

1. INTRODUCTION

OLED creation research and development began from the late eighties [1–4]. In the beginning, the ba-sic purpose of OLEDs were displays, fi rst of all for mobile devices. However, in the early 2000 s, de-velopment of light sources based on OLEDs began. OLEDs consist of many functional layers of 1–100 nm thickness each, and so production technology

PROSPECTIVE USE OF ORGANIC LIGHT EMISSION DIODES FOR INFORMATION DISPLAYS AND FOR ILLUMINATION

Nicolai N. USOV

Cyclone Central Scientifi c-and-Research Institute Open Society, Moscow E-mail: [email protected]

Light & Engineering Vol. 20, No. 1

10

of OLED display production capacities. In 2010, Samsung Mobile Displays (SMD), a leader in this market, manufactured about three million displays a month with diagonal size of 2.0–4.3 inches. In the same 2010, SMD and LG Displays (LGD) companies began building of new production lines of the 5th and 8th generations to manufacture OLED displays, which should be commissioned in 2011–2012. AU Optronics (AUO) builds a line of the 3.5 generation and is going to reorient its line of the 4.5th genera-tion from production of LC displays, in which thin-fi lm transistors based on low-temperature polysili-con (LTPS TFT) are used for control, to production of OLED displays with an active matrix on the ba-sis of LTPS TFT (AMОLED). With commissioning of new lines in 2011 and 2012, the production capac-ities will increase several hundred times.

In the second quarter of 2011, SMD Company commissioned a new technological line of the 5.5th generation (Fig. 2), which fi rst of all will manufac-ture displays for widely popular smartphones Gal-axy S [9]. It should be noted that LTPS TFT panels

OLED displays are intended to replace LC dis-plays, and OLED light sources start to revolutionise the illumination world.

2. OLED DISPLAYS

OLED displays have many advantages in com-parison with LCs (Fig. 1), plasma and fi eld emission displays: saturated bright colours, fast response, high luminance, high image contrast, wide viewing angle, a smaller power consumption, low working voltage and wide functioning temperature interval.

The products with state-of-the-art OLED dis-plays, which are available for sale, range from dis-plays by LG Company of 15 inches (38.1 cm) di-agonal size, to microdisplays of MicroOLED Com-pany with a diagonal size of 0.38 inches. The main application of OLED displays at present is mo-bile devices: mobile phones, smartphones, photo and video cameras, digital audio players. Wide use of OLED displays in other devices, such as iPhones, iPads, photo and video cameras, is prevented by lack

Table. Characteristics of technological lines of generations (G) 4.5 and 5.5 [11]

Diagonal sizeof OLED display,

inches

Total number of displays on a being processed plate Advantageof new linesG5,5/G4,5

G4.5730×920 mm

G5.51300×1500 mm

3 150 486 3.3×

4.3 72 264 3.7×

10.1 8 50 6.3×

15.6 4 18 4.5×

≥ 30 30 inches×240 inches ×1

32 inches×657 inches×2 > 2×

Fig. 1. Comparison of images on an OLED screen and on LC TV screen with back illumination using inorganic light

emission diodes [8]

Fig. 2. A production site of 5.5 generation of Samsung Mobile Displays Company


11

mentioned TVs. OLEDnet Company in March 2010 published a forecast of OLED display and TV de-velopment for 2010–2016 [15]. According to this, a production technology of TVs with OLED screens up to 30 inches diagonal size is ready for today. For profi table production of TVs with screens of 40 inch-es and more to be managed, large-scale investments, further perfection of the technology and conversion to glass panels of the 8th and 9th generations are nec-essary. In order not to lose out in the competition for the market of 3D TVs and to occupy a comfortable niche, production of these new TVs should begin not later than in 2013. One should also note that OLEDs as plain light sources of a large area, are most suit-able for back illumination of LC displays.

Companies SMD and LGD intend to invest $17 billion in OLED display and TV production devel-opment until 2015 [16]. A large-scale production of OLED TVs is planned to begin in 2014.

Ample opportunities of OLED technology are shown by Mitsubishi Electric Company, which car-ries out development of big size OLED screens. In 2009–2010 Mitsubishi presented 149-and 155-inch OLED screens at some exhibitions (Fig. 4) [17, 18].

At the turn of September 2010, Mitsubishi Elec-tric Company began sales of Diamond Vision ОLED big screens at a price of $400, 000. The fi rst screen of 3.84×2.3 m size installed in a research center of Merck Company in Darmstadt (Germany), con-sists of 60 units containing 128×128 pixels with passive-matrix addressing. Luminance of the screen is 1200 cd/ m2.

Fig. 5 shows a round screen with a diameter of more than 3 m and with height of more than 1 m. Mitsubishi Electric Company presented this screen at the ISE 2011 exhibition. And a display

will be produced for this line by Sintek Photronics Company [10], which builds for this purpose a new line of the 5.5th generation with the designed capac-ity of 80,000 panels a month. It will also supply sen-sory panels.

Some idea about levels of now operating (4.5th generation) and of appearing new (5.5th generation) of the technological equipment can be obtained from the given Table.

LGD Company in April 2011 commissioned a line of the 4.5th generation to manufacture dis-plays for smartphones [12]. Productivity of the line is 12,000 substrates per month.

The trend of revenue return growth due to sales of OLED displays is shown in Fig. 3.

In 2010, OLED displays were mainly made for mobile phones. In 2010 more than 90 % of OLED displays were manufactured by SMD Company. Ac-cording to a report of Display Search Company con-cerning mobile phones market investigation for the fi rst quarter of 2011, the sales volume of OLED dis-plays for mobile phones in 2011 will reach $4 bil-lion, and in 2012 – $6.4 billion [14]. In 2015 sales volume of OLED displays with active matrix based on LTPS TFT for mobile devices will exceed the sales volume of the correspondent LC displays on the same basis. It should be noted that the work con-cerning OLED is more or less fi nanced by the gov-ernments of the correspondent countries.

At the same time, development of TVs with OLED screens is ongoing, because TVs are the core of the display market. It is supposed that OLED dis-plays will be widely used in 3D TVs as they ensure an essentially better quality of the 3D image than LC displays. SMD, LG Displays, Sony, Toshiba, Hi-tachi, Matsushita, AU Optronics and DuPont have their own programmes of development of the above

Fig. 3. Growth Dynamics of total income of OLED R displays sales (according to the version of Display Search Company being a leading researcher of the display market [13])


12

a polymeric fi lm or a metal foil as the base [21–23], because fl exibility discovers absolutely new qualities allowing to create not only curved or roll displays, but also to build OLED displays into various objects, such as clothes, furniture, instrument panels of vehi-cles, etc. In spring 2010, Sony Corporation showed a fl exible display of 4.1 inches diagonal size, which can be wound around a pencil (Fig. 7). The thickness of the display is only 80 microns. And its remark-able feature is that its active-matrix control system is based on organic thin-fi lm transistors [22].

Universal Display Corporation (UDC) (USA) has developed a fl exible sleeve display commis-

sioned by the US Department of Defense [3]. And UDC is one of the leading companies which develop OLED production technologies. Its PHОLED tech-nology is used by most of the companies which de-velop and produce OLEDs and devices on their ba-sis, both for displays and illumination. Among these are SMD, LGD, LG.Philips LCD, AUO, Konica-Mi-nolta, and Mitsubishi Electric. In this technology, phosphorescent emission organic materials are used. They are more effective for the conversion of charge carriers to photons than fl uorescent.

A research centre Flexible Display Centre (FDC) located at Arizona University, together with its part-ner has installed the Sunic System equipment of the second generation and begins manufacturing fl exible AMОLED displays with 3.8 inches diagonal size de-veloped together with UDC [24] on their base.

One more distinctive feature of the OLED tech-nology is the possibility to create translucent (semi-transparent) displays. Such displays are developed by LGD, SMD, Sony, AUO and TDK companies [25–28]. Fig. 8 shows one of them, which has a 30 % transparency [25]. In order to increase the trans-

in a globe confi guration of 6 m diameter, is exhib-ited at the Tokyo Museum of New Technologies and Innovations [19, 20]. The latter one is made of 10,352 PMОLED panels of 96×96 mm size (Fig. 6). It shows movement of clouds and other pictures of the Earth transmitted from a meteoro-logical satellite.

One of advantages of the OLED technology is the possibility to develop displays on a fl exible basis. Not only companies which manufacture OLED dis-plays, but also many other companies and research centres develop devices on a fl exible substrate using

Fig. 4. A 155-inch OLED screen of Mitsubishi Electric Company

Fig. 5. OLED screens of Mitsubishi Electric Company looking like a cylinder and a globe

Fig. 6. PMОLED panels, of which OLED screen is made as a globe


13

from monochrome devices of night vision to full-screen 3 D HD portable video players. Full-col-our microdisplays of this series are manufactured based on OLEDs with colour optical fi lters similar to those used in LC displays. The service life of the microdisplays is no less than 25, 000 h. In April 2011, eMagin presented a new microdisplay of VGA OLED-XL series with 680×520 pixel resolution and 0.44 inches diagonal size. Its power was equal to only 60 mW, i.e. twice smaller than SXGA OLED-XL series microdisplays [34]. eMagin also develops new microdisplays of 1280×1024 and 1920×1200 resolution [35].

Due to a wide interval of working ambient tem-peratures and a small power consumption, eMagin microdisplays are widely used in military equip-ment [36, 37].

Development of OLED microdisplays is also car-ried out by MicroOLED (France), Sony, Rohm (Ja-pan), Yunnan North OLiGHTEK Opto-Electronic Technology (Peoples Republic of China) and in the research centre of the Fraunhofer Institute (Germa-ny). MicroОLED and CEA-Leti developed an OLED display with an ultrahigh resolution. Its diagonal size is 0.38 inches (Fig. 9), its resolution 873×500 and its power is equal to 25 mW [38].

All of the above listed achievements are obtained based on the technology of deposition of low-mo-

parency, intensive research is performed to replace the LTPS TFT active matrix with a matrix based on more transparent transistors of indium and zinc ox-ides (ITO) [29, 30]. In May 2011, TDK announced the beginning of production of translucent passive-matrix displays of 320×240 resolution, 2.4 inches diagonal size, 40 % transparency and 150 cd/ m2 lu-minance [31]. The passive-matrix displays do not demand an active-matrix control system located on a substrate, therefore their transparency is higher.

Translucent displays are developed for laptops, MP3 players and smartphones. However, the pos-sibility of their wider application is still being in-vestigated. It is supposed that such displays will be widely used in windows for trade, exhibition, muse-um and information centres, safety systems, wind-shields of cars and optical view-fi nders of movie cameras. Samsung suggests to use translucent dis-plays in view-fi nders of photo cameras as well [28]. According to Display Bank Company forecast, sales volume of transparent displays in 2025 can reach $87.2 billion [32].

As to OLED microdisplays, they are currently manufactured exclusively by the American compa-ny eMagin [33]. They have a resolution of 800×600, 852×600 or 1280×1024 and diagonal size from 0.59 to 0.77 inches. Microdisplays are intended for per-sonal display systems of a wide application range:

Fig. 7. A fl exible OLED display of Sony Company

Fig. 8. A translucent 19-inch monitor of SMD Company Fig. 9. A microdisplay of MicroOLED Company


14

beginning of their commercial development, howev-er, the luminous effi cacy of luminaires with OLED LSs already reaches 58–60 lm/W, and in the future it will reach 150 lm/W. Unlike LEDs, OLEDs allow creating LSs which are not blinding, are uniform-ly shining, with a high quality of colour rendition. Development of light emitting diode illumination is in the list of national priorities for reasons of eco-nomic and energy effi ciency, and ecological safety for many countries, including countries of the Eu-ropean Union, Canada, People’s Republic of China, the USA and Japan. The US Department of Energy has started a long-term national research and devel-opment programme of light emitting diode illumina-tion (Multi-Year Program Plan: Solid-State Light-ing Research and Development – MYPP SSL R&D), which is intended to achieve energy saving, decrease of service costs, increase of illumination quality and intensifi cation of environmental protection. The fi rst part of the programme was adopted in 2000 and end-ed in 2008, and the second part began in March 2008 and will end in 2014. Within this programme, the Department of Energy of the USA in 2010 allocated for LED and OLED illumination research $63 mil-lion and $40.6 million respectively, for the follow-ing six directions: energy effi ciency, service life, re-liability, assemblage and quality assurance methods, infrastructure and decrease in value [44, 45].

In EU countries, a number of programs for the development OLED LSs are in effect: ОLED 100 EU [46], TOPASS [47], Fast2 Light [48], CombОLED [49], etc… Chinese, Korean, Taiwanese and Japanese companies have their own programs. And in Russia a road-map “Use of Nanotechnolo-gies in Production of Light Emission Diodes” [50] has been developed by the initiative of the Rosna-notekh State Corporation. The roadmap is mainly based on the above mentioned program MYPP SSL R&D regarding research and production of OLEDs for illumination, and it plans correspondent tasks for domestic science and industry for the near future.

OLEDs to a large extent, unlike LEDs, allow ob-taining diffusely luminous bodies of a big area and of any confi guration, including bodies with fl exible bases and translucent; radiation of any set dominant wavelength or of a wide spectrum, including white colour of any shades (using correspondent organic electro-fl uorescent materials). Besides, it is pos-sible in essence to manufacture OLEDs using in-expensive methods, such as printer or offset print-ing, including roll technologies (for example R2 R).

lecular organic material layers using the method of vacuum thermal evaporation. Achievements of the OLED structure formation technology based on po-lymeric material, are much lower. But correspondent (“polymeric”) OLEDs attract wide attention from re-searchers, because these OLEDs can be air-produced using liquid phase that gives a possibility to use in-expensive methods of printer or offset printing, in-cluding roll technology R2 R. Cambridge Display Technology (CTD), DuPont, Epson, General Elec-tric, Mitsubishi Chemical, Pioneer, Philips, Showa Denko, SMD, Sony, Toppan and other companies develop technologies of OLED formation using the liquid phase [39]. However, because of problems connected with the production of a necessary qual-ity ink and with providing high requirements for lu-minance and effi ciency of the OLEDs, achievements of this technology are smaller. In order to overcome these diffi culties, some companies try to combine the printer method of emission layer deposition with the method of vacuum thermal deposition of subse-quent functional layers [40]. At the SID-2011 con-ference, Sony, DuPont and Toppan together with Casio showed displays manufactured using printer methods [41–43]. Toppan and Casio companies pre-sented a full-colour display of 960×540 resolution, 7.42 inches diagonal size, and of 0.17 mm pixel size.

As it was noted above, mass production of OLED displays except for microdisplays, was monopo-lised by Asian companies SMD, LGD, Sony and AUO, which also develop TVs with OLED screens and plan large-scale investments into development of OLED displays. For this reason, all leading com-panies in the European Union and the USA, except for UDC, which develops fl exible displays, termi-nated work in this direction and switched over to a more interesting direction, from their point of view: development of OLEDs for illumination.

3. ILLUMINATION USING ORGANIC LIGHT EMISSION DIODES

Increasing energy provider costs and ecological requirements, the lack of fossil fuel reserves (oil, gas), insistently lead to the necessity of electric pow-er saving, including for illumination. The most pro-spective direction of illumination electric power sav-ing today is conversion to light emitting diode light sources (LS), both inorganic (LED LSs), and organic (OLED LSs). OLED LSs in comparison with tradi-tional LSs and even with LED LSs, are only at the


15

those based on low-molecular materials obtained us-ing deposition of organic layers onto glass substrates by means of thermal evaporation in a vacuum.

Besides an emission layer, a typical OLED struc-ture contains additional functional layers facilitating injection and transport of charge carriers from the electrodes to the emission layer. They also raise the effi ciency of the radiative recombination of electrons and holes (Fig. 10).

Because of the fuzziness of OLED emission ma-terial power areas, the spectrum of OLEDs appears to be wider than of that of LEDs. Generation of high quality colour rendered white light using one emis-sion material only, is problematic. Therefore, white light is created by mixing for example, blue, green and red colours. In this case several emission layers can be used, or the emission layer can be doped with molecules ensuring radiation of additional colours. Advantages and disadvantages of each method are considered in paper [51].

OLED production technology based on low-mo-lecular materials allows a wide change of functional layers number and of materials used for their man-ufacturing. In order to increase OLED effi ciency, extra layers are added: for example blocking layers preventing the escape of charge carriers from the emission layer, or layers lowering potential barri-ers between main functional layers. In doing so, one should optimise the thickness of the auxiliary layers to ensure passage of injected charge carriers to the emission layers with minimum loss [55, 56].

One effective type of light-emission structure is a tandem structure (Fig. 11) proposed by specialists from Kodak [57]. In this structure, vertically locat-ed emission layers of red, green and dark blue emis-sion colours are used with necessary functional lay-ers. As the dark blue emitter, a fl uorescent material is used, because correspondent phosphorescent ma-

OLEDs are well suited for back illumination of LC displays.

OLED LS developers are faced with the chal-lenge of increasing luminous effi cacy, luminosity and service life, whilst reducing costs. The develop-ers are working to address this challenge [44, 51, 52].

Increasing luminous effi cacy and luminance fi rst of all requires the development of new, more effec-tive organic electrofl uorescent materials, for optimi-sation of the light-emission structure, use of special optical coatings to facilitate escape of generated pho-tons from the light-emission structure. As to increas-ing service life, this is connected with development of more stable organic electrofl uorescent materials, fi rst of all of blue emission, and with improvement of OLED structure protection against adverse effect of water vapour and oxygen contained in the air.

Many companies, such as Dow Chemical, Du-Pont, Merck, Basf, Idemitsu Kosan, Mitsubishi Chemical, LG Chemical, Nippon Steel Chemical, etc., carry out intensive development of new, more effective materials. The main direction of these de-velopments is conversion from fl uorescent to phos-phorescent emission materials. In the latter, internal quantum effi ciency can theoretically reach 100 % instead of 25 %, as takes place in the former [4, 53]. And in this case, one of the main problems is obtaining a blue phosphorescent material with a long service life. In May 2010, UDC Company an-nounced the end of development of a new effective phosphorescent emitter (RGB1 B2) containing an ad-ditional emission layer emitting blue, suitable both for displays, and for light panels [54]. This emit-ter is 30 % more effective by luminous effi cacy and twice more durable than blue PHОLED emitter de-veloped by UDC earlier.

Today, the most studied OLED structures, which have reached the stage of industrial manufacture, are

Fig. 10. A typical OLED structure


16

University have found that deposition of a chlo-rine monoatomic layer onto the surface of the indi-um oxide layer doped with tin (ITO) allows elimi-nating hole injection layer and transport layers, as well as ensuring a high luminous effi cacy of the OLEDs [59]. They have developed a simple and safe method of such deposition and obtained OLED sam-ples of 10, 000 cd/ m2 luminance.

To increase the coefficient of light extraction from a light-emitting structure, additional external and/or internal optical layers are added [57, 59, 60] (Fig. 12).

As an additional external layer, polymeric micro-lenses [6] deposited by printer methods on the sub-strate, are widely applied. They almost double light extraction from OLED structures. In Fig. 13 poly-meric microlenses and an organic material lattice are shown respectively. They are developed by research-ers at the University of Michigan [61].

Specialists of PPG Industries and UDC suggest-ed making integrated glass substrates, which along with an anode layer contain additional internal and external layers raising light extraction from the light-emitting structure (Fig. 14) [60].

For today, the main ways increasing OLED lu-minous effi cacy are understood and the speed of the progress is clear enough (see, e.g. Fig. 15).

According to the MYPP SSL R&D program cor-rections [45] and on the assumption that luminous effi cacy of OLED panels asymptotically tends to 200 lm/W, one should expect that in 2015 it will reach 105 lm/W, and luminosity of 200 cm2 area panels will reach 10, 000 lm/ m2.

Light OLED panels of 3, 000–5, 000 cd/ m2 lu-minance are required for illumination.

But some key OLED parameters decrease at high values of luminance (Fig. 16).

terials have a smaller service life. Such a structure not only raises OLED luminous effi cacy, but also increases service life and fl exibility when choosing colour shades.

Increasing the number of functional layers leads to complications and a rise in the price of the tech-nology, therefore research is being carried out to simplify OLED structure. Researchers at Toronto

Fig. 11. A tandem OLED structure:EIL-1 – fi rst electron-injection layer; ETL-1 –fi rst elecron-transport layer; Phosphor Red –electrophosphorescent layer of red emission; Phosphor Green – electrophosphorescent layer of green emission; EBL-1 –fi rst electron-blocking layer; EBL-2 – second electron-blocking layer; HTL-1 –

fi rst hole-transport layer; HIL-1 –fi rst hole-injection layer; EIL-2 –second electron-injection layer; ETL-2 – second

electron-transport layer; Fluorescent Blue –electrofl uores-cent layer of blue emission; HTL-2 –second hole-transport layer; HIL-2 – second hole-injection layer; SRL – short-

circuit layer

Fig. 12. An OLED Structure with additional internal optical layers


17

had a record luminous effi cacy of 102 lm/W with 1, 000 cd/ m2 luminance and an Ra equal to 70 [62]. According to the accelerated test results, their serv-ice life was 8, 000 hours. In July 2010, UDC to-gether with Armstrong World Industries presented an OLED ceiling luminaire to the US Department

In addition, OLED luminous effi cacy depends on the general colour rendering index Ra: the greater the index, the lower the luminous effi cacy.

In the middle of 2008, UDC Company announced the manufacture of laboratory samples of white OLEDs using PHОLED technology. The samples

Fig. 13. Polymeric microlenses deposited by printer methods on the substrate (diameter of each lens is 5 micrometers)

Fig. 14. An integrated glass substrate

Fig. 15. Growth dynamics of luminous effi cacy ηv of OLED panels in accordance with the program MYPP SSL R&D of the Department of Energy of the USA [44]


18

luminous effi cacy of 160 lm/W, was obtained for structures of green emission, and maximum service life of 1, 000 000 hours with 1, 000 cd/ m2 initial lu-minance was reached for phosphorescent structures of red emission.

In autumn 2008, the new ОLED 100 eu project began [46]. It was due to be completed in 2011, and in its process OLED LSs of white emission are planned to be obtained with a luminous efficacy of 100 lm/W, 1,000 cd/ m2 luminance and of 100, 000 hours service life.

A number of European companies take part in the TOPAS project (terminating in 2014), the purpose of which is to develop OLED LSs with 1, 000 lm lu-minous fl ux, as well as translucent OLED LSs with a big area (1 m2) [47].

Beginning from the second half of 2009, Philips has been selling light panels of different colours of 30×30, 40×40 and 44×47 mm size complete with a controller and an adapter. The panels are gener-ally named as Lumiblade [68]. The rated luminance of white emission panels is 1,000 cd/ m2, and lu-minous effi cacy is 20 lm/W. To familiarise archi-tects, light designers and other potential users with the abilities of OLED LSs the ОLED design kit has been produced.

According to Philips Lighting specialists, world production of illumination devices will change con-siderably in the near future. Ecological and econom-ical factors require to transfer from the production and application of incandescent lamps to more eco-logical and energy saving LSs as soon as possible. And this in particular will be helped by OLEDs us-ing development of LSs and illumination devices

of Energy (Fig. 17). Luminous effi cacy of the lumi-naire was 58 lm/W [63]. Its development was carried out from July 2008 under a two-year contract with the Department of Energy. The cost of the contract was $1.9 million. The luminaire contains light pan-els of 15×15 cm size and built-in optical lenses. The light panels are made in accordance with PHОLED technologies of UDC. Under the contract with the Department of Energy, UDC and Moser Bauer Tech-nologies will manufacture two pilot lines for the production of light panels [64]. The fi rst line was planned to be commissioned in 2011.

At the LightFair International 2011 exhibition an American company, Acuity Brands, making lighting equipment, showed two luminaires with OLEDs for room illumination with a luminous effi cacy of 48–53 lm/W and a service life of 15, 000 hours [65]. OLED panels for these luminaires are made by LG Chem Company. The start of sales is planned for the fi rst quarter of 2012. Under a contract with the US Department of Energy, UDC together with Acuity Brands Company develop OLED panels with a con-trolled emission colour [66].

Leading European corporations and research cen-tres within OLLA incorporated project [67] have obtained, under laboratory conditions, structures of white emission with luminous effi cacy of 50.7 lm/W and service life not less than 10, 000 hours.

This project was supported by the governments of the leading European Union countries and last-ed from 2004 to 2008. Novaled (Germany) has achieved a service life of 100, 000 hours for struc-tures of white emission with luminous effi cacy of 38 lm/W and of 1, 000 cd/ m2 luminance. Maximum

Fig. 16. Characteristic dependences of “current luminous effi ciency” ζv and luminous effi cacy ηv of OLED panels on their luminance by a normal to the surface Lv [4]


19

Energy and Industrial Technology Development Or-ganization. In February 2010, Lumiotec began trial deliveries of light panels of 14.5×14.5 cm size bun-dled with a controller and adapter, which allow con-necting the panels directly to the mains and control-ling the luminance [74]. The panels have a luminous effi cacy of 10.5–10.7 lm/W, maximum luminance of 4, 000 cd/ m2 and service life up to 50, 000 hours with luminance of 1,000 cd/ m2 [75].

Light OLED panels of Lumiotec available for sale are more luminous and larger than those by Os-ram and Philips (Fig. 18), but they are several times more expensive. In spring 2011, Philips began man-ufacturing of 70×70 mm size panels with 45 lm/W luminous effi cacy, 1, 000 cd/ m2 luminance, 2, 800 K chromatic temperature and 10, 000 hours service life. These panels were developed by Konica Mi-nolta Company based on the PHОLED technology of UDC [76]. For orders of more than 100 pieces, panel price is €120.

AUO Company at the FPD 2010 exhibition, pre-sented light OLED panels of 333×314 mm size, 1, 500 cd/ m2 luminance and luminous effi cacy 50 lm/W [77].

Toshiba made portable luminaires with OLEDs to help victims of the earthquake and tsunami in Ja-pan [78]. The luminaires of 146×100×18.5 mm size and luminous fl ux of 53 lm, can work using two al-kaline ААА batteries not less than two hours at 100 % luminous fl ux and not less than ten hours at 30 %.

Kaneka Company announced its readiness to sup-ply OLED panels of heat-white, red, orange, green and blue emission with a changed luminance (maxi-mum to 5, 000 cd/ m2), luminous effi cacy 20 lm/W and service life of about 10, 000 hours. And by 2014, it is planned to increase luminous effi cacy to 60 lm/W and service life to 25, 000 hours. The working

of fundamentally new structures with characteristics, which previously were not achievable. Accordingly, Philips is expanding its production of OLED panels in Aachen (Germany), and as allocated €40 million for this [69].

Beginning from November 2009, Osram Com-pany has been selling light panels “Orbeos” [70], which have round lighting surfaces of 88 mm di-ameter, 25 lm/W luminous effi cacy, maximum lu-minance of 1, 000 cd/ m2 and a service life of about 5, 000 hours. At present Osram (Germany) [71] is constructing a production line to manufacture light OLED panels in Regensburg. Commissioning of the line was planned for the middle of 2011. Ini-tial number of the workers was to be 200. The com-pany intends to invest €50 million in the construc-tion of this line and in research concerning OLED. The main directions of this research are the devel-opment and adjustment of technological processes at the production line, increase of OLED luminance, luminous effi cacy and service life. In June 2011, Os-ram announced its records: white OLED panels both of 87 lm/W luminous effi cacy, 1, 000 cd/ m2 lumi-nance, chromatic temperature of about 4 000 K, and with 75 lm/W luminous effi cacy and 5, 000 cd/ m2 luminance. In the near future, it is planned to begin manufacture of such panels at a pilot line [72].

Novaled Company, one of the European leaders in OLED production technology, in July 2011 an-nounced white OLEDs with luminous effi cacy of 60 lm/W, luminance of 1,000 cd/ m2 and proposed serv-ice life of 100, 000 hours [73].

In 2007, Japanese companies Mitsubishi Elec-tric Works Ltd., Indemitsu Kozan Co. and Tazmo Co. established Lumiotec joint company to develop OLEDs for illumination and to commercialise them (approximately in 2012). The work of the Enterprise was partly fi nanced by a governmental agency New

Fig. 17. A ceiling luminaire with OLEDs (UDC Company) Fig. 18. Light OLED panels of Osram, Philips and Lumiotec companies


20

In May 2011, Mitsubishi Chemical and Pioneer companies reported that they had manufactured a white OLED by deposition of emission layer from a solution [82]. Luminous effi cacy of the OLED reached 52 lm/W, and service life was 20, 000 hours with luminance 1, 000 cd/ m2. The companies plan to commercialise printer methods of manufacturing light emission diodes by 2014.

Verbatim, a daughter company of Mitsubishi Ka-gaku Media, presented at the Milan lighting exhibi-tion in April 2011, the fi rst white OLED panels with a controlled chromatic temperature from 2 700 to 6 500 K, and full-colour (RGB) OLED panels.

The panels have 140×140 mm size, luminous ef-fi cacy 28 lm/W, Ra = 80 (typical), and service life 8, 000 hours [83].

UDC Company has made a considerable progress in the development of its own OLED manufactur-ing technology P2 ОLED using solutions. For green OLEDs, current luminous effi ciency of 68 cd/A and service life of 175, 000 hours are reached, and for red OLEDs accordingly – 18 cd/A and 125, 000 hours [84].

Within the Fast2 Light European project, which has united the efforts of nine companies, two re-search centres and of two universities, OLED LS development with a fl exible foil base is carried out. The objective of the project is the development of in-expensive highly effective processes of manufactur-ing “polymeric” LSs of a big area intended for in-tellectual illumination systems [49]. Fig. 20 shows a line for research and developments using a roll technology. The line is located in the Holst Centre research centre (Netherlands) [85], which takes part in the project mentioned above.

Osram together with CEA-Leti within the CombОLED project [86], along with the TOPAS

area is now equal to 10, 000 m2, and by 2014 will increase tenfold [79].

General Electric (GE) develops (from 2003) the R2 R OLED production technology for illumina-tion (Fig. 19), which will allow lowering their cost essentially [80].

In July 2010, GE (USA) and Konica Minolta (Japan) announced a joint essential breakthrough in OLED production technology due to deposi-tion of layers from a solution onto a fl exible sub-strate [81]. They obtained OLEDs luminous effi cacy 56 lm/W. This achievement means that high-effec-tive white OLEDs for illumination can be manufac-tured by a rather inexpensive method similar to those used in the printing industry. According to a fi g. of speech of GE specialists, “OLED light can be visible in the end of a tunnel”.

Fig. 19. Luminaires with OLED (General Electric Company) made using the R2R technology

Fig. 20. A line for researches and developments using the R2R technology

Fig. 21. A translucent luminaire with OLEDs (Osram Company)


21

For the purposes of a wider acquaintance of de-signers, architects and constructors with the capa-bilities of OLED application for illumination, OLED panel pilot samples deliveries are carried out. Lumi-nous effi cacy, luminance and service life of OLED LSs have reached a level sufficient for the start of their practical implementation. Infrastructure for the industrial production of “glass” light OLED pan-els instead of LC displays, practically already exists. According to foreign experts, in the next ten years, an intensive use of OLEDs will be started for illu-mination of inhabited, offi ce and production rooms.

Emergence of OLED promises to revolutionise both the world of displaying information, and the il-lumination world: from video walls to price labels in supermarkets and from lighting walls, ceilings and windows to light design of clothes, footwear, etc.

REFERENCES 1. C. W. Tang, S. A. Van Slyke, J. Applied Physics

Letters, 1987, v. 51, p. 913–915. 2. J. H. Burroughes, D. D. C. Bradley, A. R. Brown,

R. N. Marks, K. Mackay, R. H. Friend, P.L. Burns, A. B. Holmes, Nature, 1990, v. 347, p. 539.

3. K. Mullen, U. Scherf, Organic Light Emission Devices, 2006, Wiley-VCH Verlag GmbH&Co. KGaA.

project, develop transparent light OLED panels and luminaires with them (Fig. 21). Emergence of such products means the creation of lighting windows, translucent lighting walls and mirrors that will fun-damentally change light design of rooms [87].

Based on OLED, light pointers and information signs, for example, can be made. Their important ad-vantage is a low power consumption, which allows to make them autonomous.

Besides signs with OLEDs, Kenwood Company (Japan) has developed ultrathin audio speakers unit-ed with an OLED panel (Fig. 22) [88]. These speak-ers consuming energy by 80 % less than their ana-logues, are a side effect of development of evacua-tion signal boards united with a loudspeaker, which Kenwood developed for schools and offi ces.

From year to year, more and more companies become involved in the development of OLED LSs and of their associated lighting equipment. Fig. 23 published by Display Search Company in 2009 [89], shows presumed periods of development and prep-aration by leading manufacturers of commercial manufacture of luminaires with OLEDs. And in ac-cordance with a corrected forecast of the same com-pany [90], sales volume of the OLED panels will be $18 billion in 2018.

Fig. 22. Audio loudspeakers with OLED panels and evacuation signs of Kenwood Company

Fig. 23. Plans of leading companies concerning commercial manufacture of luminaires with OLEDs


22

26. http://www.OLED-info.com/transparent-OLEDs 27. h t t p : / / w w w. t r e n d h u n t e r . c o m / t r e n d s /

transparent-OLED-displays. 28. http://www.OLED-info.com/samsung-use-trans-

parent-amOLEDs-optical-camera-viewfi nders. 29. J.L. Song, J.S. Park, Y.W.Heo, J.H. Lee, J.J. Kim,

G.M. Kim, B.D. Choi, Appl. Physics Lett., 2007, v. 90, p. 022106–1-022106–3.

30. http://www.OLED-display.net/sony-117-inch-tos-tft-amOLED-panel-with-improved-lifetime.

31. h t t p : / / w w w . O L E D - d i s p l a y . n e t /tdk-start-mass-production-of-transparent-pm-OLEDs.

32. http://www.OLED-info.com/displaybank-trans-parent-displays-reach-872-billion-revenue-2025.

33. http://www.emagin.com/OLED-microdisplays/ 34. http://www.OLED-display.net/emagin-introduc-

es-lowest-power-OLED-xltm-microdisplay. 35. h t t p : / / w w w . O L E D - i n f o . c o m /

emagin-hired-house-design-team-OLED-microdisplays. 36. http:/ /www3.rockwellcoll ins.com/news/

page11333.html. 37. http://www.liteye.com/. 38. http://www.hi-news.ru/tehnologii/OLED-

mikrodisplej-s-rekordno-vysokim-razresheniem.html#more-17291.

39. h t t p : / / w w w . O L E D - d i s p l a y . n e t /OLED-inkjet-printing.

40. http://www.OLED-display.net/showa-denko-im-proves-OLED-device-w-40-light-withrawal-profi ciency.

41. http://www.OLED-display.net/events/sid-2011. 42. h t t p : / / w w w . O L E D - d i s p l a y . n e t /

dupont-showcase-printed-amOLED-displays. 43. h t t p : / / w w w . O L E D - d i s p l a y . n e t /

sony-showcase-3-inch-printed-hybrid-OLED. 44. Multi-Year Program Plan FY 09-FY 14: Sol-

id-State Lighting Research and Development U.S. МИНЭНЕРГОСША, March 2008.

45. http://apps1.eere.energy.gov/buildings/publica-tions/pdfs/ssl/ssl_mypp2010_web.pdf.

46. h t tp : / /www.OLED-info .com/OLED100 eu-wins-eus-ict-best-energy-effi ciency-project.

47. http://www.OLED-info.com/topas-project. 48. http://www.fast2 light.org/consortium.html. 49. http://www.OLED-info.com/combOLED. 50. h t tp : / /www.rusnano.com/Sect ion .aspx/

Show/27387. 51. М. Кляйн, К. Хойзер, Светотехника, 2009,

3, с. 15–24. 52. Й. Амелунг, К. Кирхгоф, М. Эритт, 2011, 1,

с. 16–20.

4. L. Zhigang, M. Hong, Organic Light Emission Materials and Devices, 2006, Taylor&Francis Group.

5. A. Ghosh, S. Van Slyke, Information Display, 2006, v. 22, 2, p. 26–31.

6. R.Heinz, Svetotekhnika, 2006, 5, p. 43–47 7. D.Z. Peng, H.L. Hsu, R. Nishikawa, Information

Display, 2007, v. 23, 2, p. 12–18. 8. http://www.OLED-display.net/ OLED -tv-15

el9500-vs-lcd-tft-technology. 9. http://www.OLED-display.net/samsung-mobile-

displays-fi rst-55 g-production-line-is-ready-to-produce-am OLED -panels; http://www. OLED -info.com/pro-duction-am OLED -panels-samsung-photo.

10. http://www.digitimes.com/news/a20110215 PD214.html.

11. h t t p : / / w w w . O L E D - d i s p l a y .n e t / s a m s u n g - m o b i l e - d i s p l a y - p l a n s - a - 8 g-pilot-line-for-55-inch-OLED-television.

12. http://www.OLED-display.net/lg-display-small-sized-amOLED-and-OLED-television-plan.

13. http://www.OLED-info.com/displaysearch-OLED-revenue-grow-7 b-2016-we-might-see-new-OLED-tvs-sony-samsung-and-lg-2009.

14. http://www.OLED-info.com/displaysearch-amOLED-mobile-phone-display-market-reach-4-billion-2011–64-billion-2012.

15. http://www.OLED-info.com/OLEDnet-pub-lished-their-OLED-market-forecast-2010–2016.

16. http://www.OLED-display.net/samsung-and-lg-display-to-invest-17-billion-dollars-in-am-OLED-un-til-2015.

17. http://www.OLED-display.net/ mitsubishi-show-155-inch-OLED-tv-consists-of-720–10-centimeter-square-panels.

18. h t t p : / / b l u m b e r b e l k . c o m / 2 0 1 0 / 0 9 /OLED_mitsubish/.

19. http:// mashable.com/2011/02/04/ mitsubishi-shows-off-enormous-360-degree-screen-pics-video/.

20. http://www.OLED-display.net/ mitsubishi-6-me-ter-OLED-globe-at-science-museum-in-tokyo.

21. http://www.OLED-info.com/flexible-OLED; http://www.OLED-display.net/fl exible-OLED.

22. h t t p : / / w w w . g i z m a g . c o m /sony-rollable-otft-driven-OLED-display/15226/.

23. http://www.OLED-display.net/udc-delivers-wrist-mounted-fl exible-ph-OLED-display-prototypes-to-us-army.

24. http://www.OLED-info.com/flexible-display-center-installed-new-gen-ii-OLED-process-tool.

25. h t t p : / / w w w . O L E D - i n f o . c o m /samsung-demonstrates-19-transparent-amOLED.


23

74. http://www.OLED-info.com/lumiotec-shows-new-OLED-panels-plans-start-mass-production-janu-ary-2010.

75. http://www.OLED-info.com/files/Lumiotec-OLED-lighting-data-sheet-april-2011.pdf com/black-body-shows-interesting-new-OLED-lighting-designs.

76. http://www.OLED-info.com/philips-lumiblade-plus-worlds-most-effi cient-OLED-lighting-panel.

77. h t t p : / / w w w . O L E D - i n f o . c o m /auo-OLED-lighting-panels-photo-and-specs.

78. h t t p : / / w w w . O L E D - i n f o . c o m /toshiba-portable-OLED-lamp-2011.

79. http://www.OLED-info.com/kaneka-start-shipping-OLED-lighting-panels-march-unveils-OLED-stra tegy.

80. http://www.OLED-info.com/ge-roll-roll-OLED-lighting-panel-printing-real-reached-over-90-yield.

81. http://www.OLED-info.com/ge-and-km-effi-ciency-breakthrough-56-lmw-roll-roll-printable-white-OLEDs.

82. http://www.OLED-info.com/ mitsubishi-and-pio-neer-fabricated-white-emissive-layer-printed-OLED-52 lmw-effi ciency.

83. http://www.OLED-info.com/verbatim-shows-their-velve-color-tunable-OLED-lighting.

84. http://www.OLED-info.com/universal-display-re-ports-advances-its-solution-processable-OLED-materials

85. http://www.holstcentre.com/. 86. http://www.OLED-info.com/osram-shows-new-

totally-transparent-OLED-lighting-panels. 87. P.Y. Ngai, J.M. Fisher, DEPARTMENT OF EN-

ERGY OF THE USA Solid State Lighting R&D Work-shop, February 2011.

88. http://www.slashgear.com/kenwood-OLED-lamps-with-integrated-speakers-2741949/.

89. http://www.OLED-info.com/OLED-lighting-take-2011-reach-6 b-revenue-2018.

90. http://www.OLED-info.com/nanomarkets-OLED-lighting-market-will-be-97-billion-2016.

53. M.A. Baldo et.al., Nature, 1998, v. 395, p. 151–154.

54. http://www.OLED-info.com/udc-announces-new-all-phosphorescent-OLED-lighting-technology.

55. G. Schwartz, S. Reineke, T.K. Rosenow, K. Walz-er, K. Leo, Advanced Func. Materials, 2009, v. 19, p. 1–15.

56. S.L. Lai, M.Y. Chan, M.K. Fung, C.S. Lee, S.T. Lee, J. Applied Physics, 2007, v. 101, 014509.

57. Tyan Yuan-Sheng, DEPARTMENT OF ENER-GY OF THE USA Solid State Lighting R&D Workshop, February 2009.

58. h t t p : / / w w w . O L E D - i n f o . c o m /chlorine-can-lead-effi cient-and-simple-OLED-designs.

59. Tyan Yuan-Sheng, DEPARTMENT OF ENERGY OF THE USA Solid State Lighting Manufacturing Work-shop, April 2010.

60. Abhinav Bhandari, Harry Buhay, Mike Weaver et.al., DEPARTMENT OF ENERGY OF THE USA Solid State Lighting R&D Workshop, April 2011.

61. http://www.technologyreview.com/Energy/21116/ 62. http://www.OLED-display.net/universal-dis-

play-corporation-s-white-OLED-technology-exceeds-100-lm-w.

63. http://www.OLED-display.net/armstrong-and-udc-show-high-effi cient-white-OLED-lighting-ceiling-system.

64. http://www.OLED-info.com/OLED-lighting-pro-duction-line-built-moser-baer-and-udc-will-cost-20-mil-lion

65. http://www.OLED-info.com/acuity-brands-de-buts-two-OLED-luminaries-planned-2012-using-lg-chem-panels.

66. http://www.OLED-info.com/udc-and-acuity-br-ands-won-2-million-МинэнергоСША-grant-develop-color-tunable-OLED-lighting-system.

67. h t t p : / / w w w. O L E D - i n f o . c o m / l i f e t i m e /the_olla_project_delivers_its_fi nal_milestone.

68. h t t p : / / w w w . O L E D - i n f o . c o m /philips-lumiblade-OLED-light-fi rst-looks.

69. h t t p : / / w w w . O L E D - d i s p l a y . n e t /philips-expand-OLED-lighting-capacity-in-aachen.

70. http://www.osram.com/osram_com/LED/OLED_Lighting/ORBEOS_Products/index.html.

71. http://www.OLED-info.com/osram-building-70-million-OLED-lighting-pilot-production-line.

72. http://www.OLED-info.com/osram-developed-87-lmw-white-OLED-panel-production-soon.

73. http://www.OLED-info.com/novaled-develops-new-100000-hours-60-lmw-white-white-pin-OLED-structure.

Nikolai N. Usov, Dr. of tech. sciences, professor, graduated from the MPETI in 1965. At present, he is the chief constructor of Cyclone Central Scientifi c-and-Research Institute Open Society

24

Overview of the Main Issues and their Effects

The term ‘unwanted light’ is an umbrella term covering any lighting that obtrudes beyond the area of the task it is intended to light. It includes a number of other types of lighting problem:

• Light Pollution “Any form of artifi cial light which shines outside

the area it is intended to illuminate, including light that is directed above the horizontal into the night sky creating sky glow (which blocks out the night time stars) or which creates a danger by glare.” (UK Govt. defi nition)

• Light Nuisance “Artifi cial light emitted from premises so as to be

prejudicial to health or a nuisance”, (Specifi c UK le-gal defi nition in the Clean Neighbourhoods & Envi-ronment Act, 2005). More generally, light nuisance is any kind of intrusive light causing personal dis-comfort or domestic disruption. There can be con-siderable overlap between light pollution and light nuisance, but they are not the same thing.

ABSTRACT

1Unwanted (or obtrusive) light, covering such is-sues as light pollution and light nuisance, is a grow-ing international problem. It is both damaging to the environment and wasteful of energy. For example, one estimate puts the energy wastage due to over-il-lumination in the USA alone at approx. 2 m. barrels of oil per day. Various countries are trying to combat these phenomena in a range of ways.

This paper will summarise – and offer a criti-cal assessment – of the various legal, voluntary and technical means that local and national governments, in North America, Europe, Asia and Australasia, are applying to these problems. The paper will conclude with a summary of ‘best practice’ measures, which could be applied more widely.

Keywords: obtrusive light, light pollution, light nuisance, sky glow

* This paper was fi rst presented at the PLDC Conference, Madrid, October 20, 2011

TACKLING UNWANTED LIGHT: AN INTERNATIONAL PERSPECTIVE *

Carl Gardner

Lighting Journal, UK E-mail: [email protected]


Fig.1. Night-time view of the world, showing the worst offending areas in terms of light pollution


25

tres – but also increasingly more widely. The stars are part of our natural legacy and a key part of our cultural heritage and these views should be protect-ed. At the same time, unwanted, intrusive light can also affect people in their homes, by degrading their local environment, spoiling their enjoyment of their property and disrupting sleep. As well as affecting humans, a wide range of fl ora and fauna can have their breeding and feeding habits disrupted by intru-sive lighting – most notably bats, moths, birds, some amphibians, turtles and other sea life.

The Main Causes of Unwanted Light

Unless we understand the underlying reasons for these phenomena, we can never devise adequate and effective measures for combating them. There are three main types of causality involved, which inter-act to produce the problem:

• The Social-Contextual Reasons• The technical causes • The scientifi c causes

• Spill Light This is a more informal term, referring to light

that obtrudes beyond the area it was intended to light into surrounding areas or on to surrounding properties.

• Light Trespass This is a commonly and erroneously used term,

which has no technical or legal defi nition and is un-tested in law. Human agents can trespass, non-sen-tient entities without intention or will, such as light, cannot. DO NOT USE.

• Sky Glow This refers to the visible brightening of the night-

time sky, due to the diffuse scattering of upwards ar-tifi cial light by particles of dust, pollution and water vapour in the atmosphere

Main Effects of Light Pollution & Sky Glow

One principal effect of light pollution and result-ing Sky Glow is to block out the view of the stars in the night-time sky – particularly from urban cen-

Figs. 2, 3 and 4. Examples of poorly designed, polluting lighting schemes

Fig.5. Diagram of types of unwanted light (from the ILP’s ‘Guidance Notes for the Reduction of Unwanted Light’, 2011)


26

In the longer term, aided and abetted by the glo-bal energy crisis and the need to reduce CO2 emis-sions, these assumptions will need to be challenged. But this cultural battle, in which those of with a pro-found respect for both light and dark should partici-pate, is (sadly) beyond the scope of this paper.

The Technical Causes

The fundamental reason for unwanted light of all types is poor or non-existent design by non-specialist personnel – sadly every electrician thinks he/she is a lighting designer these days. Our technical capacity to light has outstripped our educational infrastruc-ture and the broader knowledge base about lighting. This is exacerbated by the design-build and contrac-

The social-contextual reasons

It is truism to say that artifi cial lighting is a symp-tom of modernity – and a genuine symbol of affl u-ence and civilisation. Darkness has only a negative value in our cultures and has come to denote danger, the ‘unknown’ and primitive under-development. People see lighting as bestowing after-dark freedom to pursue leisure and commerce. They also believe that lighting can offer a degree of security and safe-ty from crime and accidents… and, of course, they are right. At the same time, the availability of more effective, less expensive means of creating artifi cial light is increasing. There is a broad cultural assump-tion in our societies that ‘the more light the better’ for a range of activities.

Figs. 6, 7. The night-time sky as we should see it… and the kind of obscuring sky glow many of us experience

Fig.8. Floodlights mounted above the horizontal, like these, are a major offender in projecting light into the sky and creating ‘sky glow’( Photo courtesy Carl Gardner)


27

up has shorter paths though the scattering layers. So it is low angled light, just above the horizontal plane, that causes the most sky glow away from the source.

The situation is even more complicated by the form of scattering created by air-borne water and pollution molecules – this is called ‘Mie scattering’. Water and pollution particles are relatively large – much larger than air molecules. The larger the par-ticle, the greater quantity of light is scattered in a forward direction, in the direction of the light emis-sion (i.e. away from the source). The effect of this is that urban light spill just above the horizontal, coupled with aerosols in the atmosphere, will create more Sky Glow outside the immediate city area – i.e. in the surrounding countryside (up to 20 km).

Indirect Refl ected Light

The contribution of indirect refl ected light must not be forgotten, either. All surfaces, except for mat-te black, have a refl ectance factor (between 1 and 100) – and refl ect a given percentage of incident light back off them, usually in a random and diffuse manner. This applies to all exterior surfaces too – roads, football pitches, car-parks, tennis courts and so on. So regardless of the skill and precision of the lighting designer/engineer, a proportion of upwards light, contributing to Sky Glow, is inevitable. We can’t do much about it (save for painting everything matte black – not a popular policy) but we should always recognise it. Remember, there is a massive amount of light in our environment – and even a small percentage of a huge amount… is, well, a lot. In fact estimates put the exterior indirect upwards light factor at around 10 % of all upwards light.

tor-led construction culture in many parts of Europe, N. America and Asia, where relatively uninformed personnel, untrained in lighting, are given the power to specify huge quantities of exterior lighting equip-ment. The inevitable result is the purchase of cheap, poorly made equipment with inadequate optical sys-tems – light fi ttings with the wrongly specifi ed watt-age (too high) and beam angle (too wide) for the task at hand.

There is a general failure to specify beam con-trol accessories, to trim the light output of the fi tting. Light fi ttings are inaccurately aimed at the object/ area being lit – for example, fl oodlights mounted above the horizontal plane. Retrospective distur-bance of light fi tting due to wind, animals and van-dals can also be an issue, but inadequate mainte-nance means that even well-designed schemes can become light pollution hazards over time.

In the public lighting sector (road and street light-ing) chronic under-investment has also led to the persistence of archaic, technologically inferior lan-terns and light sources – e.g. in the UK there are still about 3.5 m low pressure sodium (SOX) street lights on the roads, some of which may be 30+ years old.

The Scientifi c Causes

Once stray light has escaped from the light fi t-ting above the horizontal, it can interact with the at-mosphere to create Sky Glow. This is caused by the scattering of upward light by aerosols, mostly water vapour, but also dust and pollution. The longer the light path through the lowest part of the atmosphere, the more aerosol molecules the light interacts with, so the greater the scattering. Light that goes straight

Fig.9. Diagram showing of how light projected at low angles above the horizontal can create lighter scattering than light going vertically upwards


28

• Japan – Tokyo + Bisei • Shanghai, China • Sydney, Australia • New York, USA • Los Angeles, USA • Tucson, Arizona • + Regulations for ‘Dark Sky’ parks

1. United Kingdom (exc. Scotland)

In many ways the UK has some quite progressive legal policies on lighting – but they often fail in their implementation. There are three distinct approaches to lighting in UK law:

• Planning legislation covering lighting for new installations for commercial, retail, large residential and sports facilities

• Specifi c light nuisance legislation covering ex-isting lighting schemes

• Separate legislation covering the brightness of signage lighting

Zoning

The CIE/ILP has defined four environmental zones (box below) where different quantitative lim-its on lighting should apply – see Table 1. These are applicable to all three areas of UK lighting legisla-tion above. But they are not known or referred to suffi ciently by local authority planners or other of-fi cers – and not mandated in law.

To sum up, changing our cultural assumptions about the necessity of more and more artifi cial light-ing will take time. Understanding the scientifi c caus-es of Sky Glow helps to focus our efforts on the main culprit creating light pollution – light emitted just above the horizontal. Light creating social nui-sance, however, can be caused by light emitted well below the horizontal – but is generally much more local in effect. Addressing the underlying technical causes of unwanted light of all kinds will require a multi-stranded strategy…

Global Solutions

So let’s have a look at how different countries are addressing the problems. What follows is by no means exhaustive – for example, Chile, which has had some exemplary legislation for many years, is not included in this survey. Also at least nine states in the USA have some form of regulation/legisla-tion – as have at least seven prefectures in Japan. And La Palma in the Canaries and Catalonia were the fi rst places to enact lighting regulation close to astronomical observatories, but in this study we look at a similar instance from Japan.

Cities and Countries Surveyed

• United Kingdom • Slovenia • Czech Republic, Lombardy, Italy + other Ital-

ian states


29

• Post-installation reviews of schemes, particu-larly the lighting elements (by night) are rare;

• Planning conditions can’t be applied retro-spectively, so there is no jurisdiction over existing schemes.

‘Light as a Social Nuisance’ Legislation

The second piece of UK legislation, passed in 2005, is an amendment to 19 th c. ‘statutory nui-sance’ legislation to include lighting. Light as nui-sance is defined as ‘artificial light emitted from premises so as to be prejudicial to health or a nui-sance’. Local authorities must investigate complaints of poor lighting and can advise on or demand chang-es – or, if necessary, serve an abatement notice. If not complied with, offenders can be taken to court. Interestingly, too, even if the lighting has been ap-proved by Planners, it can still be subject to action under the nuisance Act – so environmental health offi cers, whose job it is to apply the legislation, may end up dealing with the after-effects of poor plan-ning decisions.

Downsides of UK Light Nuisance Legislation

• Only applies to lighting shining into properties and has no application to upwards light pollution that is not causing a nuisance.

• Although the CIE/ILP zoning system is rec-ommended by lighting bodies it is not part of the legislation.

• ‘The prejudicial to health’ part of the legisla-tion is almost impossible to prove.

• Local authority offi cers get little or no train-ing in light and lighting, so fi nd complaints diffi cult to assess.

Planning Legislation (since 1997)

All lighting proposed as part of new commercial, retail, large residential and sports facilities is subject to UK planning permission and lighting plans/ speci-fi cations must be submitted as part of the planning application. Very specifi c conditions for lighting can be set out by planners in Planning Guidelines for specifi c areas or projects – including measures to avoid light pollution and light nuisance – and these must be adhered to. Conditions on lighting might require the achievement of certain lighting levels, styles of lantern (e.g. full cut-off road lights), cur-few switch-off times, beam angles, control devices, mounting heights etc. One important condition that can be laid down in planning conditions is a future maintenance programme that can be subject to on-site checks. Non-conformity to planning conditions can lead to the scheme having to be taken out and re-installed.

Downsides to Planning System

The UK system has many advantages, when well applied, but:

• For informed input, it relies on the participa-tion of lighting designers, which is usually only the case on larger projects;

• Many Planners do not understand lighting is-sues and don’t know what to look for or ask for in planning applications;

• Planners are often more interested in the ap-pearance of lighting equipment by day, than the night-time appearance;

• As a result, poor, potentially polluting light-ing can slip by;

Fig.10. Lighting of a docks area in the UK. Docks and ports are exempted from the Act, even though they can be some of the worst offenders in light pollution terms (Photo courtesy Bob Mizon)


30

nage lighting. Since that time new digital LED and LCD signage screens have arrived on our streets, which are many times brighter than conventional signage. According to research by Kasper Hammer Nielsen published in Lighting Journal (June 2008, p. 23) the peak luminance of such screens in Times Square, New York (and new, brighter systems have been developed since then) was 8657 cd/ m2 (8.5 times brighter than the recommended UK maximum) and the average luminance of all the screens was 2632 cd/ m2 (2.6 times brighter than the recommend-ed maxima). Clearly the legally recommended lim-its have been outpaced by new technologies and we need to re-think ways of controlling these systems.

2. Slovenia

Slovenia probably has the most stringent light-ing legislation in the world (since 2007). Effective-ly, for most installations, there is a ban on all light emissions above the horizontal for luminaires over 20 W (there are some exemptions for sensitive state facilities). There is a complete ban on sky-beamers (sky-trackers) and a ban on direct lighting of facades

• There is a list of types of installations which are exempt from the Act, including transport fa-cilities, port facilities, prisons, military facilities and street lighting. These can be some of the worst offenders.

Legislation on the Lighting of Signage

The basis of ‘deemed consent’ for new signage is that maximum brightness levels are laid down, in relation to size of sign – and the type of area in-volved (i.e. the Environmental Zone). In urban areas, for signs less than 10 m2, this is a maximum of 1000 cd/ m2 – for signs more than 10 m2, a maximum of 600 cd/ m2 – much lower levels apply to rural ar-eas and small towns in Environmental Zones 1 and 2. Sign lighting also has to be designed so that light sources are not directly visible from usual viewing angles (e.g. by road users or residents).

Downsides to Signage Lighting Legislation

The main downside is that these standards were laid down over a decade ago for conventional sig-

Fig.11. Table of recommended signage luminance from the UK legislation

Figs.12,13. Conventional backlit signage and one of the new LED display screens – the new technology is many times brighter than the old signage (Photo 13 courtesy Carl Gardner)


31

only from above and luminance limits for signage range from 300 cd/ m2 maximum for signs up to 5 m2 and 500 cd/ m2 for signs 30+m2 [curiously, this is just the opposite of UK legislation] with a 23.00 curfew. Other lit surfaces should not to exceed 10 lux or 1 cd/ m2. Luminaires should have the capabil-ity of dimming by at least 30 % (unless they emit less than 1500 lumens). There is a curfew for all archi-tectural and display schemes from 23.00–5.00 and a ban on sky-beamers.

There is a mandatory labelling system for ap-proved luminaires which must have the ‘no upward emission’ symbol and include instructions for correct installation. Lombardy has extra clause specifying

of inhabited premises, with vertical illuminance lim-its on windows from incident light off other illu-minated surfaces around the building. There are also strict power allocations for public lighting – e.g. a maximum of 50 kWh per capita per annum for all road lighting and maximum total of 180 W for all exterior installations for schools and similar institutions.

For architectural lighting, there is a maximum luminance limit for all façade and monument light-ing of 1 cd/ m2 – and there are W/ m2 limits on ad-vertising signage. These range from 17 W/ m2 up to 80 W/ m2 depending on the size of the sign. Signage lighting has strict curfew time and all applied light-ing must be downwards from the top of the sign. Various clauses in this legislation come into practice between 2008 and 2017 – with fi nes for violations of legislation up to 12,000 Euros.

While there are no intrinsic downsides to the leg-islation, in terms of preventing unwanted light (ex-cept perhaps for the W/ m2 method of assessing sig-nage lighting), we must recognise that Slovenia is a predominantly rural, mountainous country and most of its terrain would probably fall into Environmental Zones 0 and 1. Most of these recommendations are too stringent for practical application in an industr-ialised country with greater population density.

3. Czech Republic, Lombardy, Italy (+ Marche several other Italian states)

These states have broadly similar policies on lighting, developed around 2000–2002. All lumi-naires must emit no light above the horizontal and ‘fl at glass’ road lanterns should be specifi ed, rather than dished bowls. All signage is to be illuminated

Fig. 14. Slovenia – some of the best anti-light pollution legislation in the world

Fig.15. Prague in the Czech Republic – good legislation but less successful in changing the lighting culture


32

ing in Italy (2007). It would appear that the Czech Republic has had less success in changing the light-ing culture.

4. Japan

Japan as a whole, and Tokyo in particular, is one of the light pollution ‘hot spots’ in Asia. Despite this, light pollution complaints are very low. Tokyo has a unique urban lighting culture in areas like Shinjuku and Kabukicho, with tall, multi-coloured dynami-cally lit facades and signage. However, all govern-ment light pollution guidelines are purely voluntary.

These include the following measures:• Curfew times on external lighting • Limitations on external lighting of residential

properties • Limitations on glare effects for road users • Limitations on Upward Light Ratio (ULR) The obvious downside is that advisory, volun-

tary guidelines have no teeth or policing and seem to have done little to prevent Japan’s worst lighting excesses.

However, some towns have been forced to fi ght this excessive lighting environment. Bisei, north-east of Fukuyama, in central Japan, has one of the country’s most important astronomical observato-ries. To protect astronomical observations, the town has passed its fi rst ordinances on light pollution back in 1989. These include:

• 22.00 curfew on external lighting;• Subsidies of up to two-thirds of the costs for

installing, remodelling or exchanging non-light pol-luting lighting equipment;

• Stipulation that the brightness of the night sky should not exceed 10 % of the natural condition;

a minimum distance between street lamps of about four times their height, designed to encourage full ‘cut off’ street lamps, and requires that all local au-thorities develop detailed lighting plans within a particular time-frame. There are penalties (fi nes) for non-compliance in all cases.

Downsides/Comment

In the case of Lombardy, home to several big players in the Italian lighting industry, lighting com-panies have driven much of the legislation, as a way of encouraging municipalities to invest in new equipment. Although 13 other Italian states have adopted some form of legislation, lighting compa-nies there are reported to have watered them down in some cases. Lombardy claims that it has the low-est per capita energy consumption for public light-

Fig.16. As this lighting map shows, Japan is one of the light pollution ‘hot spots’ of the world

Fig.17. Bisei in Japan – fi ghting back against unwanted light


33

Downsides to Shanghai’s Approach

The emission angles for road lights are far too high at 85–90 degrees from the vertical – modern fi ttings should be emitting at no more than 70 de-grees – and mounting heights only go to 6 m, so the standard doesn’t include lighting for main roads. There is also no requirement for ‘fl at glass’ road lighting luminaires.

There also doesn’t seem to be any limits on sig-nage lighting or the use of coloured dynamic light-ing – or things such as sky-beamers. Like much else in China, it is diffi cult to assess the Standard’s effec-tiveness. However, the Standard has been in place and indications are that there has not been a single light pollution case brought before the City authori-ties since that time.

6. Sydney, Australia

Sydney has installed extensive decorative light-ing around the harbour in the city centre, including Sydney Cove, Darling Harbour, the Opera House and the Harbour Bridge. There is a Sydney Exterior Lighting Strategy which sets out mandatory stand-ards and regulations on most forms of lighting. One of its express aims is ‘to prevent sky glow’. New schemes must be submitted for planning approval.

The strategy refers to the effects on residents, road-users and astronomical observations – and cov-ers illuminated signage and lighting for buildings, landscapes, car-parks etc. However, the strategy does not apply to road lighting. It does specify cur-

• Shading of the outdoor lights with no emis-sions above the horizontal level;

• All signage lighting to come from the top of signboard;

• Specifi cation of low-pressure sodium lamps for outdoor lighting is recommended;

• Where searchlights, spotlights or lasers are continuously used outdoors, it is forbidden to use appliances which project light above the horizontal;

• Shading of indoor lights by larger businesses – advised to keep light from leaking outside by using curtains and blinds.

Six other Japanese prefectures have since fol-lowed Bisei’s example.

5. Shanghai, China

In 2010 Shanghai hosted the World Expo and invested hugely in new decorative lighting installa-tions around the city. In theory, there are mandatory controls on exterior lighting schemes must comply with the Urban Environment Decorative Lighting Standard, China’s fi rst standard on external lighting. The Standard specifi es:

• 23.00 curfew for external lights;• Limitations on lighting on to residential units;• Limits on ULR to prevent sky glow;• Glare limits on road lighting, based on an as-

sumption of the greatest average luminance being 85–90 degrees from the vertical (i.e. just below the horizontal) and for different column heights up to 6 m.

Fig.18. Shanghai’s polluting lighting culture has not been controlled by the new Lighting Standard


34

7. New York, USA

New York has mandatory controls on lighting in the city, based on energy-efficiency codes for building lighting – but they don’t deal specifi cally with light nuisance. They set maximum energy con-sumption levels (lighting power density or LPD) for privately owned exterior lighting and no measure-ments for illuminance or luminance are involved. The scheme only applies to new installations – and is not retrospective. Daylight-sensing automatic switching must also be specifi ed.

At a local level, zoning controls for sections of the city can be set, governing retail lighting or illuminated signage. There is voluntary adherence only to IESNA practice on exterior lighting

few times and limitations on architectural lighting’s glare effects to road users. It specifi es ULR for light fi ttings to limit sky glow and requires them to be classifi ed according to light intensity distribution.

Downsides to Sydney’s Approach

The two big downsides to the Sydney approach are that:

• It only applies to new lighting schemes and not existing ones;

• It doesn’t cover road lighting, one of the big-gest offenders in terms of light pollution.

There are several pro-Dark Skies bodies cam-paigning in the Sydney area which still report on the extensive light pollution and sky glow around the city – particularly over rural areas.

Fig.19. Despite having an exterior lighting strategy in place, considerable light pollution is still evident over the centre of Sydney

Fig.20. New York’s ‘energy density’ approach to lighting does not control the worst excesses of unwanted light


35

The City of Los Angeles is currently undertak-ing a 40 million dollar programme to replace 27,000 conventional 70 and 100 W sodium road lights with LEDs. It will be interesting to see if this signifi cantly reduces the city’s light pollution problems.

Downsides

Similar to New York.

Other US States

The energy-based approach to lighting control has become the norm across N. America, including, for example, Ontario in Canada. Connecticut and Maine have specifi c regulations specifying that all state-funding light sources above 1800 lumens must be shielded against upward light emissions. Arizona and New Mexico have extended this principle to all outdoor lighting.

Downsides to New York’s Approach

The energy-based approach lays down no stand-ards for lighting quality – or quantitative measure-ments. It also lays down no specifi c recommenda-tions against light pollution or light nuisance. Final-ly publicly owned lighting, including street lighting, is not included.

8. Los Angeles, USA

LA is well-known as one of the most polluted – and light-polluted – cities in the world. The city has similar lighting legislation to New York, with one for two modifi cations. Again the system is based on en-ergy effi ciency limits, but it does have Zoning limits and recommendations for four different types of city area, which are something like the CIE/ILP Environ-mental zones. The legislation specifi es cut-off and shielding requirements for external luminaires and gives vague instructions for the aiming of fi ttings.

Fig. 21. Los Angeles is one of the USA’s worst offenders in terms of unwanted light

Fig.22. Renowned for its low levels of upwards lighting and high standard of legislation


36

9. Tucson, Arizona

However, some places in the USA have defi ed the norm, Fig. 22. Due to its isolated desert loca-tion, Pima, County, Arizona, which contains Tuc-son, has had lighting ordinances since 1972. A fi ve-part Zoning system is used to defi ne different types of area. Limits are based on a rather cumbersome ta-ble of installation types specifying caps of so many lumens per acre. Curfew times range from 22.00 to midnight, depending on installation type – and un-shielded fi xtures (i.e. not full cut-off) are not to ex-ceed 3000 lm. The ordinances specifi cally ban:

• Bottom mounted sign lighting;• Mercury vapour lamps and fi xtures;• Laser light sources above the horizontal;• Searchlights for advertising purposes.

Dark Sky Parks and Reserves

The International Dark Sky Association has now given ‘Dark Sky’ status to eight parks around the world – fi ve in North America and three in Europe (one in Scotland, the Galloway Forest Park, set up in 2009, and two in Hungary). The island of Sark

Figs. 23, 24. Two of the recognised ‘Dark Sky Parks’ – the Natural Bridges National Monument Park in Utah, USA (top) and Pennsylvania’s Cherry Springs Park (bottom)

Fig. 25. Cover of the Gloucester Lighting Strategy 2008 report – lighting strategies should become standard for

all larger towns and cities, in order to control future developments


37

Best Practice Guidelines: Some Modest Proposals

Following this survey, it is possible to advance some proposals based on some of the best features and characteristics of different countries and cit-ies. Some readers might ask, why not simply adopt the IES-IDA ‘Model Lighting Ordinance’ which has recently been updated? Firstly, while the docu-ment includes many good proposals (e.g. the prin-ciple of zoning) it is very much geared to the North American scene, where local councils can adopt their own by-laws. Secondly, much of the methodology is based on ‘site lumen limits’ which is not a method used in Europe and elsewhere. Thirdly, the degree of mathematical calculation involved is not some-thing that untrained local authority offi cers could un-dertake. Finally the MLO is very engineering-driven with little consideration for lighting design quality.

1. The Importance of Integrated Lighting Plans

Lighting designers have been pushing for com-prehensive lighting plans/strategies for towns and cities for years – and they have never been more needed than now. These should be drawn up for all towns and cities above a certain size and cover pro-

in the Channel Islands has also been given ‘Dark Sky Community’ status and has similar lighting re-strictions. Recently too, 80 sq. km of Exmoor Park in England were given ‘Dark Sky Reserve’ status.

The criteria for ‘Dark Sky’ Parks are very rigor-ous. The main ones are:

• Parks must be recognised protected areas and have a minimum area of 700 sq.km;

• The park must be totally surrounded by a ‘buffer’ Zone 1 (rural) area;

• An inventory of all light sources inside the park must be prepared and each one individually assessed. They are subject to ongoing monitoring;

• Any lighting fi xtures containing lamps emit-ting 1000 lm+ shall use fully shielded fi xtures emit-ting no light at or above the horizontal;

• The type of lamp (colour, effi ciency, technol-ogy) must be carefully chosen for its energy-effi -ciency and minimal impact to wildlife, star-gazing activities and nocturnal scenery;

• Motion sensing switching to be used;• The minimum amount of light to be used for

specifi c agreed areas and tasks;• Most importantly, the municipality, residents

and businesses show strong commitment to main-taining or improving ‘dark sky’ attributes;

• There are three quality levels to ‘Dark Sky’ status – Gold, Silver and Bronze.

Fig.26. Anti-glare louvres and other devices that should be fi tted to fl oodlights and spotlights as standard (Photo courtesy Thorn Lighting)


38

f) Schemes should be subject to mandatory post-installation checks (especially after dark) at speci-fi ed intervals;

g) In an effort to prevent unwanted light spill, one useful proposal would be that lighting equipment on all schemes submitted for planning approval should be equipped with anti-glare/ anti-spill devices as standard. These are only available with better light-ing systems, so it would ensure the specifi cation of better quality equipment, Fig. 26.

3. A Framework for Controlling Existing Lighting Schemes

The current UK Light Nuisance legislation is inadequate:

a) It needs to lose its property-based bias – all complaints about bad lighting should be covered by law, both personal nuisance and upwards light pollution;

b) Full cut-off road lighting to be installed within a specifi ed period;

c) Curfew limits on non-essential lighting – and the installation of motion-sensing and dimming sys-tems on road lighting as standard;

d) One controversial proposal is that we should adopt of narrow-spectrum lighting (SOX) for rural areas (Zone 0 and Zone 1), lighting companies are currently phasing out SOX lamps (do we perhaps need a nature-friendly LED replacement?);

e) Given the science of Sky Glow outlined ear-lier, more attention should be given to controlling lighting just above the horizontal, a blanket ban on all upwards lighting would limit much well-designed

posals for future development of all types of light-ing. They would include all the necessary standards and anti-light pollution proposals and regulations that are appropriate for the town/city in question. Ideally, lighting plans would sit alongside other lo-cal plans and planning norms and have the same le-gal status, Fig. 25.

2. A Framework for Controlling New Lighting Schemes

Broadly speaking, the UK planning system of-fers a good model as a starting point, but it requires a number of improvements, namely:

a) The adoption of a version of the CIE/ILP Envi-ronmental zones table and associated lighting limits (see Section) as part of Planning Guidance;

b) Mandatory requirement for the submission of photometrically accurate computer-generated ren-derings of the night-time appearance of architectural lighting schemes above a certain size;

c) For all area lighting schemes (sports facilities, car-parks etc.) designers/ manufacturers to provide vertical illuminance plots for nearby properties, as well as horizontal plots, in order to assess potential light spill and light nuisance;

d) Where no lighting specialists are involved in the initial planning proposals, they should be subject to compulsory review by lighting design specialists;

e) Schemes should be subject to mandatory post-installation maintenance programme as one of the planning conditions (exists in UK planning frame-work, but rarely insisted on);

Fig.27. Monochromatic light sources, such as low pressure sodium (SOX) could be retained for rural locations, as they are much less disruptive to wildlife breeding and feeding habits


39

h) All limits on façade and monument lighting should be expressed in luminance and illuminance – the US W/ m2 system is inadequate;

i) All lighting regulations must be mandatory – as we have seen, advisory guidelines have had little or no success.

architectural lighting, which uses burial fi ttings or ground- or wall-mounted luminaires, and there is no evidence that these minority of lighting schemes are a major contributor to sky-glow, whereas public lighting is certainly a major offender;

f) Whatever regulations are adopted there should be minimal exemptions (e.g. military or transport security lighting);

g) New lighting regulations are defi nitely need-ed for the control of LED and LCD signage, which is massively brighter than conventional predecessors;

Carl Gardner,BA MSc (Arch) FILP is director of CSG Lighting Consultancy Ltd. and Editor of the bi-monthly publication, Lighting Journal in the UK. He is one of the UK’s foremost commentators and analysts on lighting and lighting design. Alongside architectural lighting design services, he organises training courses for local authorities on preventing light nuisance and recently advised the City Council in Hong Kong on the prevention of unwanted light

Fig.28. Attractive architectural lighting schemes such as this in York, England, would be impossible if there was a blanket ban on lighting above the horizontal. Yet their polluting effect is minimal, as vertical lighting is much less polluting than

light just above the horizontal (Photo courtesy of Carl Gardner)

40

At that time intensive use of HTLGs for illumination of industrial enterprises and shops began. Illumina-tion devices with HTLGs were developed, for exam-

ABSTRACT

A design method of natural inner illumination using hollow tubular light guides conducting natu-ral light from a building’s outer shell to the ground fl oor and basement rooms is considered. An evalu-ation of their power effi ciency is given. A possibil-ity of energy saving that reaches 60 % in June and July due to this illumination method is shown using an example.

Keywords: natural illumination, artifi cial illumi-nation, combined illumination, hollow light guide, tubular light guide, energy effi ciency, horizontal il-luminance, daylight factor, SOLARSPOT system, electric power saving

1. INTRODUCTION

At present new technologies including energy saving are widely introduced into building engi-neering. In particular, hollow tubular light guides (HTLG) have been used for natural room illumi-nation. These are light guiding devices containing a receiver of light radiation, a light guiding hollow tubular channel connecting light (not necessary in a straight line) using multiple refl ections, and also a light distributing device transmitting light from the specifi ed channel (“tube”) to the room (Fig. 1).

HTLGs were developed for the first time in the 1980s in the USSR by J.B. Aizenberg, G.B. Bukhman, V.M. Pyatigorsky, A.A. Korobko, etc. [1, 2]. They also developed illumination design methods using slit HTLGs (light leaves them through a light-transmitting slit in a tube mirrored from within [3]).

HOLLOW TUBULAR LIGHT GUIDES: THEIR APPLICATION FOR NATURAL ILLUMINATION OF BUILDINGS AND ENERGY SAVING

Alexei K. Solovyov

Moscow State Building University E-mail: [email protected]


Fig. 1. Layout of a hollow tubular light guide:1-transparent dome of РММА-HI; 2 – roof adapter; 3 –

HTLG; 4 – RIR* device intercepting and changing direc-tion of luminous fl ux; 4 – fastener for RIR* device


41

As opposed to normal light openings, HTLGs do not cause room heat losses in winter and prevent heat coming into them in summer. Thereby they save energy for heating, ventilation and cooling of the rooms. But their main economic advantage is the decreased power consumption for electric illumina-tion of rooms due to the use of natural light, where natural illumination without HTLG is impossible.

3. USE OF HOLLOW TUBULAR LIGHT GUIDES IN BUILDINGS

At present, illumination devices with HTLGs of the SOLARSPOT series (Italy) [5] are used in the Russian Federation are few in number. As this takes place, HTLGs of 375 mm diameter are used more often, though HTLGs of 250, 530, 650 and 900 mm diameter are also commercially accessible. Each knee of the HTLGs is connected with adjacent ele-ments using slightly conic ends. Besides, together

ple, the KOY series. They allowed a sharp reduction in the number of light sources in the shops and a de-creased cost of air-conditioning to release heat gen-erated by the light sources. The KOY devices were manufactured by Vatra Product Association. Length of slit HTLGs in them reached 24 m.

When coupled with a zenith skylight using spec-ular tubular knees, HTLGs could be used for trans-mission of natural light as well [1].

After the disintegration of the USSR, produc-tion of the KOY devices stopped and has not been restored till now. But in the west (Great Britain, It-aly, Canada and the USA), use of HTLGs is in full swing [4]. Some companies try to import them into Russia, but because of excessively high prices of these products, the process is slow. For HTLGs to be widely used in our buildings, it is necessary to create our own production of inexpensive illumina-tion devices with HTLGs.

Light-guide systems can be vertical or horizontal. In the fi rst case the light-detecting unit is placed on the roof, in the second case it is placed on the wall of a building (Fig. 2). When using HTLGs, the natu-ral illuminance coeffi cient (daylight factor) is equal to the relation of the illuminance in the room calcu-lation point to the simultaneous external illuminance on the horizontal surface.

Compared with windows and skylights that are normal light openings, which are determined with-out consideration of their effi ciency in direct sun-light, HTLGs sharply increase room illumination, when direct sunlight comes into their light detectors. Therefore, when calculating HTLG effi ciency, total natural illuminance should be considered as exter-nal illuminance.

2. ADVANTAGES OF HOLLOW TUBULAR LIGHT GUIDES

HTLGs allow receiving natural light on building roof or walls and conducting it into buildings and constructions with minimum losses, for example to ground fl oors or to basements through garret space, bypassing communications and pipelines. They also allow natural light illumination of shallow depth un-derground stations and tunnels.

In this case many merits of natural illumination remain: continuous spectrum of light, illumination rhythm corresponding to the biological clock of hu-mans and dynamics of light that allows adequate weather evaluation.

Fig. 2. Examples of HTLG system structure with a light-detecting unit on the roof (a), or on the wall (b)


42

η τ τ ξg c d mk= ⋅ ⋅ ⋅ , (1)

whereτс is general transmission factor of the HTLG

dome, its frame and of the intermediate lens (values of τс are given by the manufacturer, so for HTLG SOLARSPOT devices τс = 0.92 [5]);

τd is the diffuser transmission factor (the SO-LARSPOT devices have τd = 0.8, general (simple) glass ensures τd = 0.9, and organic glass provides τd = 0,92, but both of them do not ensure uniform light distribution over a room);

kт is the reserve coeffi cient (it takes into con-sideration the dome contamination when operating, domes should be cleaned as a minimum two times a year, if the dome is cleaned according to the Rus-sian Building Regulation Standards, then at dome cleaning once a year, kт value on average will be equal to 0.77);

ξ is the HTLG effi ciency when dome is located at the HTLG end face (ξ depends on refl ection fac-tor of specula coating of the light guide tube ρ, on relation of HTLG length L to its diameter D, i.e. as a matter of fact, on refl ection number of light rays in the light guide tube and on the slope of rays falling on the HTLG dome relative to its axis).

ξ can be determined [5] in accordance with a sim-plifi ed version of the mathematical equation of mul-tiple refl ections [6, 7]:

ξθ ρ

θ ρ

=⎛⎝⎜

⎞⎠⎟−

eLD

tg

L

Dtg

ln

ln

,

1

1 2 (2)

where θ is angle between HTLG axis and light ray axis (light incidence angle).

Most simply θ can be determined for direct solar component of open-air illuminance. At vertical posi-tion of the HTLG axis, θ is equal to zenith distance of the Sun at this moment. It means that ξ changes with the Sun altitude change. The less the altitude is (over horizon), the less is HTLG effi ciency (be-cause of an increased number of inner refl ections). Therefore here the question is only about HTLG average effi ciency, which corresponds well to the ICI overcast sky conditions. For these conditions θ = 30°. Values of ξ can be also determined in ac-cordance with the Table given in [5] (it is made us-ing expression (2)).

with the SOLARSPOT devices, ventilation systems separating air in air lines from light guide channels, and artifi cial illumination systems built in HTLGs of 650 and 900 mm diameter for night time, can be supplied.

On the whole, devices with HTLGs are more and more often used in schools and child care centres, warehouses, garages and other main and auxiliary rooms of public, industrial and residential buildings in many countries of Europe and North America due to their structural simplicity, energy effi cien-cy, comparatively low price and due to the fact that there is practically no need for regular replacement of elements.

4. CALCULATION OF THE DAYLIGHT FACTOR WHEN USING HOLLOW TUBULAR LIGHT GUIDES

A method of illuminance calculation using HTLG proposed by J. Bracale, is known [5]. However, the method gives rise to doubt regarding illuminance calculation on the operation surface from lighting round or rectangular diffusers. Accordingly, an alter-native method is presented below (see Fig. 3 and 4).

General effi ciency, or HTLG effi ciency [5] can be expressed as follows:

Fig. 3. Layout of a hollow tubular light guide for calcula-tion in accordance with Lambert’s law


43

102–2003 according to the design scheme in Fig. 4. In doing so:

ε η αθ

m d= −( )⎡⎣ ⎤⎦ ⋅ ⎛⎝⎜

⎞⎠⎟ ⋅180 180

21002/ sin , %, (10)

Calculation results according to formulae (9) and (10) completely coincide.

For small rooms, which are typical for office buildings, particularly of an area of no more than 40 m2, and of about 3.5 m ceiling height, calculation of daylight factor average value gives results accept-able by accuracy in accordance with [5]:

εα η

m avd А

S. .

/,=

−( )⎡⎣ ⎤⎦ ⋅ ⋅⋅

180 180100 %, (11)

where S is area of the room, m2.When changing direction of HTLG by means

of various knees, a change (reduction) of ξ be-cause of increase of its effective length takes place. HTLG diagrams with three and four knees are giv-en in Fig. 5. Examples of ξ determination for a tube with an angular adapter for these cases are given in the presented Table.

Refl ected light is taken into consideration just like for systems of top natural light. According to Table Б. 9 in the Standard document CП 23–102–

Luminous fl ux leaving the diffuser can be ex-pressed as

Фd = ηd·Фн, (3) where Фн is luminous flux entering HTLG from outside.

Фн = [(180 – α/180]·Ен А, lm, (4) where α is deviation angle of the tube axis from ze-nith; (180 – α)/180 is HTLG incidence factor; Ен is horizontal open-air illuminance, lx. (If average conditional daylight factor value is required to be determined when using HTLG, we consider Eн = 100 %); А is HTLG section area, m2:

А = πD2/ 4, m2. (5)

Transition from Фd to illuminance on the opera-tion surface in the room calculation point Em for a plain equiluminous radiator is made according to Lambert’s law:

ЕL A

rm

d=⋅ ⋅cos cos

,β γ

2 (6)

where Ld is luminance of the diffuser along a normal to its surface. Other designations are given in Fig. 3.

LФ

А

Еd

d d н=⋅

=−( )⎡⎣ ⎤⎦ ⋅

π

η α

π

180 180/. (7)

Thus

ЕЕ А

rm

d н=−( )⎡⎣ ⎤⎦ ⋅ ⋅ ⋅ ⋅

⋅

η α β γ

π

180 1802

/ cos cos, lx, (8)

and correspondent daylight factor εm = Еm /Ен can be expressed as

εη α β γ

πmd А

r=

−( )⎡⎣ ⎤⎦ ⋅ ⋅ ⋅

⋅⋅

180 180100

2

/ cos cos, %. (9)

εm in any point of the room from a round or rectan-gular diffuser can be also calculated in accordance with formula (12) in the Standard document CП 23–

Fig. 4. Layout of a hollow tubular light guide for calcula-tion in accordance with formula (12) in the Standard docu-

ment СП 23–102–2003


44

luminated with four HTLGs of the SOLARSPOT system of 350 mm diameter according to the design scheme (Fig. 6). The room is located in the middle of a two-storey building on its ground fl oor and can-not be illuminated through windows and doors us-ing natural light.

Calculation in accordance with Lambert’s law First determine effi ciency of the system using

the Table:L’

1 = 4.8 m; L’2 = 1.5 m; L’

3 = 0.3 m; АА 90 = 0.6 + 3×0.375 = 1.725 m.

L = 4.8 + 1.5 + 0.3 + 1.725 + 1.725 = 10.05 m; D = 0.375 m; L/D = 10.05/0.375 = 26,8.

Further according to [5, Table 2], with ρ = 0.98 and L/D = 26.8 using interpolation, found:

ξ =−

−⋅ −( ) + =

0 71 0 44

50 19 450 26 8 0 44 0 645

. .

.. . . .

For SOLARSPOT devices

2003, r2 value is determined depending on the hф/ l1 relation, on ρav value and on bay number n. Here r2 is a coeffi cient taking into consideration refl ected light from inner surfaces of the room, hф is diffuser height over the operation surface, l1 is width of the bay (room) and ρav is average refl ection factor in the room.

Refl ected component of daylight factor ео is de-termined using the expression:

ео = εm.av. (r2–1). (12)

Therefore fi rst, one should determine Em not less than in fi ve points of the room area and determine εm.av. Then determine ео according to formula (12). And fi nally, determine total daylight factor em:

еm = εm + ео. (13)

Example. Calculate the daylight factor in a room of 6×6 m fl oor size and of 3.5 m ceiling height il-

Fig. 5. An example of placing hollow tubular light guides in buildings


45

Table 1. Practical determination of ξ for a tube with an angular adapter using the method of optical equivalent length and/or L/D relation for versions with three (a) and four (b) knees, Fig. 5, [5]

а)D, mm 250 375 530 650

L’1 0.5 1 2 1

L’2 2 1.5 0.5 2

L’3 2 3 4 5

AA 30(1) 0.30 + 0.25 0.30 + 0.375 0.40 + 0.53 0.40 + 0.65AA 30(2) 0.30 + 0.25 0.30 + 0.375 0.40 + 0.53 0.40 + 0.65

L, m 5.6 6.85 8.36 10.1ξ 0.91 0.92 0.943 0.939

b)D, mm 250 375 530 650

L’1 1 1.5 2 0.5

L’2 2 1 0.5 1

L’3 2 2.5 4 3

L’4 1.5 2 0.5 4

AA 30 0.30 + 0.25 0.30 + 0.375 0.40 + 0.53 0.40 + 0.65AA 60 0.60 + 2×0,25 0.30 + 2×0.375 2(0.40 + 0.53) 2(0.40 + 0.65)AA 90 0.60 + 2×0.25 0.30 + 2×0.375 3(0.40 + 0.53) 3(0.40 + 0.65)L, m 9.5 10.75 12.58 14.8

ξ 0.847 0.883 0.905 0.91

Fig. 6. Layout of placing hollow tubular light guides at a cross section of the considered room


46

the distance from working surface diffuser centre projection to the calculation point is designated as a in Table 2.

Calculation according formula (12) in the Stand-ard document СП 3–102–2003

ηd = 0.44 (see calculation according to Lambert’s law).

In formula (10) sin (θ/2) values for different points of the room are different:

εθ

m = ⋅ ⋅ ⎛⎝⎜

⎞⎠⎟ ⋅0 44 1

21002. sin . Calculation is made

in a tabulated confi guration, Table 3.Thus calculation results in that the two last Ta-

bles are identical.Total εm in each of the specifi ed points from four

HTLGs located symmetrically (Fig. 7), are accord-ingly equal to, %: 0.21 + 0.042 + 0.042 + 0.025 = 0.321 (in point 1); 0.183 + 0.074 + 0.074 + 0.025 = 0.356 (in point 2); 0.124 + 0.124 + 0.042 + 0.042 = 0.332 (in point 3); 0.356 (in point 4); 0.321 (in point 5); 0.025 + 0.183 + 0.042 + 0.025 = 0.275 (in point 6) and 0,275 (in point 2’).

еav = εм ii

, /=∑

1

7

7 = 0.319 ≈ 0.32 %

According to Table Б. 9 in the СП 23–102–2003 (attachment 5.8) with hф/l1 = 2.7/6 = 0.45;

n = 1 and ρav = 0.5, r2 = 1.39,еav = 0.32·1.39 ≈ 0.445 %.

If calculating the average value of the daylight factor in such a room according to Bracale, we ob-tain the following:

η τ τ ξd c d mk= ⋅ ⋅ ⋅ = ⋅ ⋅ ⋅ =0 92 0 8 0 92 0 645 0 44, , , , , .

In formula (9), cos β = cos γ and r2 values change

for different points of the room

επ βπ

β

мr

r

=⋅ ⋅ ⋅ ⋅

⋅ ⋅⋅

⋅ = ⋅

0 44 1 0 375

4

100 1 547

2

2

2

2

, , cos

,cos

, %.

The calculation is made in a tabulated confi gura-

tion for each calculation point. As this takes place,

Table 2. Table of calculation

of point а, m tg

a

hф

β = β, grade cos β cos2 β r2=a2+h2ф Εm=1.547·(cos2 β/r2), %

1 0 0 0 1 1 7.29 0.21

2 0.75 0.28 15.52 0.963 0.928 7.85 0.18

3 1.5 0.56 29.05 0.874 0.764 9.54 0.12

4 2.25 0.83 39.81 0.768 0.590 12.35 0.07

5 3 1.11 48.01 0.669 0.447 16.29 0.04

6 3.75 1.39 54.25 0.584 0.341 21.35 0.025

Еav = 0.108%

Fig. 7. Design scheme of a room illuminated with four hollow tubular light guides (a plan with daylight factor

isolines from each light guide)


47

In accordance with the usage method with the room index i = 1, ρcl = 0.7, ρwl = 0.5 and ρfl = 0.3, the necessary luminous fl ux of the lamps in each lumi-naire Фl = 11 880 lm.

We shall consider four ЛХБ-40 lamps to be in each luminaire.

Then specifi c annual electric power consump-tion for the room illumination wep will be equal to 2292.5 kW·h [8].

Four HTLGs installed in the room, ensure еav = 0.445 %.

With such еav, critical external illuminance Ecr (i.e. illuminance, at which one should turn on and turn off luminaires, will be equal to

Е

ен

ср

⋅=

⋅100 300 100

0 445,= 67 416 [lx].

This is more than the maximum external total natural illuminance in Moscow in July at noon, and it means that an effect can be only reached at smooth ACCI.

Calculation of electric power saving due to HTLG use at smooth (continuous) ACCI is made in accordance with diagrams of change of external total natural illuminance. This saving corresponds to an annual quantity of natural illumination during working hours from 10 a.m to 6 p.m.

The sum of the areas of similar fi gs. for each month is equal to the annual quantity of natural illu-mination outside. To obtain annual quantity of natu-ral illumination in a room, one should multiply the natural illumination outside by the average daylight factor and by 100.

In our example, annual quantity of natural illumi-nation is equal to 191, 137 lx·h.

Annual quantity of artifi cial illumination in the room without HTLG is equal to

еav = Фd/S = ηg· [(180-α)/180]·100·3.14· (0.3752/4)/36 = 0.135 ( %) – from one HTLG;

εm.av = 0.135·4 = 0.539 %. In accordance with Bracale, taking into account refl ected light, εm.av = 0.539/ (1–0.5) = 1.079 %, i.e. it is more than twice as large.

At external natural illuminance with a clear sky, ЕQ = 25 000 lx (on the average), average illuminance in the room Eav = (25 000·0.445)/100 = 111.25 [lx], and at external natural illuminance with overcast sky, ЕD = 15 000 lx (on the average),

Eav = (15 000·0.445)/100 = 66.75 [lx].

Maximum Eav in Moscow can be equal to 53 000·0.445/100 = 236 [lx]. And this means that normalised illuminance in the statistical aspect can only be reached with the combined illumination.

And the greatest energy saving effect can obvi-ously be reached at a continuous automatically con-trolled combined illumination (ACCI), smoothly raising and lowering levels of additional artifi cial illumination when changing natural illuminance. As this takes place, an illuminated room can be used for any production purposes or as an offi ce.

As an example, we shall estimate illumination en-ergy saving of an offi ce room due to the application of HTLGs. At the total illumination, average stand-ard illuminance Es according to the Building regula-tions 23–05–95* is equal to 300 lx. The system of ar-tifi cial illumination is formed by built-in luminaires with specular refl ectors, a lattice (group 22) and fl u-orescent lamps ЛДЦ-40. Ceiling height is 3 m, the area of the room S = 36 m2, 100 lx luminaires spe-cifi c power is equal to 8.6 W/ m2; and their ge neral

power, accordingly 8 6 36 300

100928 8

,, ,

⋅ ⋅= [W].

Table 3. Table of calculation

of points а, m

θ12

=+

аtga D

hф

/,

grade

θ22

=+

аtga D

hф

/,

grade

θ/2,grade sin (θ/2) sin2 (θ/2) Εm = 0,44·sin2(θ/2)

100,%

1 0 3.97 3.97 0.036 0.0048 0.21

2 0.75 19.148 11.768 3.69 0.064 0.0041 0.18

3 1.5 32.005 25.920 3.04 0.053 0.0028 0.12

4 2.25 42.075 37.376 2.35 0.041 0.0016 0.07

5 3 49.733 46.169 1.78 0.031 0.00097 0.04

6 3.75 55.56 52.84 1.36 0.024 0.00056 0.025


48

method of calculation of daylight factor from HTLG allows determining the necessary number of these devices for the illumination of rooms. And daylight factor takes into consideration both sky diffused light, and direct sunlight. Therefore when calculating natural light usage time, one can use not data about state of diffuse but of total external natural illumi-nance. It is shown that most effi ciency is provided by hollow tubular light guides in combination with smoothly and automatically controlled combined illumination. Energy saving due to artifi cial illu-mination only in the considered example, reaches 59–58 % in June and July, 36 and 31 % respective-ly in March and September, and this energy saving is minimum in December and January: 4.7–3.5 %.

But the main advantage is that in rooms, in which normal systems of natural light cannot be installed, natural lighting with all its benefi ts will be possible.

REFERENCES

1. Aizenberg Yu.B., Bukhman G.B., Pyatig-orsky V.M.. A new principle of illumination using

300 [lx]×12 [month]×21 [day/ mounth] ×8 [hours/day] = 604, 800 lx·h.

Annual saving of the energy is equal to 31.6 %, or 321.69 kW·h.

However, the main advantage of HTLG use is that natural light will be supplied in the room in any event. And this will take place without any cost for completion of heat losses through light openings and without additional power consump-tion for cooling.

As it can be seen from Fig. 8, in accordance with this example, energy saving in December is equal to 4.7 % and in June and July it is equal to about 60 %.

MAIN CONCLUSIONS

Hollow tubular HTLGs, which are an innova-tive system of natural building illumination, allow conducting natural light into rooms that cannot be illuminated using normal systems of natural light, for example into basements, central rooms of wide buildings located at ground fl oors and transporting natural light through garret space. The developed

Fig. 8. Diagram of the change in electric power savings during the months of the year


49

6. Sastrow F., Wittwe, V.. Daylight with mirror light pipes and with fl uorescent planar concentra-tors. // Proceedings SPIE. 1987.

7. Chao, B.L. Solar tube performance and pay back analysis. / International report STI. July 29, 1996.

8. Textbook on calculation and design of natural and combined illumination (for the Building regula-tions II-4–79)/NIISF of the Gosstroy of the USSR. – Moscow: Stroyizdat, 1985.

illumination installations with slit light guides.// Svetotekhnika. – 1976. – 2.

2. Aisenberg, Yu.B., Bukhman, G.B., Korobko, A.A., Pyatigorsky, V.M.. Experiense and perspective of hollow guide lighting system development and application. / Proc. 25 th Sess. CIE: San Diego, 2003.

3. Korobko A.A.. Usage factors of illu-mination installations with slit light guides.// Svetotekhnika. – 1988. – 5.

4. Pein.T.. Development of hollow HTLGs in Great Britain.// Svetotekhnika. – 2004. – 3.

5. Bracale J.. Natural illumination of rooms using a new passive light guide system “SOLARS-POT”.// Svetotekhnika. – 2005. – 5.

Alexei K. Solovyov, Dr. of tech. sciences, professor. He is graduated from the Moscow Engineering-and-Construction Institute of V.V. Kuybyshev. At present, he is the head of the Chair of the Architecture of the Moscow State Construction University, member of the European Academy of Sciences and Arts, member of the editorial board of the Svetotekhnika Journal

50

ABSTRACT

1From ancient times, the walled city of Jerusalem was built, moved and developed according to water supply availability. Its long and world famous his-tory can still be read and understood by looking at the handmade sculpted landscape that surrounds the city. The complicity between the holy city, the land and the topography is what gives the old city of Je-rusalem gives its special character and charm.

The aims of the study was to defi ne what role the wall should play in the night-time shape and sil-houette of the old city, to generate an impression of the surrounding nightscapes, the nocturnal am-biences in the streets and the architectural lighting of the major landmarks located inside and outside the walled city.

Keywords: Jerusalem, lighting, master plan, nightscape, light pollutions

INTRODUCTION

The old city of Jerusalem has faced numerous battles, wars and conquests since its early begin-nings and it is still claimed as a capital city by both the state of Israeli and Palestinian authority today (Fig. 1.).

The city of Jerusalem, as a state capital, is not recognized by the international community and the old city is still considered to be an occupied city by the United Nations, including by the French government.

* This paper was fi rst presented at the PLDC Conference, Madrid, October 20, 2011

Jerusalem has become a world famous destina-tion for pilgrims of the three main monotheist re-ligions. The Israeli authorities in charge of the de-velopment of Jerusalem have therefore decided to create a lighting master plan for the old city and its surroundings in order to promote a night-time im-age to enhance cultural activities and support tour-ism (Figs. 2, 3).

Energy consumption, light pollution, and a sus-tainable approach were clearly some of the tactical and political challenges of this lighting master plan.

After a careful daytime and night-time analysis of the baseline situation and with the precious help of our Israeli partner, Eduardo Hübscher from Ariel Municipal Company, we were able to study and pro-pose a highly original lighting master plan (LMP), which was offi cially approved by the Israeli authori-ties in October 2010.

THE PREDOMINANT ROLE OF THE NATURAL LIGHT

The actual location of the old city, north of the oldest part of Jerusalem (the City of David) was de-termined by the two main valleys that run to the east (The Kidron River valley) and to the south west (the Hinnom River valley). The city is located on a slope oriented towards the south east and facing the high Mount of Olives in the east and surrounding by low-er hills to the south and to the west.

The leaning position and the orientation of the old city combined with the form of the entire wall create different daytime perceptions of the city de-pending on the curve of the sun, the weather con-ditions, and the season. Given the latitude of Je-

THE OLD CITY OF JERUSALEM LIGHTING MASTER PLAN *

Roger Narboni

Concepto, France E-mail: [email protected]



51

the west, the topography denies also this possible reading.

The narrow streets, mainly oriented east-west or north-south, offer a constant sunny side and strong projected shadows. The shadows cast by the build-ings are sharp-edged and strong and the sunny façades create a very high contrast due to the clear tones of the pavement and the construction materials. The large amount of sunlight added to the oriental atmosphere of the old city creates a chaos of visual information that sometimes hampers the discovery of the main landmarks (domes, church towers and minaret). Some streets are covered by numerous arches; others by improvised canopies. The ground is full of projected shadows.

The sky is not always visible and becomes a very thin strip in the narrow streets. There is a wide diver-sity of street confi gurations combined with the im-posing topography (stairs, important slopes).

The density of shops in some streets and the souk ambiences qualify the diverse quarters more than the type of population. And all these characters give its unique and world famous identity to the old city of Jerusalem.

rusalem (31.8°), the maximum sun angles at noon are: 81.8° during the summer solstice and 34.7° during the winter solstice. The north wall of the old city remains in shadow most of the day and in-deed of the year. The variations of light intensity throughout the day and the year are therefore also refl ected in the image of the walled city. The rising sun is dominated by green radiations while the sun-down brings a lot of red colourations. The chang-ing colours of the sun and the way the local stone refl ect the light continuously transform the visual appearance of the wall.

The domes, minarets and church towers that rise above the skyline of the old city propose a lot of colours and textures that react differently with the sunlight.

The inner side of the wall is rarely seen due to the orientation of the sun and the proximity of the buildings. The trees growing inside the city, drown in the powerful mineral environment, and are barely visible from afar.

The size of the old city and its peculiar shape cannot be read entirely from the east due to the proximity of the new city in the perspective. From

Fig. 1. Jerusalem

Fig. 2. Jerusalem image

Fig. 3. East night (computer rendering)


52

skyline, but it is drowned in bright high-pressure so-dium light from the esplanade.

Inside the Old City, the wall is lit from the out-side (with cold white light) except on the north east corner. All along the wall different types of light-ing (projectors on dedicated mast, on street poles, installed in the ground, on the opposite wall near Zion gate) create a lot of diverse luminous effects. The gates are not specially accentuated. The streets inside the city are mainly lit by luminaires mounted directly on the wall and equipped with high-pressure sodium lamps that do not render the natural, clear colors of the stone architecture. There is al lot of dif-ferent lighting fi xtures (lantern style, “diamond”, white ball, etc.).

The streets inside the city are lit mainly with lu-minaires directly fi xed on the wall equipped with high pressure sodium lamps that do not render clear tones of the stone architecture. The shops that pre-dominate in the Christian and the Muslim quarters are brightly lighted with a cold white light that over-whelms the street lighting at the beginning of the night but then disappears totally when the shops close.

Few historical buildings or landmarks are il-luminated inside the Old City (the golden Dome of the Rock, the Dome of Al Aksa Mosque, the Citadel of David and the Tower of San Salvador Church). Some of the numerous Minarets are sig-naled by green vertical strokes.

The inner landscape is never enhanced at night.

THE EXISTING PUBLIC LIGHTING

In the surroundings, the Jewish Cemeteries on the slope of the Mount of Olives are fl oodlit with cold white light (projectors on poles directed towards the hill). This installation generates a lot of glare for vis-itors on the top of Mount of Olives looking out onto the beautiful perspective offered by the walled city.

The Muslim cemetery near the eastern wall is also fl oodlit by projectors (cold white light) in-stalled on street poles.

This kind of lighting located far from the illumi-nated space is very consuming in term of energy. The wide and strong lighting effect disturbs from the east the possible views of the illuminated wall.

On the west side, the streets of the new city lo-cated closed to the old city are brightly lit with high pressure sodium lamps (orange coloured light) on high road poles.

The modern city is also comes very close to the northern side of the old city and the street lighting competes with the wall lighting.

The main street that surrounds the old city is close to the wall almost all the way around. The street lighting (high pressure Sodium) is frequently directed toward the exterior of the old city and pro-duces a lot of glare in the far night vision.

Few monuments or landmarks are lit and most of them are diffi cult to recognise at night. The gold-en Dome of the Rock is lit and marks the old city

Fig. 4. The map of Jerusalem outside illumination


53

is contained in all of the landscape in one single stone and was well known in ancient times as at-tested in the Hebraic, Assyrian, Persian, Greek and Latin languages).

The relief and the landscape that still tell stories of the old city’s past today will be enhanced at night in a very poetic manner by lines of luminous col-oured dots installed along the tops of the hills and inside the two river valleys. The slope of the hills that surround the old city will be voluntarily kept in the dark and the landscape that now occupies the bottom of the two valleys will be lit with delicacy. Inside the Kidron valley, the tombs will be lit with a golden light to complete the nightscape.

The wall, underlined continuously with a gold-en light and enhanced in the angles with a higher intensity, will play the most important role in the night vision (Fig. 5). The inner face of the wall will be lit also in some sectors in order to draw at night the familiar shape of the old city that can be observed from the top of Mount of Olives but also from the south vision. The towers included in the wall are lit the same way but the two side faces are kept in the dark to help read the volume and the re-lief at night.

The gates will be signaled by the golden lighting outside and inside. The details of the gates will be enhanced at night and the textures and reliefs will be underlined because of the up-lighting principle (Figs. 6, 7). On the west side, the exterior wall of the

THE LIGHTING CONCEPT FOR THE OLD CITY AND ITS SURROUNDINGS

A jasper landscape set in the darkness

Jerusalem is surrounded by a green belt, which is under development and follows the ancient val-leys of the two rivers. This green belt was consid-ered in the study as a potential area of darkness that could give the opportunity to underline and increase the contrast with the illuminated walled city (Fig. 4).

The overall night-time view of the walled city proposes something totally different from the day-time image. The lighting will allow the visitor to better understand the city morphology by night, its walled shape and its particular relationship with the topography and the surrounding landscape. The principal monuments, when illuminated, will be easily recognisable and create a mental map that guides people through the space and the amazing past of Jerusalem.

The lighting will reveal the diverse layers that composed the old city of Jerusalem: the surround-ing landscape, the wall, the buildings and streets, the trees, the vertical elements and the numerous domes. This “layered lighting” will create a unique and symbolic night-time image that evokes the im-pression of the monumental jasper landscape (jas-per being a very special and colourful mineral that

Fig. 5. West wall panorama (computer rendering)


54

leave the view of the wall free of any unsightly verti-cal poles). The intensity of the lighting will decrease naturally from the bottom to the top of the wall.

The southern wall and the western wall of the Es-planade should be lit in continuity with the southern and eastern wall of the Esplanade to show the monu-mentality of the Mount and its height.

Above this first illuminated golden layer, the main trees that rise from the wall will be lit as well to give density to the night image and composed this very particular nightscape, Fig. 8.

The lighting of the main landmarks and monu-ments than can be found inside and outside the Old City of Jerusalem will be very important because all these focal points will help the visitors to bet-ter understand the old city morphology and their way through it. All these architectural elements can be identifi ed easily by the type and color of their dome or roof, and the presence of a vertical element such as church tower or minaret. These architectur-al elements will be treated with a warm white light (3000 K) to render the colour and texture of each one and to insert them harmoniously into the Jeru-salem skyline. The lighting will be deliberately fo-cused on the higher elements of the architecture. On a lower level, the lower sections of the architecture (the façades that relate visually and are adjacent to the public space) will also underlined and revealed but from the point of view of the pedestrian, with the same quality and luminous color.

impressive Citadel of David lit the same way will an-nounce the old city silhouette.

The gold lighting will systematically come from the ground near the wall to show its monumentality and reveal its texture and reliefs (this principle will

Fig. 6. Damascus Gate

Fig. 7. Lion’s Gate


55

ment to host special events like those programmed during the annual Jerusalem light festival held every year in the month of June.

THE JAFFA GATE LIGHTING

This gate, the main entrance in the west of the old city is dominated by a minaret and the towers of the famous David’s Citadel. It is part of a larger regen-eration project under construction to lower the im-pact of car traffi c. It will be the fi rst important light-ing project that is to be realised at the end of 2011, in accordance with the lighting master plan.

The new lighting will beautify the night-time im-age and reveal the monumental walls of the Citadel. It will provide clear pedestrian access to the old city street system and is designed to contribute towards creating a pleasant nocturnal ambience in this re-newed space.

On the fi rst part of Omar Ibn El-Khattab square, from Jaffa Gate to David’s street (that is very wide), contemporary urban multiple poles (from 10 to 5 m high) are used to lit all the public space from the building to the deck railing and then on the sec-ond part (where the space is narrower), luminaires with the same design are fi xed directly on the lower façades to free the space on the ground. The street circulation, limited at 30 km/h, allowed us to re-search for the street an average illuminance level of 15 lx depreciated.

All the inner streets will be treated with warm white light. The lighting fi xtures will be pendant mounted or mounted onto the façades and they will diffuse coloured light to the back walls. The nu-merous architectural details located along the main streets (arches, windows, balconies, carved doors, fountains) will be enhanced at night by the lighting

MURISTAN SQUARE

This square located south of the Holy Sepulcher is a pedestrian square with a colourful Greek bazaar in the heart of the Christian quarter. The Mora foun-tain in the centre of the rectangular square is set up on an ellipse comprising strips of stone (Fig. 9.).

Completed in June 2010, the new lighting for the fountain is now the luminous focus of the night-time scene. The four pillars are lit up on three sides with warm white light and on the northern side an intense blue downlight fl oods the fountain, creating the inter-play of bright faces and multiple shadows that reveal its composition. Tiny LED projectors complete the light composition and enhance the grotesque masks at the base of the pillars. The stone graphics on the ground are underlined and punctuated with luminous dots (in-ground coloured LEDs). The polished stones refl ect beautifully in the surrounding lighting.

The nocturnal ambience now proposed for Muris-tan Square is very soft and friendly and the resulting night-time scenario has become a perfect environ-

Fig. 8. Jerusalem LNP nocturnal densities map


56

Fig. 9. Muristan fountain: general night view and illumination of details


57

The very specifi c and complex political situation of Jerusalem was also a major consideration, given that the proposal for a lighting master plan had to be accepted by the Israeli government as well as by the Palestinian authorities.

If light and lighting go a little way to create and further a dialogue between the two age-old neigh-bors sharing this wonderful city, then the lighting master plan process can be considered to be a great success.

Lighting designer – Melina Votadoro, Florence Serre, project managers, CONCEPTO studio

Client: Jerusalem Development Authority, Reu-ven Pinsky, director – Eduardo Hübscher from Ariel Municipal Company, project manager.

Architect of Jaffa Gate project: Gaï Igra, EECC electrical engineering

The canopies above the shops on the fi rst level (created by the Israeli architect Gaï Igra in charge of the whole regeneration project) are used to car-ry little down-light LED projectors that will bring a nice lighting on the space near the shops.

The lighting of the most interesting façades takes part in the whole nocturnal ambiance. For these façades, the canopies are also used to locate linear up-light LED fi ttings directed toward the upper lev-els that will give nice diffused lighting.

In addition, the architectural details of these in-teresting façades are enhanced at night with the help of LED micro-projectors (3000 K), almost invisible during daytime.

Designing a lighting master plan for such an old city was an amazing challenge. This kind of study questions the role of contemporary lighting in a an-cient site and how the lighting should render a herit-age site with a world famous daytime image differ-ently at night.

Roger Narboni is a visual artist and has a degree in electronic engineering. After several trips and sojourns abroad, including 3 years in New York, he settled near Paris in 1981 to devote himself to research and creative work on light and space. He has created several installations and monumental pieces on this theme and developed sensory projects including the famous exposition “La lumière dans tous ses états” at the Centre d’Art Contemporain in Orléans (1985), which was re-exhibited at the Cité des Sciences de la Villette (1986).Following these experiences he became the fi rst lighting designer in France and coined the French term concepteur lumière and then devoted himself exclusively to urban and architectural lighting. Roger Narboni has been a member of the Association française de l’Eclairage since 1992 and was one of the founders of the Association française des Concepteurs lumière et Eclairagistes, where he served as fi rst president from 1994 to 1999.With the Concepto team he has designed over 80 regional planning and development projects for lighting in France and abroad. Since 1999 he has been doing original research on the night landscape and the ecology of lighting which materialized in a new treatise La Lumière et le paysage published in 2004 (translated in Portuguese, English and Italian).In 2002, along with Laurent Fachard, he became artistic co-director of the Fête des Lumières in Lyon and in 2003 the head of the exploratory program committee. Since 2003 he has been serving as international expert with the LUCI association, the world network of cities of light

58

increase of power effi ciency” forbids “… commis-sioning buildings and constructions, which are not correspondent to the record devices and power effi -ciency requirements”.

The aim of this article is to understand, what capabilities of energy saving can be implemented in systems of interior illumination (offi ces, educa-tional and public institutions, stores, sports com-plexes, etc.), and which technical solutions ensure a considerable electric power saving when observing the established illuminance standards.

For example, we consider today’s most wide-spread illumination systems, in which gas-discharge fl uorescent lamps are used as light sources, and take as a reference point a system, in which standard fl uo-rescent lamps and standard electromagnetic starting controllers (ballasts) are used (see Fig. 1).

As can be seen from the presented example, the use of improved ballasts with low losses can save up to 7 % of electric power, electronic ballasts provide an opportunity to reach 22 % saving, and in combi-nation with an illuminance detector automatically controlling the luminance depending on daylight lev-el, one can save up to 42 %. If an elementary manual system of luminance control is added, it will be pos-sible to reach 55 % saving, and if using new genera-tion fl uorescent lamps (Т26/Т8), then – 71 %. And fi nally, the inclusion of an additional presence de-tector in the system in conjunction with most power effective fl uorescent lamps (Т16/Т5) currently al-lows achieving a maximum energy saving 61 % and 82 % accordingly.

It is no coincidence that the Federation of As-sociations of National Manufacturers of Illumina-

ABSTRACT

A simple, effective and comparatively inexpen-sive system SensaModular for interior illumination control, which is proposed by one of leaders on the world lighting market Thorn Company, is described. This modular system allows making the decision on necessity of modules for development of programs of illumination scenarios, of connection with the natural illuminance level detector, with the presence detector, as well as for switching, adjustment of the illumination program and for infra-red control. As a result, only the necessary components are connected to the base controller unit, and this ensures control of a system containing up to three luminaire groups with luminance control using a controller (a control unit). Application of presence and illuminance de-tectors is described in detail. Examples of the Sen-saModular system use are given.

Keywords: energy saving, illumination control, SensaModular system

1. INTRODUCTION

At present the challenge of increasing the of pow-er effi ciency of buildings and constructions has be-come extremely topical, not only due to gradual and constant rise in energy carrier prices, but also be-cause of a great attention, which is given by federal authorities of the Russian Federation to increasing economic effi ciency as a whole, including energy effi ciencies of illumination systems. For example, Federal law of the Russian Federation of the 23rd November 2009 261-Ф3 “On energy saving and

ENERGY SAVING CAPABILITIES WHEN USING CONTROL SYSTEMS FOR INTERIOR ILLUMINATION

Julia B. Babanova and Vadim A. Lunchev

Representative offi ce of Russia Thorn Lighting, Moscow E-mail: [email protected]



59

systems, all the more due to the fact that more tech-nically perfect and more expensive luminaires are necessary for their work. Nevertheless, two factors inspire confi dence that we are on the right track: 1) the government actively promotes energy saving and energy effi ciency ideas both on the political, and on the legislative level, and 2) in the projects, which “are constructed for themselves”, the customers and investors insist that an illumination control system and advanced illumination equipment are integrated in the building. The latter obviously leads to reduc-tion of both electric power cost and of electric power connection cost, as well of operation, lamp replace-ment and maintenance service expenses.

3. THORN SENSAMODULAR SYSTEM

What do manufacturers propose today so that a control system could be easily and simply installed, adjusted and maintained? Let’s consider as an ex-ample, the SensaModular interior illumination con-trol system proposed by Thorn Company, one of the world leaders on the lighting market.

The main advantages of this control system are simplicity, effi ciency, ease of installation and rea-sonable price. And in addition of course, is system modularity, which is similar to a Lego construction set, where the user himself can select the “bricks” necessary to construct a system needed for him en-suring the required functionality without overpaying for superfl uous functions, which equipment manu-facturers like to foist with their systems. At the stage

tion Equipment and Electrotechnical Components for Illumination Equipment in the European Union (CELMA) specially accentuates an essential electric power saving when using electronic ballasts with lu-minance control, natural illuminance detectors and detectors of people’s presence in the room.

Now it becomes clear that to achieve a consider-able energy saving, not by a few percentage points or even by ten or twenty percent, one should more drastically use modern illumination control systems in combination with modern light sources, includ-ing light-emitting diodes, because the main princi-ples of energy saving in this case remain invariable.

2. BARRIERS

What constrains the wide application of interi-or illumination control systems? First, it is the low awareness of design engineers, customers and inves-tors in the control system fi eld. Secondly, it is the perception of a control system as very complex, ex-pensive and unreliable electronic equipment. Third-ly, it is a desire to save as much money as possi-ble on illumination equipment, which, as is well known, is required during the fi nal stage of construc-tion, when frequently there is no money not only as planned for illumination equipment in accord-ance with the original budget, but generally fi nanc-ing is carried out by the residual principle: if only to install any luminaires in order to somehow light. Such realities lead to a situation, when contractors and customers in the fi rst place give up on control

Fig. 1. Stages of energy saving for modern illumination


60

the presence detector, as well for switching, illumi-nation program setting and for infra-red control. On balance, only the necessary components will be con-nected to the base controller unit: those components that provide control of a system consisting of up to three luminaire groups with luminance control using a controller (a control unit). A picture of the THORN SensaMod-3 DIG controller is given in Fig. 2.

Unlike traditional luminance control systems or integrated solutions intended for a whole building, the SensaModular system represents a simple, eco-nomically effective modular solution, which allows saving time and labour expenditures for all types of use inside one room of a building with two or three luminaire groups.

The modular structure is based on a controller, which interacts with the panel of program choice, with detectors, switches and IR receivers to control luminaire groups in such a manner so that the user

of control system confi guration, you yourself must make a decision, whether you need modules to cre-ate illumination program scenarios for connection with the detector of natural illuminance level, with

Fig. 3. System of illumination control

Fig. 2. THORN SensaMod-3 DIG controller


61

nation control can be implemented in such a way, for example, in corridors and staircases, as well as in rooms, where installation of illuminance detec-tors is not required, and IR remote controls will not be used.

The SensaModular controller (Fig. 4) is equipped with two built-in switches (in the fi g. they are black), which give the user the possibility to choose neces-sary operating modes and a presence detector delay time.

Flexibility of the system is provided by the fol-lowing operation modes:

Turn-on/turn-off mode: detection of presence and absence of people. The illumination will be turned-on, if a person is detected by the presence detector. Then illumination luminance is smoothly decreased until full switching off after the detector will detect people absence and delay time expires. If several detectors are used, the SensaModular system will only start to reduce/switch off illumination after all detectors inform on people’s absence.

Turn-off mode: detection of people’s absence.This mode is similar to the previous except for

that illumination can only be turned on manually us-ing an IR remote control, panel of program choice or push-button switches.

Corridor mode: detection of people presence.This mode provides an operation algorithm simi-

lar to the turn-on/turn-off mode but without full turn-ing illumination devices off, with their luminance re-duction to a 10 % level of the rated one.

There is also a special function of all presence detectors switching-off, which is useful when diag-nosing and commissioning the system.

The switch-off time delay is set by a separate switch in the interval from zero to one hour. It should be remembered that when using some presence de-tectors with a built-in delay, this time delay will be summarised with the delay time set in the Sen-saModular controller.

One more useful mode is connection of a stand-ard keyboard switch without blocking to the input

can obtain as a result, a fl exible and comfortable illu-mination control system, and also obtain a guarantee that all the components are completely compatible with each other, Fig. 3.

The main advantages of the proposed system are as follows:

1. The possibility to install several presence de-tectors that allow adaptation of the SensaModular to a necessary version of presence determination and to various operation modes;

2. Availability of an illuminance detector in the system ensures a simple adjustment of the Sen-saModular for a necessary fi eld of application, es-pecially in relation to the room size, its confi gura-tion, and to use of the ceiling height;

3. Adjustment of the illumination program us-ing a panel of program choice, IR remote control or by means of switches;

4. Output signals of presence detectors in the DSI/DALI standard in order to control luminaire groups, two versions of the controller: with three or with two digital outputs;

5. A simple connection by means of nonpolar two-wire connections using mounting materials with standard rated mains characteristics.

4. PRESENCE DETECTOR

Let us consider in more detail the use of presence detectors. It is known that the most expensive light is the light left in empty rooms or on empty sites. The SensaModular system allows automatic energy saving in the rooms, where people are absent. One can choose one of three versions:

One multidetector for defi nite group of lumi-naires. This version is ideally suited for areal (zone) illumination, for example, for group offi ces. It cor-responds to the recommendations of the standard EN 15139, which provides for use of presence detectors in the rooms of more than 30 m2 area.

One multidetector for all groups of luminaires. This solution is suitable for small rooms, where use of presence and illuminance detectors is usually re-quired at the same time, and illumination system operation by means of an IR remote control is also necessary.

One presence detector for all groups of lumi-naires. This version is an alternative of multi detector use, and in this case any presence detector for 230 V, which is connected to the special PD input of the controller, will be suitable. Other models of illumi-

Fig. 4. SensaModular controller


62

into sites, for example for group offi ces. The multi-detector presented in Fig 6. provides measurement of natural illumination level. It contains a built-in presence detector and a receiver of the IR control system.

When installing the system, one should take into consideration that detectors measure levels of artifi -cial and natural illumination, including the one re-fl ected from the working surface, so both used ma-terials, and their colours are important. Besides, for each group of luminaires one multidetector is re-quired, therefore the illumination can be perceived as more obsessive. Overlapping effect of detection zones of people’s presence can also take place.

The next solution is to use a “tracing” photocell directed towards a window and detecting the level of natural illumination only (Fig. 7).

This photocell is installed on the ceiling and has an effect on all groups of the luminaires. Advan-tages of such a solution consist in that visual dis-tortions of illumination are reduced to a minimum, because one detector is used for no more than three groups of luminaires. Besides, because of direct de-tection of natural illumination level, energy saving appears to be maximised. An optimal version of such a solution is when it is used for luminaires installed in rows and/or at ceiling height of no more than 3 m, as well as in rooms, where connection with people presence is not required, for example in classrooms, production rooms, gyms, etc. (Fig. 8).

The last version is use of one detector “directed downwards” for all groups of luminaires. Such a solution is suitable for rooms of a little size, where there is no necessity of zone illumination, for ex-ample, for separate offi ces and conference rooms. In this case, for detection of illuminance level, one detector directed to the working surface is only in-stalled. It is obvious that in this case, illumination stability is also ensured regardless of lamp ageing and luminaire contamination. All three functions (presence, illuminance and IR remote control) are integrated in a single multidetector (Fig. 9).

of the presence detector of the controller. And in this case the “staircase” function can be involved: activa-tion of the switch is perceived as presence of a per-son, and illumination is turned on. After the set time delay expires, illumination luminance decreases until full switch-off (turn-off mode) or to a level of 10 % of the rated one (corridor mode).

5. ILLUMINANCE DETECTOR

Now we will consider the use of the illuminance detector in more detail. The most simple version is the application with each luminaire group of the illuminance detector directed downwards on to the working surface (Fig. 5), for example the surface of a table, for illuminance level to be detected.

As this takes place, illuminance maintenance is ensured at a required level regardless of the lamp ageing and luminaire contamination. The solution is ideally suited for rooms of a big size separated

Fig. 5. Illuminance detector directed downwards

Fig. 6. A room with three multidetectors

Fig. 7. “Tracking” photocell

Fig. 8. A room with a “tracking” photocell


63

Requirements and wishes: illuminance level in every area (zone) adapts to the activity types; electric power saving; simplicity of installation and commissioning without involving specialists and without necessity of their training.

Solution: multidetectors detect presence/absence of people and control luminance according to day-light level; individual hand luminance control for il-luminance level correction in a zone is carried out, if necessary using keyboard switches without lock-ing or using an IR remote control. IR remote control is also used for multidetectors to be related to spe-cifi c groups of luminaires (Fig. 11).

The mode of presence detection and delay time of the switching off are set using switches on the controller. Illuminance levels are programmed by means of IR remote control or keyboard switch without locking, temporarily connected to the LUX input of the controller.

When implementing the solution, a standard mains wiring, only nonpolar DSI/DALI and wires of the detectors are used.

7.2. Classroom

Main parameters: up to 25 users; area is less than 70 m2; activity type is lessons, break, cleaning; loca-tion of luminaires in rows, two groups, in parallel to the window; height of the ceiling is 3 m.

It should be noted that in this case materials and colours of the working surface infl uence the sys-tem’s operation. Recommendations of the standard EN 15193 for rooms of more than 30 m2 area are not taken into consideration.

6. PROGRAMS

What is the purpose of the illumination pro-grams? In reality, no room is used for only one type of activity or work. That is the reason why the Sen-saModular system allows a user setting illumina-tion programs and adapting illumination for various tasks. A built-in illumination control panel makes it possible to set and select three scenarios of illumi-nation, and also to disable them (Fig. 10).

The activated scenario is highlighted by a green indicator. Three pairs of buttons are used for lu-minance control of individual luminaire groups, and two additional buttons allow adjusting lumi-nance of all luminaires simultaneously. The panel of program choice in the standard version has mat-te steel surface treatment with an attractive bronze decoration.

One more advantage of the SensaModular sys-tem is the function of automatic determination of the connected luminaire types in accordance with the DSI/DALI standards, and the correspondent adjust-ment of output control signals. Up to 50 luminaires can be connected to each controller output, if they are operated according to the DSI protocol, or up to 25 luminaires operated according to the DALI pro-tocol. If necessary, number of the connected lumi-naires can be increased by connecting additional am-plifi ers of control signals.

7. TYPICAL SOLUTIONS

As typical examples of illumination control sys-tem implementation, we consider illumination of a group offi ce of a graphic design agency and of an elementary school classroom.

7.1. Offi ce

Key parameters: up to 15 users; less than 100 m2 area; activity type is work at a personal computer, work at a table, break, cleaning; location of the lu-minaires is by three zones, with a group of the lu-minaires built in the ceiling over each group of the tables; height of the ceiling is 2.8 m.

Fig. 9. A room with one multidetector

Fig. 10. Control panel of the SensaModular illumination system


64

locking “opened using a key”, which are intended to correct illumination levels, if necessary (teach-ers, the household manager, cleaners have the keys) (Fig. 12).

The mode of presence detection and delay time of switching off are set using switches on the con-troller; illuminance levels are programmed by means of a press button mounted in the photocell, or using a keyboard switch without locking, which is tempo-rarily connected to the LUX input of the controller.

Requirements and wishes: electric power sav-ing; stability against vandals, additional protection; simplicity of use; simplicity of installation and com-missioning without involving specialists and without necessity of training.

Solution: daylight detector directed to a window for all groups of luminaires; wall passive infra-red detector or presence detector for all groups of lumi-naires; hand switching off and luminance control using standard double keyboard switches without

Fig. 11. Illumination of an offi ce

Fig. 12. Illumination of a classroom


65

It should be noted that the cost of this control system is lower than the cost of other European and Russian analogues.

9. SHORT TERM PROSPECTS

Already today it is clear that in the near future, not a single project of illumination will do without an integrated control system. By analogy with the fact that many manufacturers of illumination equip-ment, especially of offi ce equipment, have already refused production of luminaires with out-of-date electromagnetic ballasts and completely shifted to electronic ballasts, we expect a gradual changeover from luminaires without a control system to the lu-minaires with an integrated interface (driver) to be used as a part of the control system. It is obvious that one should move in this direction “from simple to complex”, as much as possible using the gained experience and projects, in which reliable, easily in-stalled, adjusted and operated equipment is applied.

We are sure that this equipment will help design-er engineers, light engineers, control system special-ists and “smart house and offi ce” specialists to use energy saving and energy effi ciency potential in il-lumination systems of various applications.

Similar to the previous example, when imple-menting the solution, a standard mains wiring, non-polar DSI/DALI and wires of the detectors are only used.

8. IMPLEMENTED PROJECTS

The SensaModular control system of illumina-tion is installed at the Inter-parliamentary assem-bly building, in St.-Petersburg, in the assembly and exhibition halls (Fig. 13). The system operates with groups of the Thorn BaseLED light-emitting diode luminaires of light directed downwards, and with the Thorn Shalice 2 x18 W, Titus 2 x58 W and West-minister 1 x36 W luminaires. The system is installed by specialists of ТТК and Arkhilait companies, and since 2009 has given a good account of itself as highly reliable, simple in maintenance and use.

All of the Thorn luminaires used are completed with DALI drivers with dimming function. The con-trol system operates with any light sources and con-trollers from any manufacturers compatible with the DSI/DALI standards.

Fig. 13. The assembly hall of the Inter-parliamentary as-sembly, St.-Petersburg

Julia B. Babanova, an engineer, graduated from the Department of Cosmonautics of the Moscow Aviation Institute in 1992. At present, Julia Babanova is the Regional Director of the Thorn Lighting representative offi ce for Russia and CIS in Russia

Vadim A. Lunchev, a physicist, graduated from the Physical Department of the Moscow State University of M.V. Lomonosov in 1984. Now he is the Head of the Thorn Lighting representative offi ce in Russia

66

evaluation of road lighting energy and environmen-tal performance.

2. GRAPHICAL OVERVIEW

The assessments quantitatively appraise key per-formance parameters of lighting infrastructure con-struction and operation, Fig. 1:

• Economic Impact – Whole-of-life cost and in-vestment Net Present Value;

• Energy Impact – Operational energy;• Environmental – Carbon pollution and mer-

cury waste.The equipment technologies accommodated

include:• All relevant lamping options (both current and

legacy benchmarking options);• HID, CFL, LED luminaires;• Magnetic & electronic control gear;• Step-dim controls, Central Management

Systems;• Microgeneration power systems (solar, wind

and hybrid net-zero systems).The process begins with road lighting design sce-

narios developed separately via established lighting design good practice in accordance with Standards based regulatory requirements.

The major system performance determinants, lu-minaire gross wattage and column spacing data, are inputted to the RULHAM model (along with other detail input data) for the assessment of a given light-ing scenario.

The RULHAM software calculates life cycle im-pacts and enables “what-if” comparisons to be con-

ABSTRACT

1This paper presents the results of a four year re-search project at Massey University – Centre for En-ergy Research – Auckland, New Zealand. The out-come is termed a “Road and Urban Lighting Holis-tic Assessment Model” (R.U.L.H.A.M.), a software based modelling tool that facilitates the structured and comparative performance assessment of in-stalled road lighting design and technology scenar-ios using transparent and auditable standards based methodologies. The goal was to develop a practically applicable planning tool for lighting designers, in-vestment analysts and infrastructure asset managers.

Keywords: a road and urban lighting holistic as-sessment model, software, standard, planning tool, from cradle to grave

1. INTRODUCTION

RULHAM is a computer based assessment mod-el to calculate the economic and environmental per-formance of road lighting systems to assess and compare holistic performance over whole-of life, from cradle to grave.

The model is currently available for New Zealand and Australian road lighting schemes.

RULHAM software has been incorporated into the New Zealand Electricity Commission “Right-Light” Efficient Road Lighting Programme as a Government endorsed methodology for the holistic

* On basis of the paper presented ai the PLDC Conference, Madrid, October 20, 2011

Light & Engineering SvetotekhnikaVol. 20, No. 1, pp. 66-70, 2012 No. 2, 2011

DEVELOPMENT OF A ROAD AND URBAN LIGHTING HOLISTIC ASSESSMENT MODEL*

Bryan King

Lighting Management Consultants Ltd, Auckland, New Zealand E-mail: [email protected]


67

• LAMPS (Improved metal halide, improved HPS, CFL, in-

duction, plasma, SSL/LED);• LUMINAIRES (High IP, high photometric performance, high

optical control, long life, self clean optics, upgrade-able LED optics);

• CONTROLS (Electronic gear, step-dim controls, full-dim con-

trols, Central Management Systems);• COLUMNS (Passive safe, recycled materials, ground

sockets);• POWER SYSTEMS (Solar PV, wind, hybrid net-zero systems).RULHAM is structured to accommodate the

many new and upgraded Standards and guidelines the world of public lighting and the built environ-ment, such as:

• AS/NZS1158 – Lighting for Roads and Public Spaces (EN13201, IESNA RP-8);

• IESNA LM79/LM80 – LED Performance Measurement;

• ISO14040 – Life Cycle Assessment;• ISO14025 – Environmental Product Declara-

tions;• ISO15686–5 – Life Cycle Cost: Buildings &

Constructed Assets;• BSI PAS 2050 – Lifecycle GHG Emissions:

Goods and Services;• BS EN 16001 – Energy Management Systems;• ISO 50001 – Energy Management Systems….

coming soon.

veniently explored, to work towards optimised solu-tions on an iterative basis.

The model allows the quantifi cation and identi-fi cation of “best-fi t” improved design and technol-ogy options for Local Councils and Road Control-ling Authorities.

It underpins this with rigorous economic, energy and environmental assessment to assist with real-world investment analysis and procurement deci-sion making.

ISO Standards based, whole-of-life, cradle-to-grave, analytical methods have been packaged to explore and evaluate the performance of road light-ing scenarios and to compare “Best Available Tech-nology” (BAT), with existing or “Business as Usual” (BAU) approaches.

3. WHY THE QUANTITATIVE ASSESSMENT IS NEEDED

There are many new lighting technologies now available, lamps, light sources, luminaires, control systems, power systems.

There have been many new or modifi ed Stand-ards, guidelines and approaches introduced on light-ing design, lighting and energy performance, Life Cycle Management, Life Cycle Assessment, Life Cycle Inventory and Life Cycle Costing.

The question is: how do we effectively assess and compare the various and different, design tech-niques, technologies and approaches?

RULHAM can accommodate a wide range of new and traditional technologies, such as:

Fig. 1.


68

For a road lighting system, this may be……“1 km of roadway lighting to AS/NZS1158

Category P3 over whole-of-life” The Functional Unit encompasses the quantity

of construction to be measured, the lighting design standards conformance required, the type of stand-ards classifi cation and the length of project life.

7. THE RULHAM MODEL

The model has been constructed to capture all of the relevant cost, energy and environmental in-puts into each category at each stage of life. The be-haviour of the various separate elements is modelled in detail over the selected project life of up to 50 years (typically 20–30 years for roadlighting). All el-ements are then calculated and aggregated to form a quantifi ed total picture of the performance of the se-lected project. The following lighting scheme design stage outputs are entered into the software. These are determined by separate (conventional) lighting de-sign processes and calculations:

• Column spacing;• Number of luminaires per column;• Construction works duration;• Roadway cable crossings;• Operating hours;• System duty cycle (dimming uptime);• Microgeneration power input;• Luminaire power factor;• Lamp/luminaire type selection;• Luminaire gross wattage.The above parameters combine to capture the

“capital density”, the “materials density” and the “energy density” per kilometre of roadway, for each lighting scheme.

8. ECONOMIC IMPACT

The economic impact of road lighting systems consists of the combination of capital costs, oper-ating costs and end-of-life costs. These can be ex-pressed simply as an upfront lump-sum capital cost, annual operating costs, plus the cost of site restitu-tion and the conclusion of the functional life. How-ever, for investment analysis and improved deploy-ment of the asset manager’s budget, more sophis-ticated methods are required. Such methods must recognise the time value of money, the required rate of return and package the calculation result in a way, which facilitates effective decisionmaking.

Additionally, regulatory authorities such as High-ways Agencies Procurement Requirements often mandate that whole-of-life multiple bottom line ap-proaches be adopted in order to qualify for govern-ment funding subsidies.

The RULHAM model is geared around the re-quirements of the New Zealand “NZTA – Procure-ment Manual 2009”. This is a “new age” document refl ecting international best practice in the determi-nation of “Best Value” procurement.

4. BASIC PRINCIPLES

What are the basic principles of road lighting evaluation?

Road and urban lighting schemes are systems.They need to be assessed as a system, assessed

over whole of life and assessed to recognised standards.

There are four main impact categories that drive the economic energy and environmental perform-ance of road lighting systems, such as:

Economic Impact – Net Present Value over life – $ NPV;

Energy Impact – Energy consumed over life – kWh;

Carbon Impact – Emissions over life – Tonnes CO2;

Waste Impact – Mercury waste over life – mg Hg.Over the 3 phases of project life: The Embodied

or Capital Phase, the Operational Phase and the End-of-Life Phase.

5. SOFTWARE STRUCTURE

The RULHAM software is written in a combina-tion of C++ and Visual Basic programming languag-es. This allows a custom tailored user interface and supports a multi-choice user option, prompting/selec-tion operation with an internal product technology at-tribute and cost database. An important aspect of the model is to use a user friendly format, which conceals database and calculation functions behind the scenes.

6. THE HOLISTIC METRIC – THE ISO 14040 FUNCTIONAL UNITS

Life Cycle Assessment to ISO 14040 requires de-termination of “The Functional Unit”

This is the self-determined unit of measure for a designed system within the built environment.


69

– Maintenance vehicle energy – petrol/diesel fuel.

A particular challenge was the ability to accom-modate adaptive lighting and dimming control ac-tivities. This has been handled with a “duty cycle” calculation, leading to a “percentage of uptime” fi -gure that represents the net outcome of a certain dim-ming profi le.

10. CARBON IMPACT

The carbon impact aspect of the model is a refl ec-tion of the operating energy activities (as above) of a chosen lighting scenario.

CO2/kWh conversion factors for purchased elec-tricity are obtained from offi cial government sourc-es, as appropriate for the generation mix. Micro-generated electricity of course has no operating carbon impact. The operating fuel carbon impact of installation, service and maintenance vehicles is accommodated.This includes both km of trans-port use plus hours of operation for lifting/access equipment use.

11. MERCURY IMPACT

The most tangible and high profi le environmental toxin from lighting systems is generally regarded as mercury. This comprises direct mercury waste and indirect mercury emissions from power generation by coal fi red power stations.

As the New Zealand electricity generation is around 80 % from renewable sources the emitted mercury calculation has been omitted, but the model can easily be adapted to accommodate indirect mer-cury emissions.

The embodied mercury of the initial installa-tion is a simple calculation of the quantity lamps per km of roadway times the mercury content of the lamp/ light source concerned.

The operating mercury calculation is the same as the above but factors in the number of lamp replace-ment intervals over the life of the project, plus safe disposal of the fi nal cycle.

12. COMPARATIVE REPORTING

The graphical Comparative Report generated is a singular bar-chart report that delivers an over-view of the performance of all modelled scenarios (up to six).

The RULHAM model uses a Net Present Value (NPV) calculation model to express the lifetime. NPV of the various scenarios under evaluation with a variable investment return hurdle rate is selected. Capital costs include design, management, equip-ment, installation, commissioning and financing costs. Operating costs include electricity, electricity distribution, maintenance labour, consumables and replacements. These costs could be offset against microgenerated electricity feed-in fi nancial credits (if applicable).

End of life costs comprise deconstruction, dis-posal, reuse, recycle and reinstatement.

The model summates the initial capital costs, built up from an internal database of equipment, ma-terials and works costs. The annual and cyclical costs (i.e. relamping) are plotted on an annual cash fl ow basis (from the internal cost database) with a bal-loon cost at the end of project for site restitution and material disposal/recycling. All of these cost drivers form the inputs to the NPV calculation algorithm, which provides the NPV value for each investment scenario. The objective is to determine the lowest lifetime project NPV (actually a Net Present Cost) whilst delivering a suitable lighting result.

9. ENERGY IMPACT

The embodied energy aspects of lighting infra-structure are not currently addressed by the RUL-HAM model. It was considered that this was too ad-vanced to be incorporated into a practical everyday working tool, at this point. It is however, intended that in the near future an extension be provided to assemble an embodied energy and carbon quantifi ca-tion process, using a framework such as PAS 2050. The adoption of materials LCI databases such as ILCD and taking the assessment a pragmatic 90/10 level of detail should be workable.

With the rise in popularity of off-grid power sys-tems for road lighting (with no operational ener-gy) such embodied impact assessment processes should prove effective in the evaluation of design and equipment options.

Operating energy quantification includes all of the following aspects:

– Lighting energy density/km of road;– Lamp wattage;– Control gear – watts loss, power factor;– Dimming capability;– Microgeneration contribution;


70

ed and confi gured in accordance with US and/or EU prevailing requirements and cost structures.

Indications of interest in collaboration in this task would be most welcome.

Additionally, embodied and end-of-life energy and environmental impact calculation phases are to be developed and included in the RULHAM model.

From the fi eld of environmental engineering, ma-terials and product Life Cycle Inventories (LCI’s) and ISO 14025 standards based Environmental Prod-uct Descriptions (EPD’s) are rapidly evolving and expanding. These resources and processes house the information raw materials to enable practical product embodied energy data to be compiled and calcula-tions undertaken. This is particularly important for the realistic performance assessment of microgen-eration, and hybrid power net-zero lighting systems which side-step the conventional purchased electric-ity operational energy phase.

The tabular Scheme Assessment Reports are in-dividual reports covering the detailed outputs each modelled scenario (up to six), including:

• Economic Report – Capital cost, annual cash fl ows, investment Net Present Value;

• Energy Assessment Report – Life cycle energy, including both purchased and generated electricity plus distribution line losses;

• Carbon Assessment Report – Life cycle car-bon, including impacts from both purchased and generated electricity plus distribution line losses;

• Mercury Waste Assessment Report – Life cy-cle mercury waste.

13. OUTCOMES FROM QUANTITATIVE MODELLING

• To CALCULATE… the multifaceted perform-ance of various options;

• To COMPARE… the various options (BAU vs various possibilities);

• To IDENTIFY… the best value option;• To COMMUNICATE… the best value option.The RULHAM model is useful for:• Designers, to make more informed technol-

ogy and design choices and to better understand the wider consequences of design choices and equip-ment selection;

• Financiers, to evaluate competing investment proposals and rank and the rate fi nancial outcomes, particularly those that purport to deliver energy and fi nancial savings;

• Asset Managers, to target investment propos-als and gain improved return on investment;

• Policymakers, to support higher performing options that are aligned with Government objectives.

14. CONCLUSIONS

“If you can’t measure it… you can’t manage it!” This time honoured adage holds true to this day.

Substantive road lighting performance improve-ment in practice are unlikely to be achieved unless quantifi able economic, energy and environmental gains can be measured and justifi ed in the real world. RULHAM is a practical tool designed to assist in de-livering on this challenge.

15. FUTURE DEVELOPMENTS

The RULHAM model is currently geared to New Zealand cost structures and to Australia/New Zea-land standards practices, the model could be expand-

Bryan Kingis the Director of Lighting Management Consultants Ltd., New Zeeland. A me-chanical engineer, with a Diploma in Business and Industrial Administration and a Master of Business Administration from the

University of Auckland, Bryan is currently completing a Master of Technology in Energy Management at Massey University Centre for Energy Research. Bryan was founding Chairman of Lighting Council NZ, is a member of the Illuminating Engineering Society of ANZ, a member of the Life Cycle Association of NZ and the Energy Management Association of NZ. He is currently on the AS/NZS1158 Public Lighting Standards Committee and has been an APEC NZ CFL Standards delegate, a member of the Ministry for the Environment Lighting Product Stewardship Steering Group and the Electricity Commission Effi cient Lighting Advisory Panel. Bryan is a fre-quent visitor to international lighting conferenc-es, seminars and trade fairs, is a regular presenter on lighting, energy and management matters at NZ conferences and has been a guest lecturer at the University of Auckland, Auckland University of Technology and Massey University

71

Carl GardnerTackling Unwanted Light: An International Perspective

Figs. 2, 3 and 4. Examples of poorly designed, polluting lighting schemes

Fig. 14. Slovenia – some of the best anti-light pollu-tion legislation in the world

Fig.18. Shanghai’s polluting lighting culture has not been con-trolled by the new Lighting Standard

Fig.20. New York’s ‘energy density’ approach to lighting does not control the worst excesses of unwanted light

72

Roger NarboniThe Old City of Jerusalem Lighting Master Plan

Fig. 1. Jerusalem

Fig. 6. Damascus Gate

Fig. 8. Jerusalem LNP nocturnal densities map

73

Roger NarboniThe Old City of Jerusalem Lighting Master Plan

Fig. 9. Muristan fountain: general night view and illumination of details

74

Nicolai N. Usov Prospective Use of Organic Light Emission Diodes for Information Displays and for Illumination

Fig. 1. Comparison of images on an OLED screen and on LC TV screen with back illumination using inorganic light emission diodes [8]

Fig. 2. A production site of 5.5 generation of Samsung

Mobile Displays Company

Fig. 7. A fl exible OLED display of Sony Company

75

This paper is a survey of the best LED suppliers in Norway, and their ability to deliver “acceptable” road lighting installations. In cooperation with the Norwegian Public Road Administration, depart-ment of Buskerud, an installation of 18 road poles was made available for the study. This installa-tion was divided into three sections, each equipped with six luminaries from three different suppliers. An invitation was sent to a broad selection of LED suppliers (12) with a description of the installation and a request to give an offer of their best solution, this was done in June-July 2010. The best three was chosen after a theoretical evaluation and installed at the test location in September-October 2010. Af-terwards both a fi eld study and a lab study were conducted on the installations performance. Both electrical and photometric parameters was meas-ured and compared to the supplier’s initial claims (rated) and data sheets. The lab measurements were done in the period from September-October 2010, while the fi eld measurements were done in Decem-ber 2010.

2. METHOD

Using several measures this study investigates the LED-luminaires performance compared to the supplier’s own data sheets and lighting calculations.

• Lumen output linked with electrical power were investigated in lab.

• Electrical parameters, both during switch-on and during continuous operation (stabile values) were investigated in lab.

• Levels and uniformity based on illuminance were investigated in fi eld.

• Levels and uniformity based on luminance were investigated in fi eld.

A BSTRACT

12 LED suppliers were invited to provide their best solution for a given geometry of road. After a theoretical evaluation of lighting calculations, the best three were chosen. The selected luminaires were tested according to information given from the sup-pliers and relevant standards, and guidelines, con-cerning electrical and lighting performances both in laboratory and in fi eld. The study shows diver-gence between rated values and measured values of both electrical and photometric parameters. Espe-cially poor electrical performances during dimming were found. The paper also discusses and compares the luminance yield of LED with the traditional in-stallation of high pressure sodium lamps with elec-tronic ballast.

Keywords: LED, road lighting, luminous effi -cacy, compliance

1. INTRODUCTION

During the last decade an increasing focus on energy effi ciency has led to a global initiative to re-duce use of electrical energy in all public and private sectors. As an example we have the “20 20 by 2020” Europe’s climate change opportunity (EU, 2008). Norway has close to 1.3 million road lighting lumi-naires, representing an annual electrical consump-tion of about 0.8–0.9 TW×h (Enova, 2010). Related to the development within LED technology, LED is assumed to take a large market share in the near future on outdoor lighting, both within refurbishment and new installations. In this context it is important to be aware of the actual performances concerning both lighting-quality and other more general quality parameters that the technology delivers.

USE OF LED FOR ROAD LIGHTING

Pål J. Larsen

Norwegian Scientifi c and Technical University, Trondheim E-mail: [email protected]



76

2.2.1. Illuminance measurements in fi eld

Measurements were done with a calibrated hand-held luxmeter type HagnerLux meter EC1, according to procedures described in EN-13201–4 (CEN, 2003 b). As can be seen in Fig. 2 there are some surround-ing buildings adjacent to the measurement area, but the illuminance contribution are assumed to be negli-gible. Hence measurements are assumed to be made without infl uence of surrounding light.

Registered values were the next: Eavg, Emax, Emin, Uo, Ulill,, SR

As described in EN 13201–3 (CEN, 2003 a) a grid of 6 x10 measuring points were established on the road surface. Illuminances were measured at each point. Eavg were calculated as the mean of all these points, Emax and Emin is the respectively high-est and lowest value presented in the grid points. Uo were calculated based on the Eavg/Emin value derived from measurements. Ul based on illumi-nance is not a value that is necessary in road light-ing, therefore this value is not discussed in the paper but it is presented in the table for comparison. This is derived from the centre line of each lane (grid row 2 and 5) where the average of all points is divided by the minimum value. For the SR value an addi-tional grid row were established 3 meters outside the edge of the road and 3 meters inside the edge on each side, and 10 measuring points were established in each row. The average of the 20 points outside the road was divided by the average of the 20 points in-

2.1. Field experiment location

The road (FV283 Drammen-Åserud) is a public county road between the city of Drammen and the suburb of Mjøndalen in the county of Buskerud, 50 km south-west of Oslo.

The chosen parcel of the road is a straight leg of 700 m. There is two-way traffi c with 1 m central separation; each lane is 3.25 m wide [Figs. 1, 2]. 10 meter high poles are placed at an individual distance of 33 m, 3.5 m from the edge of the road (each lumi-naire is mounted with a 2 m bracket). The road has the possibility of diversion of traffi c at night time, so it was possible to do extensive measurements on the road surface without traffi c.

2.2. Experiment setup

The fi eld study investigates performances based on photopic illuminance and luminance. The lab study investigates both electrical and photometric parameters.

Fig. 1. Cross section of test road

Fig. 2. Air photo of the selected leg of FV283


77

of switch on, approximately 5 repeated measure-ments were logged with the oscilloscope. The larg-est value is presented in the paper. For the “time for stabile current”, the time from switch on to the point amplitude of the current approaches the nominal am-plitude is recorded and presented.

2.3. Constriction

The fi eld measurements were done while there was snow in the terrain. The surroundings refl ect-ed light-contribution to the illuminance on the road has been neglected. Luminance measurements were conducted at a time there was no snow on the road, but the temperature was around –15C. Even though the surface was free of snow, the temperature and humidity of the air infl uenced the surface refl ective properties visibly, hence the evaluation according to the design criteria’s are done based on measured il-luminance values.

The age of the lighting luminaire and burning hours of light source has not been recorded, but are in the region of 100–200 hours. The wear of the surface, or corresponding properties, has not been recorded.

Field measurements of illuminance and lumi-nance have been done only one time, as an instant measurement. The ambient temperature was regis-tered to be –15 C at time of measurements. No com-pensation has been done to justify the comparison to the lab values measured at an average ambient tem-perature of 22 C.

This study investigates luminance yield related and compared to the current international standards for road lighting. It does not consider different visu-al conditions e.g. mesopic/photopic “compensation” of luminance (CIE, 2010) or target visibility.

The study does not emphasize an evaluation on cost over lifetime. Some parameters are included on cost, but these are only based on the installation and are thereby not representative to compare systems for a LCC analysis.

2.4. Experiment hypothesis

Prestudies have shown divergences between the rated luminance yield from LED installations and the actual yield in real life. The prestudies lead to the hypothesis:

• Cos phi is lower than 0,9 when installation is operated as dimmed;

side the road, and the ratios of this were presented as the Surrounding Ratios.

2.2.2. Luminous fl ux measurements in lab

All luminous fl ux measurements were done in an integrating sphere. Ambient temperature was regis-tered to be approximately 22 °C (3 degrees off the rated ambient temperature to measure LED lumi-naires which is 25 °C (IES, 2008)), but this is as-sumed to be negligible in this comparison study (see section 2.3 constriction). All levels of measurements were recorded after a time for stabilization, approxi-mately 5 minutes.

2.2.3. Luminance measurements in fi eld

Luminance measurements were done with a cali-brated Canon EOS 350 from TechnoTeam, processed in Techno Team’s LMK 2000 Mobile Advanced soft-ware. This measurement setup is assumed to be cat-egorised as accuracy class B referred to DIN stand-ard 5032–7 (DIN, 1985). Camera was mounted on a tripod and measurements were done according to procedures described in EN-13201–4 (CEN, 2003).

Registered values: Lavg, Uo and UlThe same way as illuminance measurements are

done, the luminance measurements are based on a standardised measurement grid of 6*10 measuring points described in EN 13201–3 (CEN, 2003 a). Uo and Ul are derived the same way as Uo and Ul for the illuminance based paramteres, hence based on lumi-nance. The Lavg though are taken directly from the “Mobile Advance software“, which derive the aver-age value of the whole measurement area.

2.2.4. Electrical measurements in lab

To register the current inrush during switch-on, an oscilloscope type Textronix TDS was used. To log the voltage, current and Cos phi of the individual luminaries, a calibrated power quality analyser type Fluke 43 was used. Also the THD (Total Harmonic Distortion) represented by the accumulated harmon-ics up to the 50th was registered by the Fluke 43.

All power supplies in lab were done by a stabi-lised source at 230 V. All measurements were done as system measurements, including consumption for the whole fi xture.

As the electrical inrush amplitude varies with the position of the sinus curve for the voltage at time


78

3.1.2. Luminous effi cacy – measured in lab

Supplier 1 and 2 delivered 1–10 V dimmable bal-lasts while supplier 3 delivered only static output ballasts. Luminous effi cacy, expressed by their re-spective lm/W ratio with an input voltage are shown in range (10–1) V in Fig. 4. Note that supplier 3 are only indicated at 10 V (100 %) in the diagram.

3.1.3. Lighting performance in fi eld

Field measurements were taken for illuminance and luminance. The “conventional solution” is based on a 150 W high pressure sodium lamp’s luminaires. This is a theoretical comparison that is only calcu-lated and not measured.

Results are listed for three categories:• “Calc from supplier” – values from lighting

calculation original from suppliers;• “Calc based on datasheet” – values from light-

ing calculation based on values from datasheets pro-vided by supplier;

• “Measured” – Measured values in fi eld that are multiplied with the given maintenance factor (0,8) from the calculations to be comparable to lighting calculations.

All measurements are done after approximately 100–200 hours burning time. The demand for lon-gitudinal uniformity are given based on luminance, the Ul-value based on illuminance are listed only for relative comparison. All measurements are assumed to be stabilised values.

Based on the measurements presented in Tables 1 and 2 certain parameters were derived. As the con-striction chapter explains this paper will not discuss LCC-calculations or lifetime of LED, yet a relative

• Installation generally perform as promised concerning levels of light;

• Installations generally perform poorer than rated (claimed from supplier) concerning uniformity of light, especially longitudinal uniformity;

• Some LED suppliers fail to comply to the reg-ulations concerning total harmonic distortion (shall be lower than 8 %, when registered over time accord-ing to EN 50160 (CEN, 2010));

• There are divergence between rated perform-ance (claimed from suppliers) and actual perform-ance concerning luminous fl ux [lm] and thereby the luminous effi cacy [lm/w].

3. RESULTS

3.1. Photometric parameters

3.3.1. Luminous fl ux – measured in lab

The process of selecting suppliers was done as a real project at the Public road Authority of Norway.

11 suppliers delivered an offer (out of the 12 invited), containing a lighting calculation based on the described geometry of the road (see Fig. 1) and basic rated values of their luminaires. The best three were chosen based on an evaluation of their delivered lighting calculations and rated perform-ance. The graph shows these initial values as “Val-ues from lighting calculations”. When they were informed of being chosen to deliver 7 luminaires, they were asked once again to provide information on the luminaires they were about to deliver (shown in Fig. 3 as “Values from data sheet”). These values are in Fig. 3 compared to the luminous fl ux measured in lab (shown in Fig. 3 as “Values measured in lab”).

Fig. 3. Luminaires luminous fl ux from different issues

Note: Supplier 1 did not give any additional datasheets upon request; therefore the column “values from datasheet” are not presented.

Fig. 4. Luminaires luminous effi cacy


79

urement on the conventional luminaire gave a fac-tor of 6,6.

The time from switch on to the luminaires reached a stable current wave, was fairly short. All three suppliers had reached stable values within 2 seconds, and the fastest were stabilized after ap-proximately 0,1 second. The conventional luminaire are not shown in Fig. 6, but stabilized in the range of 3–3,5 minutes.

3.2.2. Power factor

Of the three luminaires, supplier 1 and 2 had a power factor above 0,9 at 100 % power. Of the two dimmable luminaires, supplier 2 held the power factor above 0,9 until 20 % power, while supplier 1

comparison valid for this investigation is included showing the cost per lumen from the luminaire and the cost per achieved cd/ m2 on the road surface only using investment costs of the luminaire. Comparison of these two factors gives an indication of the rela-tive optimization of the different luminaires. This is shown in Table 3.

3.2. Electrical

3.2.1. Inrush current

As shown in Fig. 5 supplier 1 had about 11 times the nominal current at time of switch on, while sup-plier 2 and 3 had about 35 times the nominal cur-rent the fi rst cycle of the 50 Hz. Comparable meas-

Table 1. Calculated and measured luminance’s fi eld

Table 2. Calculated and measured illuminance’s fi eld


80

cy). Based on this it is strange to see such high vari-ance between the luminous fl ux used by the suppli-ers in their lighting calculations and what their own datasheets states. Regarding rated luminous flux (taken from suppliers initial lighting calculations) to actual measured luminous fl ux one can both expect and accept a certain divergence, based on the the-ory of LED and the diffi culty of supplying a large number of identical luminaires. In our test of three luminaires, supplier 2 had a divergence of 6 %, which is acceptable. The suppliers 1 and 3 had respectively divergence of 25 and 18 %. Basically one can argue they do not deliver the product they claim.

But it should also be noted that a large infl uenc-ing factor for the diverging luminous fl ux was the fact that they also delivered lower wattage than rated (respectively 11,2 and 4,7 %).

4.1.2. Luminous effi cacy

Considering the lm/W ratio gives a good indica-tion on the luminaires ability to deliver good light at a low electrical consumption, which is one of the qualities one strives for in an electrical roadway in-stallation. For the traditional high pressure sodium lamp, the effi ciency increases as the wattage rises. Typically for a 150 W lamps the effi ciency is just above 100 lm/w, while for the 250 W lamps it in-

went below 0.9 already at 60 % output. Comparable measurements have also been done for the 150 W electronically ballasted high pressure sodium lu-minaire. This luminaire started with a power factor at 0.99 at 100 % and sank to approximately 0.95 at 20 % output.

3.2.3. Total harmonic distortion

Tests were done as a “stabile instant measure-ment”, and showed variations from 9–15 % rela-tive harmonics component at 100 % output. Of the 2 dimmable luminaires, supplier 2 had below 10 % harmonics component until 50 % output, while sup-plier 1 increased to above 20 % already at 60 % out-put. Note that the relative components are related to the total dimmed value at each level, not to the total value at 100 %.

4. DISCUSSIONS

4.1. Photometric parameters

4.1.1. Luminous fl ux of the luminaires

The luminous fl ux of the luminaires linked with the electrical consumption, is the most used criteria for the luminaires performance (luminous effi ca-

Fig. 5. Inrush/Nominal Fig. 6. Time before stable current

Table 3. Energy performance and cost calculations

Note: 1 Kr = 0,13 Euro (24.2.2011). Defi ned m2 is the road surface area between two poles.


81

and the “worst” luminaire, supplier 1, had 15 %. A normal demand for a tender is that the products are to “fulfi l all relevant demands in all normal operation levels”. Therefore this paper also presents values for the products driven in dimmed condition. The 2 dim-mable luminaires showed an increasing harmonic component when dimmed (in % related to the dimmed total). Supplier 2 were fairly compliant and decreas-ing slowly, while supplier 1 must be said to fail to de-liver as requested, by rising above 20 % at 60 % pow-er and reaching 30 % just below 30 % power!

4.1.4. Lighting quality performance – measured in fi eld

As the demands for a road installation in Norway normally are given in the MEW lighting class, lumi-nance is the correct way to evaluate a road installa-tion. Due to the circumstances mentioned in section “2.3 – constriction” this paper is based on illumi-nance measurements.

Eavg All suppliers fulfi l the requirements regarding

the level of light (Eavg). So the amount of light that reaches the surface is suffi cient to fulfi l the demands.

Emax All suppliers have a higher maximum illumi-

nance than the calculations claims.Emin Supplier 1 and 2 performs as calculations claim.

Supplier 3 gives a higher Emin than assumed, this contributes to a better uniformity but does not infl u-ence the Eavg signifi cantly while the areas of the low illuminance are small.

Uniformity (UO and Ul)The high Emax does not influence the overall

uniformity due to the fact that this is only present

creases to about 120 lm/W. Note that this values de-scribes the light source not the whole luminaire. If the light output ratio of the conventional luminaire of 0.80 is included the “real” number is respectively 80 and 96 lm/W. Of our three test luminaries two had an lm/W ratio around 70. Supplier 3 however stood out and had a measured luminous effi cacy 92 lm/W. The project initial assumption was that the luminar-ies would deliver between 60–80 lm/W, which was the case for supplier 1 and 2.

The two dimmable luminaires behaviour were more or less expected. A slightly increasing effi -ciency is observed when luminaires are dimmed, un-til the level is low and the electronics in the drivers become an infl uencing factor, thereby the effi ciency slightly decreases again.

Considering the rapid development in effi ciency in the LED technology, one can expect a good devel-opment on these ratios in the near future.

4.1.3. Total harmonic distortion

For an electrical product to be legally distributed in Europe it has to be CE marked and thereby ful-fi lling all relevant demands. Concerning the currents harmonic components the EN 50160 (CEN, 2010) states a maximum limit for accumulated harmonics (up to 40) below 8 % as a mean value over 10 min, below 5 % as a mean value over 1 week. Our nation-al guideline NEK400 (NK64, 2010) gives correction factors for design of the electrical distribution if the THD factor is above 10 %. All of these demands are based on the stable values when the product is run at 100 %.

Tests were done as an instant measurement and not a logging over a long period. The test still gives a good indication of the level. Considering the demand of maximum 8 % none of the luminaires complied,

Fig. 7. Measured power factor component Fig. 8. Measured total harmonic distortion ( % THD)


82

tioning that if each side had been treated individu-ally, which might be more correct when poles are placed single sided, there would have been a more divided result. Table 4 shows the SR for the 3 sup-pliers divided into SRnear and SRfar, related to loca-tion of luminaires. The table shows that treated in-dividually supplier 1 and 3 has too low surrounding light at the far side of the road, with a ratio of 0,34 and 0,35.

4.1.5. Energy performance indicators

Through the last decade several energy perform-ance indicators has been used to evaluate road instal-lations. Fig. 9 shows some examples of comparing similar installations effi ciency using different ratios.

Luminous effi cacy [lm/w] is the initial indica-tor for the light source itself, stating how much light each energy unit produces. As we can see the tradi-tional high pressure sodium alternative still has the most effi cient ratio, but in this case it has to be noted that the [lm/w] ratio for the conventional luminaire presented in Fig. 9 represents the light source and not the total for the luminaire as it does for the LED luminaires. In Figs. 10 and 11 the datasheets stated L.O.R. (Light Output Ratio) of the conventional lu-minaire has been taken into account (for the investi-gated luminaire this factor is 0.80), and here we see a shift in the comparison.

As LED is a smaller light source than the tra-ditional, in theory it should be possible to direct the light more accurate. With the assumption that all necessary quality demands (uniformity, glare and SR) are fulfi lled, for the same geometry, one can evaluate ratios comparing the resulting illu-minance and luminance values for the investigat-ed road. In this paper [(W/ m2)/Lavg] and [(W/ m2)/Eavg] are presented. If for the four alternatives the best in each category represents the value 1 at the y-scale, a comparison of the different values will show us how they perform.

We can see in Fig. 9 that the high [lm/W] ratio for the conventional lamp is lost in the luminaire, and that LED luminaire 1 has a lower electrical con-sumption per luminance unit on the road. Luminaire 2 and 3 perform similarly based on luminance de-spite supplier 3 higher initial installed fl ux. Some-what surprisingly luminaire 3 perform relatively better when the evaluation is based on illuminance instead of luminance, so does the conventional lu-minaire compared to luminaire 1 and 2.

in small areas and don’t mark the total uniformity. All suppliers are better than the demand for Uo>0,4, but it is worth noticing that while supplier 3 perform identical to the calculated value, supplier 1 and 2 has an 16–17 % deviation from the calculation. If con-sidering a normal measuring deviation of 10 % the suppliers are still is 6–7 % off.

The longitudinal uniformity is affected by the high Emax values. Evaluating the longitudinal uni-formity based on illuminance (the correct evalua-tion is based on luminance according to the demands (Norwegian Public Road Authority, 2009) we see that supplier 1 and 2 (that has the highest maxi-mum illuminance points) has a poor longitudinal uniformity, while supplier 3 has a much better rela-tive longitudinal uniformity. As the demand refers to luminance it is not possible to compare this to the requirement, but one may assume supplier 1 and 2 might be below the demand based on these results.

SR All three suppliers deliver poorer SR than prom-

ised, but they all reach the minimum demand of 0,5. The SR-ratio as defi ned in EN-13201 part 3 and 4 (CEN, 2003), states that the ratio is the average join-ing together both sides of the road. It is worth men-

Fig. 10. Relative comparison of optics performance

Fig. 9. Energy performance indicators


83

circuit breakers minimal current. This dimensioning factor is not anymore important as our three test lu-minaires show, by using 0,1–2 seconds to reach their stabile nominal currents.

The other relevant electrical factor to consider at time of switch on, is the instant inrush the fi rst period of the 50 Hz. The instant inrush has tradi-tionally been solved by using slow characteristics on the circuit breakers, for the magnetically bal-lasted luminaire, which have been assumed to be 15–20 times the nominal current (for the electroni-cally driven ballast investigated in this paper a factor of six was found). For our test luminaires supplier 1 had an inrush current that normally can be handled by slow characteristics circuit breakers, while sup-plier 2 and 3 (found 35 times the nominal current at time of switch on) might represent a practical prob-lem in large installations. This might be an additional argument to use an advanced control system with ad-dressable luminaires giving the possibility to switch on the luminaires with a 2–3 seconds individual de-lay, instead of all at once.

4.2.3. Power factor

The Norwegian Public road authority gives a de-mand in their handbook for street and road lighting (Norwegian Public Road Authority, 2009) that the “light installation is to keep the Cos phi factor above 0.9”. For a dimmable installation it is also usually given a demand that the Cos phi should be above

Isolating the optics component, Fig. 10 presents the relative use of the installed lumen compared to the luminance on the road. For each category the lu-minaire with the highest value is set as the reference. Comparing the 3 individual LED suppliers one can derive some interesting facts. Supplier 2 and 3 has signifi cantly higher lumen packages than supplier 1. But supplier 1 uses the installed fl ux best and have the best ratio between installed fl ux and luminance level, but still at a lower level than supplier 2 due to lower initial fl ux. Hence supplier 1 has the best optics, in relative performance. So comparing the optics performance ranging from best to worst, we have supplier 1, 2, conv. and 3. Note that this com-parison assumes all other quality criteria’s to be ful-fi lled, and only compares the suppliers based on in-stalled lumen and achieved luminance on the road. This shows that LED already are competitive to the traditional solution in some aspects. But if one also look at the costs, as shown in table 3, both the “cost per lumen” and “cost per cd/ m2” shows that (at least for these high lumen packages) LED has a challenge to be competitive to the traditional solu-tions. A relative comparison on installation costs are shown in Fig. 11. One can argue the longer lifetime and thereby lower life cycle costs gives LED a lower maintenance and life cycle advantage. But still pre-sumed 3–4 times longer lifetime, are hard to argue when the cost of purchase are 3,5–6 times.

But it can be argued that as the LED luminous effi cacy is expected to increase the next years, the LED will reduce the gap to the traditional solutions also in this aspect. Also if the study had been done at a lower wattage level the gap would have been smaller. So for small streets (lighting class MEW4 or MEW5) alternatively a bicycle or pedestrian road (CE4 or CE5 lighting class) LED might already be a competitive solution.

4.2. Electrical

4.2.2. Inrush current

The inrush current is an important factor when designing an installation. The total inrushes when switching on a leg of installation can in worst case become the dimensioning factor for the installation. Traditionally the long time for the traditional mag-netically ballast driven luminaire to reach stabile current (twice the nominal current, decreasing for 2–5 minutes) has been a dimensioning factor for the

Table 4. SR-ratio

SR SRnear SRfar

Supplier 1 0,52 0,66 0,34

Supplier 2 0,72 0,63 0,83

Supplier 3 0,50 0,67 0,35

Fig. 11. Relative comparison of installation costs


84

tremely high cost for maintenance might diffuse this conclusion, but not as a general statement.

• Not fulfi lling what they claim concerning elec-trical and photometric parameters makes the markets attitude to the technology more negative. More cor-rect sales information would make the technologies reputation more serious for the market and would, for the long run, be advisable for the suppliers.

• It is concerning that the products do not com-ply fully to the relevant standards when it comes with electrical quality parameters such as Cos phi and total harmonic distortion. Especially the rapid degeneration when dimming is concerning.

• The defi nition of SR should be reconsidered. When poles are placed single sided in many cases the SR ratio on the far side, considering each side separately, are too low despite compliance to the “overall-SR” demand.

REFERENCES:

1. CEN. (2003 a). EN-13201–3 Road lighting part 3 Calculation of Performance. CEN.

2. CEN. (2003 b). EN-13201–4 Road lighting part 4 Methods of measuring light performance. CEN.

3. CEN. (2010). EN-50160 Voltage Characteristics of Electricity Supplied by Public Distribution Networks.CEN.

4. CIE. (2010).CIE 115 Lighting of Roads for motor and pedestrian traffi c.CIE.

5. CIE. (2010). CIE 191 Recommended System for Mesopic Photometry Based on Visual Performance. CIE.

6. DIN. (1985). DIN 5032–7 Lichtmessung, klasse-neinteilung von Beluchtungstärke und Leucgtdichtemess-geräten. DIN.

7. DSB. (2006). FEF – Forskriftomelektriskeforsyn-ingsanlegg. Standard Norge.

8. Enova. (2010). Veilys i Norge.Unpublished.9. EU. (2008, 1 23). 20 20 by 2020 Europe’s climate

change oppotunity. Downloaded 11 11, 2010 from EUR-Lex Access to Europen Union Law:

10. http://eur-lex.europa.eu/LexUriServ/LexUriServ.do? uri=COM:2008:0030: FIN: EN: PDF

IES. (2008). LM 79 approved Method: Electrical and Photometric Measurement of Solid-State Lighting Prod-ucts. IES.

11. NK64. (2010). NEK 400 – Electrical low voltage installations.Standard Norge. Norwegian Public Road Au-thority. (2009). Handbook 264: Prosjekteringavvegbelysn-ingsanlegg.Norwegian Public Road Authority.

0.9 “in all normal operating levels”. Yet neither the national handbook nor other relevant standards/reg-ulations gives a reference to keep the Cos Phi above 0,9 also at dimmed levels.

The fact that one of the suppliers delivers a prod-uct with an initial power factor of less than 0,9 is not good (Fig. 7). Also not good is supplier 1 rapidly decreasing power factor as it is dimmed. This factor are normally not given much consideration, because of the low impact on the design process (it does not contribute to the vital dimensioning factors), and it is not normally addressed from the suppliers side even though they have the right to claim compensa-tion if the factor is lower than 0.9 (DSB, 2006).

This is a criterion that is seldom verifi ed in op-eration, but considering the total harmonic distor-tion in LED installations (and also other road light installations with electronic gear), it might be use-ful for the future to give some regard to this. Also as this investigation shows that during dimming the power factor is, for some suppliers, rapidly decreas-ing it should be taken into consideration in the rel-evant electrical demands that the criterion are to be fulfi lled in “all relevant operating levels” not just at 100 %. These electrical disturbances are possible to compensate through the electrical control gear, so a better awareness and increased focus should result in a minimization of these factors.

5. CONCLUSIONS

• A road lighting luminaire needs a large in-stalled lumen output to reach the demands regarding lighting level and uniformity. Still the traditionally high pressure sodium lamp is the best solution con-sidering both cost of installation and performance at least for the bigger roads as is the case of this study. But isolating the optics component the best LED lu-minaires has lower optical losses than the conven-tional luminaire (high pressure sodium).

• To be competitive to the traditional installa-tion the LED luminaire has to become more effi cient (higher [lm/w] ratio) and especially lower the “cost per lumen” of installation. The best LED luminaire has a lower electrical consumption per “luminance unit” on the road (ignoring diverging achievements of other quantitative quality criteria’s in this study). Using LCC-calculations LED generally has a long-er lifetime than the traditional lamp but still they are too expensive to become a relevant competitor. In-corporating the replacement cost on roads with ex-


85

APPENDIX:

Pдl J. Larsen, postgraduate doctorate in way to get Ph.D. from Norwegian Scientifi c and Technical University of Trondheim in fi eld of new devices for illumination management and energy effi ciency in street’s and road’s lighting

Table 5. Conventional Lighting Details and Conventional Luminance

Note: 1 Kr = 0,13 Euro (24.2.2011).

86

to show that work performance would increase as a function of increased lighting [1]. The results failed to establish such a mechanistic relationship between lighting conditions and human responses. These studies prompted wide discussion on the effects of intervening psychological and psychosocial vari-ables, such as attitudes and work load. Further re-search has specifi ed a large number of mechanisms by which lighting quality is seen to affect perform-ance, well-being or comfort. These mechanisms can be roughly categorized into psychological process-es, such as perceived control over the environment and affective responses [2], and psychobiological factors, such as visibility, stimulation and circadian rhythm [3].

In order to develop lighting design guidelines, it is important to obtain evidence of the effects of lighting conditions in fi eld settings. Some fi eld studies have objectively measured work output in industrial workplaces. A review of these studies showed an increase in output and a decrease in re-jects, but the extent to which this could be attributed to illuminance was left unclear [4]. Moreover, most of the reviewed studies were old, with only two stud-ies conducted between 1980 and 1991. More recent fi eld studies in industrial environments [5,6] have demonstrated an increase in productivity for assem-bly workers during raised horizontal illuminance. However, the mechanisms behind these effects re-mained indistinct. Objective performance measures can also capture only a narrow range of the effects of lighting.

ABSTRACT

The aim of the study was to show that the im-provement of lighting conditions to meet EN 12464–1 standard is benefi cial for work performance and well-being. The improved lighting conditions were designed to meet the recommendations of EN 12464–1. Lighting measurements and questionnaire survey were conducted before and after a lighting renovation. Twenty workers completed both surveys. After the renovation, the average illuminances had at least doubled and provided a horizontal illuminance of at least 300 lx in the task areas. The uniformity of the lighting increased signifi cantly. The estimated energy consumption decreased by 60 %. Complaints about too little light and uneven lighting decreased. Satisfaction with lighting conditions as a whole im-proved. The need for extra light decreased. Fewer errors were reported to occur at work. Complaints about eye symptoms, tiredness, and motivational dif-fi culties decreased. The results indicate that uplifting the industrial lighting was benefi cial to the workers.

Keywords: lighting, intervention, work per-formance, productivity, environmental satisfaction, satisfaction

1. INTRODUCTION

The influence of lighting levels on work per-formance has been studied since the early 20th cen-tury. In fact, one of the fi rst longitudinal experi-ments in workplaces, the Hawthorne studies, aimed

LIGHTING IMPROVEMENT AND SUBJECTIVE WORKING CONDITIONS IN AN INDUSTRIAL WORKPLACE

Annu Haapakangas1, Jukka Keränen1, Marko Nyman2, Valtteri Hongisto1

1 Finnish Institute of Occupational Health, Indoor environment laboratory, Lemminkäisenkatu, Turku, Finland.2 KaNmanCo Ltd, Helsinki, Finland.

E-mail: [email protected]

Light & Engineering SvetotekhnikaVol. 20, No. 1, pp. 86-96, 2012 No. 3, 2012


87

of the surfaces surrounding the investigated worksta-tions were 0.3–0.5.

The visual demands of the work tasks were cus-tomary. The workers operated automatic plastic ex-trusion machines from remote control panels. They also collected the manufactured plastic parts from the machine. Part of the work consisted of the main-tenance of the machines. In some of the maintenance operations, the workers needed more light which was provided by working lamps. These lamps were not used during industrial production. One of the work-stations is presented in Fig. 1.

The lifetime of the old lighting system was end-ing and workers were complaining about inadequate lighting. The lighting system was outdated and some of the luminaires were no longer taken care of. Ac-cording to the production manager, most of the em-ployees (approximately 90 %) had a positive attitude towards the lighting renovation and the associated

Questionnaire measures can be used to assess perceptions of lighting and its subjective effects on health and performance. However, studies re-lying on self-reports have typically lacked objec-tive measurement of lighting conditions completely, e.g., [7], or have failed to demonstrate a relationship between objective conditions and subjective out-comes, e.g., [8]. Some longitudinal fi eld experiments have demonstrated a change in the perceived condi-tions after an environmental change. For example, Niemelä et al. [9] used a standardized indoor envi-ronment questionnaire that included environmental symptoms and psychosocial factors, but the effect of lighting could not be distinguished from the other concurrent environmental improvements.

Previous work indicates that direct measurement of work performance alone will not produce reliable answers about the effects of specifi c indoor environ-mental factors. The literature lacks a study where workers’ perceptions of a lighting improvement in an industrial workplace are exhaustively studied. Our research group has previously conducted interven-tion studies in offi ce environments in which the re-duction of offi ce noise has been investigated with subjective measures. The results of these experi-ments have been encouraging as the questionnaires have sensitively detected the changes in the per-ceived conditions and the reactions of the workers have been as expected, e.g., [10–12].

The purpose of this study is to evaluate perceived working conditions before and after a lighting ren-ovation. The study was conducted in an industri-al workplace, where an aged lighting system was renovated to meet current recommendations of the European standard EN 12464–1 [13]. The changes in subjective perceptions after meeting the recom-mended lighting conditions were studied. Evalua-tion of the psychosocial work environment was in-cluded. A rough examination of energy consumption was also conducted.

2. MATERIALS AND METHODS

2.1. Description of research site and the study

The intervention study was undertaken in a Finn-ish factory that manufactures plastic parts for vehicle industry. The interiors consisted of white brick walls, light gray concrete walls, light gray concrete fl oor, and gray concrete ceiling. The estimated refl ectance

Fig. 1. Luminance images of one of the workstations in the manufacturing plant before (top) and after (bottom) the lighting renovation, the luminance scale [cd/ m2] is pre-

sented below


88

No windows existed on wall surfaces. Small ceil-ing windows existed above the passage areas. Be-cause the room height was large and window density was small, sunlight did not affect illuminance levels in the task areas and did not cause glare problems. Therefore, this study concentrates on the effects of artifi cial lighting.

2.3. Subjects

Before the renovation, 26 workers completed the questionnaire and 29 workers after the renovation. All respondents worked in the factory hall and of-fi ce employees were excluded from this study. The response rates were 92.8 % and 93.5 %, respective-ly. Due to turnover, only 20 workers (experimen-tal group) participated in both questionnaire sur-veys. The analyses were performed for this group, all male, aged between 22 and 63 years (median 42 years). Demographic details about the respond-ents are presented in Table 1. Selection of a control group from the same workplace was not possible be-cause the lighting renovation concerned the whole workplace.

2.4. Data gathering

The fi rst lighting measurements and question-naire survey were conducted in March and April 2007. The second questionnaire was administered in a similar way in January and February 2008, about four months after the lighting renovation.The com-pany reported that they introduced the survey as a part of work and actively encouraged participation but informed it was voluntary. According to the pro-duction manager, two workers did not participate in either of the surveys. The questionnaires were completed individually at the workplace in an of-fi ce room close to the factory halls. Data were gath-ered either via the Internet (Digium Enterprise soft-ware) or using a paper version of the questionnaire. The paper versions were returned to the manager in a sealed envelope to ensure the confi dentiality of responses. As almost half of the employees were Swedish-speaking, a Swedish questionnaire, trans-lated by a licensed translator, was also available.

2.5. Questionnaire

The questionnaire is presented in Appendix 1. It consisted of the following sections: Background

research project. A few workers expressed that they found the renovation unnecessary. The workers par-ticipated in the identifi cation of problem areas. The preliminary renovation plan, suggested by Finnish Institute of Occupational Health, was presented to the employees and openly discussed. Most of the employees (approximately 90 %) supported the plan and, after this, the investment decision was made.

Minor noise abatement and ventilation improve-ments were also conducted during the study period in certain workstations. The study period also includ-ed minor organizational changes, such as improve-ment of machine safety, increase of production, some changes in personnel, and increase of control in terms of the implementation of a clock card and restricting smoking to specifi c areas. The economic situation of the company was stable during the research period.

2.2. Lighting systems

The old lighting system included luminaires both with fl uorescent tube lamps (107 pcs, 2 x 58 W) and mercury vapor lamps (84 pcs, 2 x 250 W). Both lamps operated with magnetic ballasts which may cause light fl icker at 100 Hz. However, visual evalu-ation did not indicate noticeable fl icker in any parts of the factory. The mercury vapor lamps caused glare problems, strong shadows and uneven light distribu-tion. All the lamps had poor color rendering proper-ties. The color rendering index (CRI) of the old fl u-orescent lamps was less than 60 and the correlated color temperature (CCT) was approximately 3000 K. CRI of the mercury lamps was less than 50. The rec-ommended minimum value in EN 12464–1 is 80.

The old luminaires (altogether 191 pcs) were re-placed with new luminaires (304 pcs) containing two 35 W T5 fl uorescent lamps and electronic bal-lasts. The new lighting system was designed using modeling software (Dialux) to satisfy the recom-mendations of EN 12464–1 standard [13], i.e., hori-zontal illuminance was at least 300 lx in task areas and at least 100 lx in passage areas. The number of the luminaires and their locations were designed to provide light evenly in task areas and passage ways. The light distribution of the new luminaires was wider and no mercury vapor lamps were used which eliminated the glare problems. According to the manufacturer, (CRI and CCT were based on the manufacture’s information) CCT of the new lamps was 4000 K and CRI was 80, which increased the quality of light signifi cantly.


89

2.6. Lighting measurements

The measurements were carried out identically both before and after the renovation. The illumi-nances were measured carefully in four task areas and in passage areas on a horizontal level at a height of 0.85 meters. The four task areas of 10–20 m2 were around the operated machines and they were occu-pied continuously. The lighting conditions in these four task areas represented typical lighting conditions of all task areas in the factory. The task areas were divided into a grid of 60 x 60 cm. The illuminances were measured in the centre of each square using a calibrated illuminance meter. Uniformity of the task area illumination was determined from the meas-ured illuminances. An imaging photometer was used to estimate glare problems in the selected worksta-tions. The images were taken to represent a view to the operated machine (Fig. 1). The luminance-ratio

factors, Indoor environment, Lighting quality, Ef-fects of lighting conditions, Symptoms and their re-lation to lighting, and Psychosocial environment. Questions assessing overall indoor environment, effects of lighting conditions and symptoms were modifi ed from an acoustic environment question-naire [12]. Most questions concerning lighting were generated by the research team after identifying es-sential factors of lighting in the literature [2, 3, 15]. One question about workplace colours was taken from [8]. Psychosocial environment was assessed with several items taken from the national ques-tionnaire survey ‘Work and Health in Finland’ [14]. Most questions were answered on a 5-point Likert scale (e.g. never, seldom, sometimes, often, very of-ten). The respondents were instructed to base their responses on the conditions of a typical work day during the past month.

Table 1. Demographic data of the experimental group (N=20), who participated both before and after the renovation, and excluded respondents in both phases


90

3.2. Questionnaire survey

Indoor environment. The disturbance of lighting conditions was reduced after the lighting renovation (Z = –3.1, p =.002), Fig. 2. The disturbance of ther-mal conditions also decreased (Z = –2.7, p =.007). Other indoor environmental conditions were per-ceived similarly before and after the lighting reno-vation. Satisfaction with the indoor environment as a whole did not change.

Lighting quality. After the lighting renova-tion, complaints about too little light decreased(Z = –2.9, p =.003), as did complaints about un-even light (Z = –2.8, p =.005) and shadows(Z = –2.1, p =.035), Fig. 3. Reflections from light coming from outside were perceived to be less of a bother (Z = –2.0, p =.046). Satisfaction with the possibilities to adjust lighting improved(Z = –3.4, p =.001). Satisfaction with the lighting conditions as a whole increased(Z = –3.6, p <.001), Fig. 4. The need for an extra, movable light decreased (Z = –2.2, p =.026).

Effects of lighting conditions. Compromising of the quality of work because of lighting problems decreased after the lighting renovation (Z = –2.3, p =.022), as did changing work station due to lighting (Z = –2.1, p =.038), Fig. 5. Self-reported initiatives for lighting improvements decreased (Z = –2.0, p =.049). There was a tendency to work less in an un-comfortable position in order to see better (Z = –1.8, p =.075). After the lighting renovation, fewer errors were reported to occur at work (Z = –2.6, p =.009), Fig. 6. Ten out of 19 workers reported fewer diffi cul-ties in fi nding tools and other necessary items after the renovation, but the difference did not reach sta-tistical signifi cance (p =.06).

There was a general trend towards few-er symptoms after the lighting renovation, Fig. 7. Dry eyes (Z = –2.0, p =.046), tired eyes (Z = –2.0, p =.046), tiredness and over-strain(Z = –2.5, p =.013) and motivational diffi culties (Z = –2.1, p =.032) decreased. The subjects who indicated having symptoms also rated the extent to

between the task view area and the surrounding area were inspected and bright light spots were searched. The luminances were averaged to estimate the overall change in luminance distribution. No daylight pen-etrated to the measurement areas. The measurements after the renovation were carried out after the stabili-zation of the new lighting system. CRI and CCT were documented using the manufacturer’s information.

2.7. Energy consumption

The calculated energy consumption of lighting was based on the nominal lighting effi ciency [lm/W] during maximum light output. It was assumed that all the luminaires were switched on during the day shift.

2.8. Statistical analysis

The questionnaire data were analyzed with SPSS 16.0 statistical software (SPSS Inc, Chicago, IL, USA). Wilcoxon Signed-Rank Test was used to an-alyze the changes between the fi rst and the second questionnaire survey. The number of respondents varied between 19 and 20 in individual analyses due to incidental missing responses. Statistically signifi -cant differences were reported when the p-value was below 0.05.

3 RESULTS

3.1. Lighting

The measurement results of illuminance in four task areas and passage areas are presented in Table 2. The measurements confi rmed that the new lighting conditions meet the recommendations of EN 12464–1. The luminance images showed that the luminance rose from values 1–10 cd/ m2 to 5–50 cd/ m2. The risk of glare problems decreased because the lumi-nance distribution was signifi cantly more even. The quality of the light sources increased signifi cantly, because CRI and CCT improved.

Table 2. Illuminance measurement results in four task areas before and after the lighting renovation. In other task areas illuminances were at least 300 lx


91

only rush at work was signifi cantly reduced (Z = –2.3, p =.021).

3.3. Energy consumption

The average lighting effi ciency of the luminaires was 45 lm/W before and 100 lm/W after the reno-vation. The improvement was obtained by the use of more effi cient lamps and improved electrical and optical properties of the luminaires. The instantane-ous electric power decreased from 61 kW to 23 kW

which their symptoms might be related to workplace lighting. Due to the small number of responses, a statistical analysis of these data was not possible.

Psychosocial environment. The distributions of most psychosocial variables corresponded well to the values of industrial workplaces in Finland [14], except that job satisfaction and support received from the manager were slightly lower before the lighting renovation, but not after it. There was a trend towards more positive assessments of all psy-chosocial factors after the lighting renovation, but

Fig.2. Disturbance of indoor environmental factors in the work area, means and standard errors in bars, medians indicated by digits

Fig.3. Disturbance of factors related to lighting, means and standard errors in bars, medians indicated by digits


92

ovation. Before the renovation, the objective meas-urements revealed signifi cant defects in the lighting conditions. The measured illuminances in the four task areas were clearly below the recommendations for industrial work. The new lighting system in-creased illumination and uniformity to meet the rec-ommendations of ISO 12464–1 [13]. After the reno-vation, complaints about specifi c lighting factors and about lighting in general were reduced, and satisfac-tion with overall lighting conditions was increased. No new complaints, e.g., of increased glare or excess light, emerged.

It is worth noting that satisfaction with possibili-ties to adjust lighting also improved even though such possibilities were not provided by the lighting renovation. This could be viewed as an indirect in-dicator of successfully implemented indoor lighting as it probably indicated the decreased need for ad-justable lighting resulting from suffi cient lighting in general.

The effects of lighting on work performance could not be objectively measured. Some gross measures of monthly machine operation rates and percentages of faulty products were available, but the interpretation of these data was hindered by several confounding variables, such as changing business conditions, changes in product demands, absentee-ism, machine repairs, and changes in the composition and number of staff. A reliable measurement of pro-

during the day shift. The total annual energy con-sumption of lighting was reduced from 530 MWh to 200 MWh. The costs of the renovation could be refunded within four years considering only the sav-ings in electric energy consumption using the price of electricity in 2007.

4. DISCUSSION

This study was able to describe a change in the subjective work conditions following a lighting ren-

Fig. 4. Satisfaction with lighting as a whole, %

Fig. 5. Behavioural coping methods used because of lighting problems, means and standard errors in bars, medians indicated by digits


93

from the lighting improvement, or whether there are confounding factors or methodological issues that affect the interpretation of the results. It was recog-nized that the before-after design of this study has many risks in terms of internal validity [16], par-ticularly as it was not possible to include a no-treat-ment control group due to the size of the company. The small sample size also prevented the use of sta-tistical analyses that could have detected mediating explanations, such as psychosocial factors. There were also changes in personnel, with six respond-ents leaving and nine new being recruited between the fi rst and the second survey. The demographic details of the respondents (Table 1) show that the six individuals who left the workplace in the mid-dle of the study were somewhat more dissatisfi ed in general than the remaining 20 individuals who responded in both surveys (experimental group). However, the observed results did not result from

ductivity effects would have required individual performance data from each worker, which was not possible in this workplace. However, the subjective ratings suggested that positive changes in work per-formance did occur. Particularly, questions related to the quality of work showed an improvement. Com-promising of the quality of work decreased and fewer errors were reported to occur after the lighting im-provement. Some support for improved work well-being, which could be viewed as an indirect measure of productivity, was obtained as eye symptoms, tired-ness and motivational diffi culties decreased. The fail-ure to fi nd signifi cant improvements in ergonomics was probably explained by the nature of the work as some work tasks required crawling under machines with portable working lamps, problems which could not be solved with general lighting.

The most important question to address is whether the observed changes really resulted

Fig.7. The occurrence of recent symptoms, means and standard errors in bars, medians indicated by digits

Fig. 6. The occurrence of diffi culties related to lighting, means and standard errors in bars, medians indicated by digits


94

of psychosocial factors seems minor for the percep-tions of lighting on several grounds. A careful inves-tigation of the experimental group revealed a group of six individuals whose job satisfaction was low before the renovation but had improved after the lighting renovation, whereas 13 workers attained a more satisfi ed attitude towards their job throughout the study. Job satisfaction can be seen as an outcome variable of the psychosocial and organizational envi-ronment as a whole. Therefore, it could be assumed that the group of 13 workers was not signifi cantly affected by any changes in the psychosocial work environment. The changes in lighting perceptions remained signifi cant when the analyses were repeat-ed, leaving out the dissatisfi ed group of six workers. Thus, the negative perceptions before the lighting improvement were not merely a spillover of general resentment towards the employer, and that the im-provement in perceived lighting did not simply mir-ror the improved psychosocial climate of the unsat-isfi ed sub-group.

The results might still have been distorted by the social desirability effect meaning that the workers would have given answers that they assumed would please the employer, i.e., by exaggerating shortcom-ings in the fi rst survey and improvements in the sec-ond. However, it was unlikely that this mechanism was responsible for the results because such a strat-egy would have resulted in a more consistent pattern of responses across questionnaire items, whereas the present fi ndings showed robust results in some sec-tions of the questionnaire but very few in others.

The results suggested a possible trend of reduced symptoms after the lighting renovation. The effect of lighting on worker well-being is likely quite com-plex, as it can generally be expected to be small in relation to that of psychosocial factors and concur-rent issues in the worker’s personal life. The present data did not allow such causal conclusions to be drawn, but demonstrates the importance of develop-ing intervention research to overcome the limitations encountered in this study.

5. ACKNOWLEDGEMENTS

We thank the production director of Parlok Ltd, Jari Salminen, for his patience and help in the practi-cal implementation of the study. We thank Marjaana Lahtinen for valuable comments on the manuscript. The study was funded by the Finnish Work Environ-ment Fund, which is gratefully acknowledged.

changes in pre- and post-sample as the analyses were only conducted for those for whom data was available from both surveys (a repeated-measures design). The nine new workers, who had not expe-rienced the poor lighting conditions nor had had the possibility to participate in planning the renovation, seemed to experience the conditions in a similar way as the experimental group of 20 workers (Ta-ble 1). With the response rate of 94 %, the second survey therefore represents the subjective conditions after renovation very well.

Even though the small sample size and the ex-perimental design do not allow making reliable causal conclusions, there are some mediating fac-tors that can be excluded as unlikely explanations. The role of the other environmental changes seems minor. The noise control measures were small and local and the noise complaints did not decrease. The decrease in thermal complaints possibly refl ected both decreased heat stress from lighting and the time of the year as the second questionnaire survey was conducted in spring time while the fi rst survey was conducted in winter time. However, the fi nd-ing that environmental satisfaction did not change supported the conclusion that major environmental changes did not take place. The fact that the light-ing improvement had no effect on environmental satisfaction could be attributed to the fi nding that lighting was experienced as a smaller problem than the other environmental problems, Fig. 2. Thus, the remaining problems might have cancelled the posi-tive effect of lighting improvement on the percep-tion of the environment as a whole. This outcome emphasizes the need to evaluate and develop work environments as a whole. To improve the envi-ronmental satisfaction, the development should be started from the greatest problem, which in this case would have been noise. Other studies also suggest that ventilation, thermal conditions and noise may in general be perceived as more important factors in industrial environments than lighting [15]. How-ever, noise control in industrial workplaces was very challenging in this workplace. Even a moder-ate reduction of noise level could not be achieved without signifi cant investments and hindrances to the production process.

The effect of psychosocial factors on the per-ceptions of lighting environment and on the effects of lighting requires separate examination as the psy-chosocial variables could have different effects on these two types of subjective outcomes. The effect


95

Appendix 1. Lighting environment questionnaire

Year of birth

Gender (female, male)

Job title (floor manager, processing industry worker, machine operator,

other, what?)

I wear glasses (yes, no)

I have been diagnosed with an eye disease or eye injury (yes, no)

How often do the following working environment factors bother

you in your work area? (never, seldom, sometimes, often, very often)

- draught

- too high or too low temperature or changes in temperature

- too dry or too damp air

- bad air quality

- smells

- noise

- too dim or too bright light, glare or reflections on the monitor

How satisfied are you with your working environment as a whole?

(very satisfied, quite satisfied, neither satisfied nor dissatisfied, quite

dissatisfied, very dissatisfied)

In the daytime, how much light shines on your work area from a

skylight? (none at all, a little, some, quite much, very much)

How often do the following factors related to lighting disturb your

concentration at work? (never, seldom, sometimes, often, very often)

- lack of light

- too much light

- uneven lighting

- the shadows caused by electric lighting or light from outside

- reflections caused by electric light (e.g. on the monitor)

- reflections caused by light from outside (e.g. on the monitor)

- bright spots (e.g. bare light bulbs, light-coloured surfaces in sunlight,

etc)

- flickering, fluttering or blinking lights

- the distortion of people’s skin colour or some other familiar object

Which of the following do you consider the worst fault with the

lighting of your work area?

- lack of light

- too much light

- uneven lighting

- the shadows caused by electric lighting or light from outside

- reflections caused by electric light (e.g. on the monitor)

- reflections caused by light from outside (e.g. on the monitor)

- bright spots (e.g. bare light bulbs, light-coloured surfaces in sunlight,

etc)

- flickering, fluttering or blinking lights

- the distortion of people’s skin colour or some other familiar object

- the hue of the light

- other, what?

How satisfied are you with your present possibilities to adjust or

direct your lighting? (very satisfied, quite satisfied, neither satisfied

nor dissatisfied, quite dissatisfied, very dissatisfied)

How satisfied are you with the lighting of your work area as an

entity? (very satisfied, quite satisfied, neither satisfied nor dissatisfied,

quite dissatisfied, very dissatisfied)

How colourful is your working environment? (no colours, very

subdued or monotonous, some colours, but not very colourful, colourful,

very colourful)

How often have you needed to act in the following ways because

of the lighting conditions in your working environment? (never,

seldom, sometimes, often, very often)

- turned off or dimmed disturbing light fixtures

- interrupted your work, left work or taken an extra break

- struggled harder than usually

- compromised on the quality of my work

- postponed work to a later time or worked overtime

- worked in an uncomfortable position to be able to see better

- moved to another work area

- discussed problems concerning lighting with a colleague

- made a suggestion about improving lighting conditions to those in a position

to make decisions

How necessary would it be to have an extra, movable lamp in your work

area? (very necessary, quite necessary, quite unnecessary, very

unnecessary)

How often do you have the following difficulties because of lighting?

(never, seldom, sometimes, often, very often)

- it is difficult to find tools or other necessary items

- dangerous situation developed at work

- mistakes are made at work

- I feel insecure at work

How often have you had the following feelings or symptoms lately?

(never, seldom, occassionally, often, very often)

- pain in my neck, back or shoulders

- headaches

- tired eyes

- dry eyes

- watery eyes

- tiredness or stress

- difficulties concentrating

- memory problems

- depression or melancholy

- sleeping difficulties

- difficulties with motivation

- irritability

- something else, what?

Are the symptoms or feelings listed above the result of the lighting in

your work area, or are they due to your other work or something else in

your life? (completely due to other factors; mostly due to other factors, but

perhaps also due to lighting; due to both lighting and other factors; mostly due

to lighting, but perhaps also due to other factors; completely due to lighting)

- same 13 items as in the previous question

How satisfied are you with your work as a whole? (very satisfied, quite

satisfied, neither satisfied nor dissatisfied, quite dissatisfied, very dissatisfied)

How often do you need to rush to get your work completed? (never, quite

seldom, now and then, quite often, very often)

Are you able to use your knowledge and skills at work? (very much, quite

much, somewhat, quite little, very little)

Can you affect matters that concern you personally at work? (very much,

quite much, somewhat, quite little, very little)

Do you get support and help from your boss when you need it? (very

much, quite much, somewhat, quite little, very little)

Do you get support and help from your colleagues when you need it?

(very much, quite much, somewhat, quite little, very little)

How has the planning and implementation of change usually taken place

in your workplace (e.g. information, choosing of employees, training,

participating in planning)? (very well; quite well; not well, but not poorly

either; quite poorly; very poorly)

How much can you depend on the fact that the management will make

wise decisions concerning the future of your workplace? (very much,

quite much, to some extent, quite little, very little)

How well do the following statements hold true for your work area? (very

well; quite well; not well, but not poorly either; quite poorly; very poorly)

- The flow of information between colleagues works well.

- Cooperation is direct and pleasant.

- The working environment is refreshing and invigorating.

- Work area is pleasant.

Do you have any other comments about lighting?


96

places, European committee for standardization, Brus-sels, 2002.

14. Perkiö-Mäkelä, M., Hirvonen, M., Elo, A., Ervasti, J., Huuhtanen, P., Kandolin, I., et al. Työ ja terveys – haastattelututkimus 2006: Taulukkoraport-ti. [Work and Health – Interview Study in 2006: Tabu-lar Report]. Finnish Institute of Occupational Health, Helsinki, 2006.

15. Juslén, H. Lighting, productivity and preferred illuminances – fi eld studies in the industrial environ-ment. Doctoral dissertation, Espoo: Helsinki University of Technology, 2007, Report 42.

16. Shadish, W.R., Cook, T.D., & Campbell, D.T., 2002, Experimental and quasi-experimental designs for generalized causal inference. Boston, MA: Houghton Miffl in.

REFERENCES

1. Sundstrom, E. Work places – The psychology of the physical environment in offi ces and factories, Cambridge University press, USA, 1986.

2. Veitch, J. A. Psychological processes infl uencing lighting quality. Journal of the Illuminating Engineer-ing Society. 2001, 30 (1), pp.124–140.

3. van Bommel, W. J. M., van der Beld, G. J. Light-ing for work: a review of visual and biological ef-fects. Lighting Research & Technology, 2004, 36 (4), pp.255–269.

4. Juslén, H., Tenner, A. Mechanism involved in en-hancing human performance by changing the lighting in the industrial workplace. International Journal of In-dustrial Ergonomics, 2005, 35 (9), pp.843–855.

5. Juslén, H., Fassian, M. Lighting and productiv-ity – night shift fi eld study in the industrial environ-ment. Light & Engineering, 2005, 13 (2), pp.59–62.

6. Juslén, H., Wouters, M., Tenner, A. The infl u-ence of controllable task-lighting on productivity: a fi eld study in a factory. Applied Ergonomics, 2007, 38, pp.39–44.

7. Grimaldi, S., Partonen, T., Saarni, S. I., Aromaa, A., Lönnqvist, J. Indoors illumination and seasonal changes in mood and behavior are associated with the health-related quality of life. Health and Quality of Life Outcomes. 2008, 6:56.

8. Küller, R., Ballal, S., Laike, T., Mikellides, B., Tonello, G. The impact of light and colour on psycho-logical mood: a cross-cultural study of indoor work en-vironments. Ergonomics, 2006, 49 (14), pp.1496–1507.

9. Niemelä, R., Rautio, S., Hannula, M., Reijula, K. Work environment effects on labor productivity: an in-tervention study in a storage building. Am J Ind Med., 2002, 42, pp.328–335.

10. Helenius, R., Hongisto, V. The effect of acous-tical improvement of an open-plan offi ce on workers. Proceedings of Inter-Noise 2004, paper 674, Aug 21–25, Prague, Czech Republik.

11. Hongisto, V. Effects of sound masking on work-ers – a case study in a landscaped offi ce. 9 th Interna-tional Congress on Noise as a Public Health Problem (ICBEN) 2008, July 21–25, Mashantucket, Connecti-cut, USA.

12. Kaarlela-Tuomaala, A., Helenius, R., Keskinen, E., Hongisto, V. Acoustic environment and its effects on work in private offi ce rooms and open offi ces – lon-gitudinal study during relocation. Ergonomics, 2009, 52 (11), pp.1423–1444.

13. European standard EN 12464–1 Light and light-ing – Lighting of work places – Part 1: Indoor work

Annu Haapakangas, Master of Science in Psychology, she is responsible for the psychological aspects of research in the laboratory

Jukka Keränen, Master of Science in Physics, he is responsible for acoustical and lighting research in the laboratory

Marko Nyman, Master of Science in Technology. He is a senior lighting engineer in KaNmanCo Ltd. He worked in Finnish Institute of Occupational Health during this project

Valtteri Hongisto, Doctor of Science in Technology, works as a senior research scientist. He works also as an adjunct professor in building acoustics in Aalto University in Helsinki

97

1. INTRODUCTION

Luminous fl ux is economically the most impor-tant photometric quantity. It is necessary to derive its unit, lumen, directly from the SI base unit, can-dela, for luminous intensity. The fundamental re-alization of the luminous fl ux unit is by goniopho-tometry, even realization by the absolute integrating sphere method [1] needs - strictly analyzed - the rel-ative luminous intensity distribution of the radiation of the lamp under test found by goniophotometric measurement.

The principle of goniophotometry is independ-ent of the technical realizations of the different types of goniophotometers. A goniophotometer measures the illuminance E r, ,ϑ φ( ) on a closed envelope around a lamp measured with a V λ( ) -matched pho-tometer for all directions ϑ φ,( ) of emission within the full solid angle ( 4π sr ) as shown in Fig. 1.

Here the radius vector r ϑ φ,( ) denotes the po-sition of the photometer in sphere-coordinates and r ϑ φ,( ) is always orthogonal to the entrance win-dow (facing the lamp) of the photometer. The radius vector r ϑ φ,( ) also describes the distance between the lamp and the photometer. The angle ϑ desig-nates the photometer’s position between the poles in a traditional way ( 0 ≤ ≤ϑ π ). The angle φ des-ignates the photometer’s position in the equator di-rection ( 0 2≤ ≤φ π ). The luminous fl ux value Ф of the lamp can be calculated from the illuminance E ϑ φ,( ) , either from the luminous intensity defi ni-tion I Ф= d d/ ω or from the defi nition of the illu-minance E A= d dФ / .

ABSTRACT

1The Physikalisch-Technische Bundesanstalt (PTB) is the national metrology institute (NMI) of Germany and it focuses on the realization, main-tenance and distribution of the SI units. Therefore, the determination of the luminous fl ux unit by goni-ophotometry has a long tradition. This paper gives an overview about the measurement systems devel-oped, used and improved over more than 50 years in PTB. In the fi fties and sixties of the past century a “single-lever-goniophotometer” was used to realize the unit lumen (lm) of luminous fl ux derived directly from the SI unit candela (cd) by luminous intensity transfer standard lamps. Later on, a large three frame gimbal mounted goniophotometer (radius 2500 mm) which allowed an automated movement of a pho-tometer and simultaneously an additional tristimu-lus head around the measured lamp was used. For the measurement of small lamps and LEDs, a simi-lar small three frame gimbal mounted goniopho-tometer (the so-called mini- goniophotometer) with a radius of only 300 mm was designed in the early 1980 s. Compact goniophotometers especially for LED-measurements were also developed. In 2006, the fi rst robot goniophotometer was installed in the new Albert Einstein Building of PTB.

Keywords: photometry, luminous fl ux, lumen, goniophotometry, single lever goniophotometer, gimbal mounted goniophotometer, compact gonio-photometer, robot goniophotometer, robotic

* On basis of the report presented at the 27th CIE Session in Sun City, South Africa, July 2011

A BRIEF HISTORY OF TRACEABLE GONIOPHOTOMETRY AT PTB*

Matthias Lindemann, Robert Maass, and Georg Sauter

Physikalisch-Technische Bundesanstalt, Braunschweig, Germany [email protected]



98

Φ = ( ) ( )∫∫r E2

0

2

d d0

φ ϑ φ ϑ ϑ, sin .ππ

(6)

On the other hand equation (2) with equation (4) and r = const will lead to equation (6), as well. But we must keep in mind that for large lamps equation (5) might be wrong ( r rL > ) but the luminous fl ux value according to equation (6) is right if the pho-tometer evaluates the light correctly with the cosine of its incidence angle.

2. GONIOPHOTOMETER SYSTEMS HISTORY AT PTB

As already partially discussed in section 1, a suitable goniophotometer system for luminous fl ux measurements should meet the following main requirements:

1. Calibration via a luminous intensity lamp as-suring traceability to the national realization of the unit candela;

2. Relative movement between the lamp and the photometer with a rectangular intersection of the ϑ- and φ -axis;

3. V λ( ) -matched photometer (luminous fl ux);4. Automatic data analysis;5. Independent measurement of the luminous fl ux

value from the spatial distribution and the spectral distribution of the measured irradiance of light;

6. Temperature independence of the photometer;7. Any burning position of the lamp under test;

Ф I= ( )∫ ϑ φ ω, ,4πsr

d (1)

or

Ф E Ar

= ( )∫ ϑ φ, ,4 2π

d (2)

where the solid angle element dω of equation (1) is given by

d d dω ϑ ϑ φ= ( )sin , (3)

and the area element dA of equation (2) is given by

d d dA r= ( ) ( )2 ϑ φ ϑ ϑ φ, sin . (4)

The radius r of the radius vector r ϑ φ,( ) can be

treated as a constant value because usually it is not changed during the goniophotometric measurements. Then the luminous intensity distribution I ϑ φ,( ) can be determined from the illuminance E ϑ φ,( ) by the inverse square law.

I r Eϑ φ ϑ κ, , .( ) = ( )2 (5)

If the photometric distance rL of the lamp under test is smaller than the measurement radius r of the goniophotometer then we can write equation (1) with equation (3) and r = const as follows:

Fig. 1. Illuminance E as a function of the radius vector r ϑ φ,( )


99

To simplify the further explanations we assume no offsets and rotations between the mentioned coordi-nate systems. That means the lamp coordinate sys-tem is equal the to the centre coordinate system.

Next we will discuss the properties, advantag-es and disadvantages of different goniophotometer systems used at PTB over the decades and whether they meet the listed requirements or not. We will start in the fi fties and sixties years of the past centu-ry with a single lever goniophotometer and we will end with the state of the art robot goniophotometer, introduced in 2007.

2.1. Single Lever Goniophotometer (1950–1975)

In this time period a manually operated single le-ver goniophotometer as depicted in Fig. 2 was used. First in the fi fties this instrument was equipped with a visual photometer. This photometer allowed the relative measurement of illuminance with an inter-nally installed auxiliary lamp and reducer system [2, page 321]. The traceability to the national luminous intensity unit the candela was assured by the cali-bration of this visual photometer by luminous inten-sity transfer lamps (this complies with requirement 1) and according to distance measurements based on the national length unit the meter.

Furthermore the visual photometer had to be in-stalled at a fi xed place as shown in Fig. 2 due to the position dependency of the auxiliary lamp inside the photometer. Hence a mirror/prism system was used

8. Air-conditioned environment;9. Small stray light section;10.Motor driven movement for the ϑ - and φ

-axis;11. Tristimulus head (correlated colour tempera-

ture of luminous fl ux) and/or relative spectral distri-bution of the measured irradiance of light);

12. Measurement without movement of the lamp under test;

13. Complete scan of illuminance/luminous in-tensity distribution of the lamp under test;

14. Monitoring of lamp’s output to correct for ag-ing during the measurement;

15. Temperature-conditioned walls (minimiz-es radiation exchange between discharge lamp and wall);

16. On-line stray light correction (especially for stray light tubes with greater fi eld of view for meas-urements of large lamps like tubular fluorescent lamps);

17. Goniophotometer room luminaries should be shadowed during measurement;

18. Variation of the measurement radius r (near fi eld, far fi eld measurements with imaging systems).

In goniophotometry, we have to handle different coordinate systems. There are at least the lamp coor-dinate system, the room coordinate system, and the goniophotometer coordinate system. Fig. 1 a shows these different coordinate systems with possible off-sets and rotations between them.

Please note the goniophotometer coordinate sys-tem is called “Centre” coordinate system in Fig. 1 a.

Fig. 1a. Coordinate systems


100

of the large lever (ϑ ) and turning the lamp (φ ) as well as the adjustment of the photometer, this meth-od was very time-consuming. For example; let us as-sume 600 points to be measured (from today’s view-point not a great many) and let us say 25 seconds per point. It took more than four hours to measure a lamp with a relative homogenous spatial distribution.

In 1963 an international comparison of the lu-minous fl ux unit of discharge lamps without fl uo-rescent material was carried out. These lamps had a very inhomogeneous spatial distribution of light. So it was impossible to measure these lamps with the described method because the number of meas-uring points needed would have been much too large. The solution as described in [3] was to replace the visual photometer by a rough V λ( ) -matched photomultiplier, adding an electrical drive to the φ -axis and a high precision capacitor for integrat-ing the photocurrent of the photomultiplier. Then for each ϑ -adjustment of the large lever the drive of the φ -axis rotated the lamp with constant ve-locity while the precision capacitor was integrating the photocurrent of the photomultiplier. After about t = 5 minutes for one turn in the φ -direction, the

charge Q C U= ⋅ of the precision capacitor repre-sented the mean illuminance

E k CU

ti = ⋅ . (9)

of a measured zone i . Again we get the luminous fl ux value from equation (8). There is no need to know the value of kC because it has been eliminat-ed during the calibration process with a known illu-minance from a luminous intensity lamp but it must be constant over the total measurement time. Later on, in the sixties, the visual photometer was also replaced by a modern V λ( ) -matched photometer based on fi ltered silicon photodiodes in combination with a photocurrent to voltage converter and an x/y-plotter to plot the illuminance distribution of a zone. Then the mean illuminance was determined in a graphical integration or a numerical integration way.

We will see in the next sections that this method of “analogue zone integration” is very similar to the digital evaluation performed in goniophotometers, today.

These last two improvements met the require-ment 3 and partially complied with requirement 4.

to redirect the light from the lamp under test to the visual photometer. More detailed information about a system like this can be found here [2, p. 319]. This arrangement complies with requirement 2.

For a complete luminous flux measurement a number illuminance values for manually adjusted different sphere-coordinates ϑ φ,( ) has to be meas-ured. If the fi ctive sphere of Fig. 1 is divided into parallel zones in the ϑ -direction and each zone i is divided into n sections of equal length with an illu-minance of E j , it is possible to calculate the mean illuminance Ei of each zone i by

E Ei n i jj

n

==

∑1

1, , (7)

and the luminous fl ux value by numerical integration with equation (8).

Ф r Ei ii

m

i= ( ) − ( )( )+=∑2 2

0

π cos cos .ϑ ϑ 1

1

(8)

The total number of single illuminance measure-ments depends on the spatial distribution of light of the lamp under test. Due to the manual adjustment

Fig. 2. Single lever goniophotometer 1950-1975


101

pivot-mounted inside the φ -frame but the work-ing range of this last frame is limited to a half turn. All three frames are mechanically balanced and motor driven. Their speed was manually adjust-able by 10-turn precision potentiometers because the frame speed regulation was of an analogue type. The angle determination for all frames was done by 13-bit angle encoders. The photometer was at-tached to the ϑ -frame (Fig. 4) which gave a fi xed measurement radius of 2500 mm. It was possible to attach a second photometer or a tristimulus head at the opposite end of the same ϑ -frame. In this case the working range was limited to 4 176° ≤ ≤ °ϑ by the two lamp holder rods, which were very suit-able for the measurement of tubular fluorescent lamps. In case of single socket lamps one photom-eter was detached and the crank of this frame side was turned to the give clearance to the lamp holder rod. This allowed a working range of 0 176° ≤ ≤ °ϑ .

According to equation (6) we have to integrate the measured illuminance over the full solid angle of 4π sr to get the luminous fl ux value. So it is obvi-ous to use a digital time integrating method to collect and store measured illuminance values. Fig. 5 shows the signal fl ow. The photometer and/or tristimulus head as well as the photocurrent-to-voltage convert-er including the voltage-to-frequency converter were of a temperature regulated type, too.

The photocurrent from the photometer with the luminous responsivity s normalized to CIE illu-minant A is converted with Rg in a voltage U and further converted in a frequency f with the relat-ed conversion factor w f . This frequency is count-ed by a digital counter as the signal. The quartz

2.2. Gimbal Mounted Goniophotometer (1976–2007)

The existing single lever goniophotometer re-ally needs a “helping hand” to move it manually from point to point. The computation of the lumi-nous fl ux value was carried out “by hand”, too. This procedure needed a long time and was together with additional requirements for the measurement of large tubular fl uorescent lamps, micro lamps and of course LEDs basically the reason to introduce a new goniophotometer.

2.2.1. Large Gimbal Mounted Goniophotometer

After intensively rebuilding several rooms of the Kösters Building, it was possible to assemble the newly constructed large gimbal mounted goniopho-tometer [4] in the new goniophotometer room of ap-prox. 7 to 10 to 8 m3. Fig. 3 shows a photograph of the large gimbal mounted goniophotometer. It also shows a tubular fl uorescent lamp under test.

This room was also equipped with additional regulated electrical room heating to adjust the air temperature according to the required measurement conditions. In principle a gimbal mounted goniopho-tometer as in Fig. 4 consists of three frames where the outer frame allows any spatial orientation of the lamp under test by turning this so-called α -frame around the y -axis. Inside this outer α -frame the φ -frame is pivot-mounted which provides an end-less variety of ways of turning this frame around the z -axis. Finally there is the ϑ -frame which is also

Fig. 3. Large gimbal mounted goniophotometer 1976-2007


102

Φ = ⋅ ⋅ ⋅ ⋅

⋅ −( ) ⋅+=

−

∑

c c c r

Ei ii

m

i

str spec elec

1

2 2

0

1

π

cos cos .ϑ ϑ(11)

During a measurement the walls of the goniopho-tometer room are also illuminated by the lamp un-der test as well as the goniophotometer itself. This illumination is also seen by the photometer attached to the goniophotometer and leads to a stray light il-lumination on the sensitive area of the photometer. This stray light illumination also depends on the fi eld of view of the photometer given by the stray light re-duction tube in front of the photometer. That means the stray light correction factor cstr has to be deter-mined for any used stray light reduction tube with its specifi c fi eld of view. This was carried out by differ-ent baffl es of different diameters which were posi-tioned between the lamp and the photometer during a normal luminous fl ux measurement. A detailed de-scription of the method can be found here [5].

The technical realization of a photometer is al-ways an approximation to V λ( ) , only. That means in most cases a correction of this mismatch has to be applied. Generally for all kinds of lamps with a rela-tive spectral power distribution S λ( ) the correction factor is obtained from equation (12)

frequency fq is counted simultaneously by a sec-ond counter to get the measured signal frequency back during the data analysis. This method assures the above-mentioned signal/time integration. As al-ready stated the fi ctive sphere around the lamp un-der test is divided into zones and sections (Fig. 5, upper part). Due to the analogue speed regulation of the two moving frames, positive and negative ac-celeration phases at the beginning and ending of a full scan and load alternation of the rotating sys-tem, a small wow and fl utter is unavoidable. ∆φ j , the section length, reduced this infl uence. With this weighting the mean illuminance Ei is computable with equation (10)

Es w R

fy

tfi

f gj

j

n

qi, j

i, j

=⋅ ⋅ ⋅( )

⋅ ⋅ −⎛

⎝⎜⎜

⎞

⎠⎟⎟

=∑1

2 10π

∆∆

∆φ , (10)

where f0 is the dark frequency (shadowed photom-eter) and ∆yi, j are the difference signal counts and ∆ti, j difference time counts with respect to the sec-tion start and end. Again the luminous fl ux value is determined by equation (8), but we have to add corrections for stray light cstr , the spectral mismatch of the photometer cspec and deviations from the stat-ed electrical measuring conditions celec as described in equation (11)

Fig. 4. Large gimbal mounted goniophotometer 1976-2007


103

for example my =0.018, the photometer’s match is of good quality. The exponent my (and the others) can be found from a fi t of known pairs of E T T( )( ), from a luminous intensity lamp at different working points as described here [6, 7]. Finally the correction factor for tungsten lamps can be written as follows

cT

T

my

specA

=⎛

⎝⎜

⎞

⎠⎟ (14)

Also in case of tungsten lamps operated close to

CIE illuminant A the determination of celec is easy because the value of luminous fl ux is a function of the electrical operating current J and we can use the following equation (15) to fi nd celec

cJ

J

mJ

elec =⎛

⎝⎜

⎞

⎠⎟

−

0

(15)

where J 0 is the nominal current and mJ ≅ 6 84. . More information can be found here [7].

The data analysis for the tristimulus head was carried out in the same way to get the chromaticity ( , )x y of the luminous fl ux from the three channels of the tristimulus head

cV S

s S

V P T

s P

spekrel

A

rel

d

d

d

=( ) ⋅ ( )( ) ⋅ ( )

( ) ⋅ ( )( ) ⋅

∫∫

∫∫

λ λ λ

λ λ λ

λ λ λ

λ

,

λλ λ,.

TA d( )

(12)

However, we have to know the relative spectral responsivity srel λ( ) of the photometer. P Tλ, A( ) indicates a Planckian radiator at a distribution tem-perature of TA K= 2856 . In case of tungsten lamps there is a more simple way without knowing srel λ( ) to correct for a spectral mismatch according to equa-tion (13). Where E T( ) is the corrected illuminance for a distribution temperature T from

E Ty T

s

T

T

my

( ) =( ) ⎛

⎝⎜

⎞

⎠⎟

A

, (13)

where y T( ) is the measured photocurrent and s is the luminous responsivtiy of the photometer for CIE illuminant A. A property of the photometer is the ex-ponent my . The index “ y ” indicates an exponent for a V λ( ) -matched photometer or the y -channel of a tristimulus head. The other channels will have dif-ferent exponents. If this value of my is close to zero,

Fig. 5. Signal fl ow


104

2.2.2. Small Gimbal Mounted Goniophotometer

For the measurement of small lamps and LEDs a similar three frame gimbal mounted goniophotome-ter with a radius of only 300 mm was designed in the early 1980 s. The photograph in Fig. 6 shows this so-called “mini-goniophotometer”.

The theory of operation was exactly the same as the large gimbal mounted goniophotometer. Also the signal fl ow including voltage-to-frequency convert-ers and counters was the same. All aspects of this system have thus already been discussed under sec-tion 2.2.1. In 2007 this instrument was replaced by a goniophotometer especially designed for measure-ments of LEDs [10].

2.3. Robot Goniophotometer (2006 …)

At the end of the nineties the planning of a new type of goniophotometer became more and more concrete, because PTB decided to construct a new building for the optics division, which is now called Albert Einstein-Building. So in principle this new building was designed “around” the new robot-goni-ophotometer. As for back as in 1995/1996 G. Sauter showed that goniophotometry by robots is possible and has an advantage compared to a gimbal mounted system [8]. As also described in [8] so-called service robots are suitable to realize a robot goniophotome-

xФ

Ф Ф ФX

X Y Z

=+ + (16)

yФ

Ф Ф ФY

X Y Z

=+ + (17)

Unfortunately, only the total luminous fl ux value

with the related chromaticity of the source could be determined, the spatial information was lost due to poor computer performance. In 1991 however, all computers, analogue systems, wiring and further parts were replaced by modern equipment. From that time on all the spatial data were stored in suitable media. As long ago as in the year 1997 this gimbal mounted goniophotometer was equipped additional-ly with an array spectrometer to measure the relative power distribution of the radiation synchronized to the luminous fl ux measurement. But the photometer and the input optic of the array spectrometer had dif-ferent fi elds of view and slightly different positions on the goniophotometer system which complicated the fi nal data analysis. However, now it was possible to get spatial spectra information of lamps under test.

In its last version from 1997 this goniophotom-eter met the requirements 1–14 from the list at the beginning of section 2. The relative expanded k = 2 measurement uncertainty for luminous fl ux values of standard lamps was 0.60 %.

Fig. 6. Small gimbal mounted goniophotometer 1982-2007


105

angle encoder from the desired spherical coordi-nates r, ,ϑ φ( ) . During the measurement of a lamp the robot’s control system reads these seven angles (counts) from a given list every 10 ms and tries to adjust each axis according to this value. After the adjusting process all seven angles are read synchro-nously and stored in a database with the related photometric data. In principle the photometric sig-nal fl ow is very similar to the large gimbal mounted goniophotometer as described in section 2.2.1. But there is one additional feature, because we have a distributed data acquisition system, it is necessary to use a time stamp of the same clock (1 MHz) for all systems to synchronize the data later during the analysis process.

As the described setting of each axis is the result of a regulation process, the actual value will differ from the desired angle. Therefore, we have to recal-culate the actual positions/orientation of the photom-eter from the actual axes’ angles.

It is common in robotics to describe a (rigid) ro-bot like a mechanical chain with so-called Denavit-Hartenberg Parameters (Later called DH param-eters), see [9]. Table 1 shows the DH parameters of one robot 1 as an example.

With these parameters and the actual angle θi for each axis, it is possible to calculate the position/ori-entation of the 7 th axis relative to the robot’s base

ter. Their design is slim with long arms and they are of light weight construction to move photometers and hold lamps. They also have a high fl exibility to move the photometer (s) on desired tracks with dif-ferent radii (1000 3000mm mm)≤ ≤r over the fi cti-tious surface of a sphere around the lamp. Each ro-bot is of seven degree of freedom (DOF) type which guarantees the required fl exibility and range of op-eration (Fig. 7). In fact the robot goniophotometer consists of two robots to cover the whole solid angle of 4π sr and one similar robot to position the lamp under test in the goniophotometer room. All robot supports were directly attached to the building’s con-crete to assure the long term stability of their spatial position. Fig. 8 shows the goniophotometer room with robots measuring an LED-traffi c light.

In contrast to a gimbal mounted goniophotom-eter the measurement radius is variable by turning all the seven axes into appropriate positions. Due to the large redundancy of a seven DOF robot, the tar-get position/orientation of the photometer is reach-able with many different positions of the seven axes. Hence we need a strategy to fi nd positions of the axes. That means in our case to fi nd the axes posi-tions in such a way, that the distances of all other parts of the robot from the centre are maximized. Special software is used to calculate all seven an-gles of the axes expressed in counts of each axes’

Fig. 7. Arm of robot goniophotometer


106

base frame to the centre coordinate system and the photometer in relation to axis 7. This is simply done by adding a base frame matrix F and a tool matrix W . Both are of the same style as equation (18). Then we have the position and orientation of the photometer P * with equation (19).

P F T T T T T T T W* = • • • • • • • •1 2 3 4 5 6 7 (19)

frame. With 7 matrix equations of type (18) the po-sition (marked grey)/orientation (marked dark grey) of each axis is calculable

So the position/orientation of axis 7 relative to the base frame of the robot is the vector product T T T T T T T1 2 3 4 5 6 7• • • • • • . But we are interested in the position/orientation of the photometer in rela-tion to the centre of a fi ctitious sphere. For this we have to know the position/orientation of the robot’s

Table 1. DH parameters of robot 1

axisi

∆θi

offset to rotation around Z-axis

α i

rotation around X-axis

di

translation in Z-axis

ai

translation in X-axis

1 -1.438° 45.024° 0.000 mm -0.400 mm

2 191.088° -90.026° 2686.322 mm 0.696 mm

3 178.585° 89.996° 274.088 mm 0.781 mm

4 6.229° -90.022° 1410.499 mm 0.474 mm

5 -0.141° -89.955° 173.036 mm -0.779 mm

6 -0.924° 90.032° 1191.007 mm 0.223 mm

7 180.443° 90.007° 91.130 mm 0.000 mm

T

a

i

i i i i i i i i i i i

=

+( ) − +( ) +( ) +cos sin cos sin sin cos∆ ∆ ∆ ∆θ θ θ θ α θ θ α θ θ(( )+( ) +( ) − +( ) +sin cos cos cos sin sin∆ ∆ ∆ ∆θ θ θ θ α θ θ α θ θi i i i i i i i i i ia (( )

⎛

⎝

⎜⎜⎜⎜

⎞

⎠

⎟⎟⎟⎟

0

0 0 0 1

sin cosα αi i id(18)

Fig. 8. Robot goniophotometer 2006...


107

environment’s temperature Tamb .The parameter R stands for a large data set to

describe the robot’s properties like the dimensions of the robot, mass distribution and so on. The com-putation of C ( ) is complicated and was created by the robot’s manufacturer. Due to competition on the robot market, C ( ) must be treated as confi den-tial and cannot be published. But the effect of C ( ) is publishable. Fig. 10 shows this effect.

The correction range reaches from approx. − < <2 3mm mmr depending on the photometer’s ϑ φ,( ) position. The equator area needs more cor-

rection than the polar area due to larger centrifugal forces. From the position part of P the current ra-dius r is obtained. Since the data fl ow is nearly the same as in shown in Fig. 5, the determination of the luminous fl ux value is in principle the same as us-ing equation (11) with equation (10), but due to two photometers we have to do this separately for both robots. Equations (21) and (22) are examples of ro-bot 1. Please, note that the current radius r i1, is with-in the sum of equation (22). The goniophotometer is also equipped with a monitoring photometer. This photometer is not moved during the measurement but is always pointed from a fi xed point to the lamp under test which allows a correction for a possible aging of the lamp during the measurement process as a function of time ( c t imon 1,( ) ).

Fig. 9 shows all the robot’s joints with their own coordinate systems (1..7) as well as the base frame and tool (=photometer) coordinate systems in a sche-matic way. P* is referenced to the centre coordinate systems. Additionally a typically track of the pho-tometer is shown. Due to the lamp holder stick (not shown here) at the upper equator region, the tracks are compressed to avoid a collision with this lamp holder stick.

As already mentioned, equation (19) is valid only for a rigid robot. In practice the robot’s orientation/position P* is infl uenced by static gravity forces and dynamic forces during the movement which causes bending of the robot’s structure as well as by chang-ing of the structure length caused by temperature differences to the reference temperature. Thus, we have to add a correction term to equation (19) like the following:

P Ct t

T R

F T T T T T

= ⎛⎝⎜

⎞⎠⎟ •

• • • • • •

θ θθ θ

1 71 7

amb

1 2 3 4

d

d

d

d..... , ..... , ,

55 6 7• • •T T W ,

(20)

where Ct t

T Rθ θθ θ

1 71 7

ambd

d

d

d..... , ..... , ,

⎛⎝⎜

⎞⎠⎟ describes

the final correction based on the angles θ θ1 7.....

of all axes, their time derivates d

d

d

d1 7θ θt t

..... and the

Fig. 9. Coordinate systems of robot goniophotometer and sample track


108

Φ Φ Φ= ⋅ +( )celec 1 2 . (23)The determination of the stray light correction

cstr in the robot goniophotometer room is different from the method used for the gimbal mounted goni-ophotometer: It is diffi cult to mount baffl es between the photometer and the lamp as described in [5]. Here we calculate the correction cstr directly from the luminance distribution of the walls of the gonio-photometer room. With the knowledge of the geo-metrical and photometric properties of the room, the luminous intensity distribution of the lamp and the correction could be calculated. Here the effect of the refl ected light on the walls and the arms of the robots is measured by an additional so-called “back-looking photometer” as described in [11].

Es w R

fy

tf

i

jj

ni, j

i, j

1,1,f 1,g

1, q1,

1,

1

=( )

⋅

⋅ −⎛

⎝⎜⎜

=∑

1

2 1

11 0

π

∆∆

∆φ ,

⎞⎞

⎠⎟⎟,

(21)

Φ1 str spec

mon 1, 1, 1

= ⋅ ⋅ ⋅

⋅ ⋅ ( ) ⋅ −( )+

c c

r c ti i i i

1 1

12

1

2, ,

, , cos cos

π

ϑ ϑii

m

iE=

−

∑ ⋅0

11

1, .(22)

Similar equations are valid for the data of robot 2 and so the luminous fl ux is the sum of Φ1 andΦ2 corrected with the factor celec from equation (15).

Fig. 10. Effect of radius correction

Fig. 11. Photometer (on the left) and Back-Looking-Photometer (on the right)


109

squares of 0.1 m x 0.1 m are used. Of course the ef-fect of stray light on the photometer’s active area de-pends on the fi eld of view of the used stray light tube in front of the photometer. Currently there are two types of stray light tubes in use with fi elds of view from approx. 15° up to 22.5°. Their typical correc-tion factors are 0.9995 and 0.9990. Each photometer is equipped with a back-looking stray light photom-eter to determined the luminance distribution with its own fi eld of view. Note that each back-looking pho-

Fig. 11 shows a photograph of a robots “hand” with the photometer and back-looking photometer. The back-looking photometer is a little bit tilted. Otherwise the measured illuminance is infl uenced by the shadow of the photometer on the wall.

Fig. 12 shows a sample stray light situation of the goniophotometer room with fi nite element grating of the luminance caused by the lamp under test. This example works with fi nite squares of 0.5 m x 0.5 m to make the grating visible. In reality smaller

Table 2. Uncertainty budget

Quantity Symbol Value Uncertainty Unit Sensitivity Contribu-tion/lm

Relative contribution

Luminous responsivity 1 s1

3.065 E-10 9.195 E-13 A lx-1 1.88 E+12 1.73 0.00150

Luminous responsivity 2 s2

4.111 E-10 1.2333 E-12 A lx-1 1.43 E+12 1.77 0.00153

Spectral matching exponent 1 m y1,

0.021 0.015 1 20 0.30 0.00026

Spectral matching exponent 2 m y2,

0.036 0.015 1 18 0.27 0.00023

CCT or distribution temperature T 2762 10 K 0.012 0.12 0.00011

Uncertainty in r ∆r 0.000 0.0020 m 901 1.80 0.00156

Calibration factor for frequency F cFH

1.000 0.0005 1 1141 0.57 0.00049

Calibration factor for frequency f cFD

1.000 0.0050 1 10.4 0.05 0.00005

Voltage to frequency converter 1 w f1,

50084.9 25 Hz V-1 0.012 0.31 0.00027

Voltage to frequency converter 2 w f2,

50056.1 25 Hz V-1 0.0118 0.29 0.00026

Amplifi cation gain 1 R g1,

1.000 E+09 99996 W 6.223 E-07 0.06 0.00005

Amplifi cation gain 2 R g2,

9.997 E+08 99966 W 5.430 E-07 0.05 0.00005

Calibration factor for cmon

ccm1.000 0.0005 1 1162 0.58 0.00050

Correction factor for stray light cstr

0.9995 0.0005 1 1134 0.57 0.00049

Luminous fl ux Ф 1154.1 3.27 0.0028


110

Only the more important contributions to the combined measurement uncertainty are listed in Table 2.

Thus the expanded k = 2 measurement uncer-tainty will be 0.56 % at the moment. That is nearly the same as we reached with the large gimbal mount-ed system. We must keep in mind that the very high fl exibility of this new goniophotometer complicates the position determination of the photometer. How-ever, in future a position look-up-table is planned to

tometer will measure the luminance distribution used as a correction for the opposite photometer. Now it is possible to calculate the luminous fl ux value and related standard measurement uncertainty. Table 2 il-lustrates the results from a simplifi ed Monte Carlo simulation (MC) of the calibration of an OSRAM Wi 4 luminous fl ux standard lamp. The spatial luminous intensity distribution is shown in Fig. 13.

The MC is based on the following equation (24) and detailed values are shown in Table 2.

Φ = ⋅⎛

⎝⎜

⎞

⎠⎟ ⋅

⋅

⎛

⎝⎜

⎞

⎠⎟ + ⋅ ⋅ (

−

cJ

J

T

Tr r c c t

m

m

i i

j

y

str

Acm mon

1,

0

12

1( ), ,∆ )) ⋅ −( ) ⋅

⋅ ⋅ − ⋅

+=

−=∑

∑cos cos

, ,

ϑ ϑ

φ

1, 1, 1i

1, FH FD1

1

i i

m jj

n

i jc F c

0

11

1∆ ff

s w R

T

Tr r c c t

f g

m

i i

y

1 0

1

22

2

,

, ,( )

( )⋅ ⋅

+

⎛

⎝⎜

⎞

⎠⎟ + ⋅ ⋅ ( )

1, 1,

Acm mon

2,

∆⋅⋅ −( ) ⋅

⋅ ⋅ − ⋅

+=

−=∑

∑cos cos

, ,

ϑ ϑ

φ

2, 2, 1

1, FH FD2

2

i ii

m jj

n

i jc F c f

0

11

2∆ 22 0

2

,

.

( )⋅ ⋅

⎛

⎝

⎜⎜⎜⎜⎜⎜⎜⎜

⎞

⎠

⎟⎟⎟⎟⎟⎟⎟⎟s w Rf g2, 2,

(24)

Fig. 12. Measurement and effect of stray light


111

REFERENCES

1. OHNO, Y. 1995. Realization of NIST Lumi-nous Flux Scale Using an Integrating Sphere with an External Source. CIE Proceedings, 23 rd Session New Delhi.

2. SEWIG, R. 1938. Handbuch der Lichttechnik, Band 1, Pages 319 and 321, Berlin, Springer Verlag.

3. FÖRSTE, D. 1966. Zur Lichtstrommessung an Lampen mit ungleichmäßiger Lichtverteilung, PTB-Mitteilungen, 1/66, pages 18 to 20, Braunsch-weig, PTB.

4. FÖRSTE, D. 1979. Ein Goniophotometer zur genauen Bestimmung des Lichtstroms, Licht-Forsc-hung, 1. Jahrgang, 3/79, pages 30 to 36, Heidelberg, Hüthig Verlag.

5. FÖRSTE, D. 1980. Elimination des Fremdlich-ts bei der Lichtstrombestimmung mit dem Goniopho-tometer, Licht-Forschung, 2. Jahrgang, 1/80, pages 27 to 29, Heidelberg, Hüthig Verlag.

6. ERB, W. 1997. PTB Network for realization and maintenance of the candela, Metrologia, 34/97, pages 115 to 124, Bristol, UK.

7. SAUTER, G. 1996. Kalibrierung von Lich-tstärke-Normallampen und Angabe der Messun-sicherheit, PTB interner Bericht, Braunschweig, PTB.

8. SAUTER, G. 1996. Goniophotometry: new calibration method and instrument design, Metro-logia, 32 (1995/96), pages 685 to 688, Bristol, UK.

minimize the radius uncertainty. An on-line-laser-tracer system to measure the photometer’s position is also possible.

Meanwhile the robot goniophotometer is suffi -ciently characterised to be used for all traditional (il-luminance based) calibrations and even an interna-tional RMO-comparison which is in progress. This capability complies with the listed requirements.

The next step at PTB uses the “basic concept of the robot goniophotometer” with imaging lumi-nance meters replacing the “traditional photometers” to operate the system as “Near-Field-Goniophotom-eter” allowing the direct traceability of measured luminance distributions to maintained photometric units. This future step will comply with the last point in our list of requirements.

3. CONCLUSION

The history of goniophotometry at PTB over more than 50 years is presented. Beginning with a manually operated single lever system with visual photometer via different gimbal mounted goniopho-tometers up to a state of the art robot goniophotome-ter for the 21st century. The latter has the potential to operate in “traditional” and in “near fi eld” mode over many years. This will solve the upcoming demands for traceability of photometric units for instant from solid state lighting technology.

Fig. 13. Spatial luminous intensity distribution of OSRAM Wi4 standard lamp


112

11. BIZJAK, G. 2009. Determination of Stray Light at the PTB Goniophotometer Facility, MA-PAN – Journal of Metrology Society of India, Vol. 24, No 3, pages 163–173, New Delhi, India.

9 . h t t p : / / e n . w i k i p e d i a . o r g / w i k i /Denavit-Hartenberg_Parameters.

10. LINDEMANN, M. 2009. Photometry and Colorimetry of Reference LEDs by Using a Compact Goniophotometer, MAPAN – Journal of Metrology Society of India, Vol. 24, No 3, pages 143–152, New Delhi, India.

Georg Sauter,Dr. He fi nished his studies at the “Technische Universität of Braunschweig” as assistant professor with investigations on the electrical properties of small-gap semiconductors. In 1978 he joined the Photometry laboratory at the Physikalisch-Technische Bundesanstalt (PTB), the National Metrology Institute of Germany. Since 1989 he has been responsible for all photometric calibration work in the laboratory and also for the accreditation of photometric laboratories within the German calibration service (DKD). In the 90ies he was the convener for a working group of the Comité Consultatif de Photométrie et Radiométrie (CCPR) to organize the fi rst CCPR intercomparison of luminous intensity and luminous fl ux with lamps as transfer standards. He retired in 2007, but he is still active in standardization organizations such as DIN and CEN and he is the German representative and associate director in the Division 2 of the CIE

Matthias Lindemann has got his education (1969 – 1982) at different courses in electronics and microcomputer techniques, Precision Engineering, Technikerschule der Stadt Braunschweig, Trainee, Precision Mechanic, Siemens AG. From 2007 up to now, he is a head of the Goniophotometry Working Group at PTB. His current work is connected with derivation and dissemination of the photometric unit for the luminous fl ux and the related angular dependence and colorimetric characteristic quantities like luminous intensity distribution, distribution temperature and tristimulus values, as well as the specifi c characteristics for semiconductor radiators with non-coherent radiation

Robert Maass, graduated as the certifi ed engineer in automation from Technikerschule der Berufsbildenden Schulen Peine in 1977. At present, he is the member of the Goniophotometry Working Group at PTB (Physikalisch-Technische Bundesanstalt). One of his main tasks is to be in close cooperation with the industry focused on a contemporary transfer of the developed novel methods and the resulting experience in the industrial applications and third party projects focused on traceability of photometric and radiometric units as well as colorimetric characteristic numbers from PTB to industries

113

tometric, radiometric, and colorimetric properties of LEDs. The standard conditions named A or B for determination of the (averaged) luminous or radiant intensity of LEDs are commonly and extensively applied by LED-manufacturers for production con-trol, device-grading and serve as basis for preparing of photometric and radiometric data sheets. Howev-er, optical radiation safety-related measurements are so far not covered by CIE 127. The possible photo-biological hazard of LEDs depends on the so called “physiological” radiance (i.e. spatially averaged and hazard effective weighted radiance) and relates to the potential hazard to the retina of the eye at direct long-term source viewing.

Due to the enormous gradients of the LED spec-tral distributions as well as of the applicable hazard-related action spectra, especially the spectral weight-ing to be done usually requires additional and ex-pensive instruments, i. e. spectrally high-resolving or correspondingly sophistically and spectrally well fi ltered measurement equipment.

Thus, – for economic reasons – it seems to be rea-sonable to analyze, under which conditions, modi-fi cations or limitations the existing measurement setups according to CIE 127 or data sheet values can be used to deduce required safety statements as well.

From a series of related measurements and calcu-lations with LEDs as well as with other light sources helpful conversion factors could be established for transforming the photometric data into correspond-ing hazard-related data. Especially in conjunction with the assessment of phosphor-converted white

ABSTRACT

1Optical radiation hazards which may be caused by lamps and other broadband light sources are as-sessed by application of CIE Standard S009:2002 “Photobiological Safety of Lamps and Lamp Sys-tems”. Related studies done by CIE TC 6–55 re-vealed that this standard is also adequate for the assessment of potential photobiological hazards of LEDs, rather than the previously applied laser safety regulations.

The present contribution shows how the safety-relevant data can be derived from the routinely de-termined luminous intensity of LEDs emitting in the visual spectral range. Also it will be shown how these photometric data can be transferred into corre-sponding spectrally weighted effective safety-related radiance values for comparison with the applicable threshold limit values of the standard CIE S009.

Keywords: photobiological hazards, LED

1. INTRODUCTION

Measurement procedures for characteriza-tion of optical properties of LEDs are already de-scribed in detail in document CIE 127 “Measure-ment of LEDs” as compiled by CIE TC 2–45. This document recommends defi ned measurement con-ditions, which should be used to specify the pho-

* On basis of the report presented at the 27th CIE Session in Sun City, South Africa, July 2011

SIMPLIFIED APPROACH FOR CLASSIFICATION THE POTENTIAL PHOTOBIOLOGICAL HAZARDS

OF LEDS ACCORDING TO CIE S009*

Werner Halbritter2, Werner Horak1, and Werner Jordan2

1 Siemens AG, Corporate Radiation Safety,2 OSRAM GmbH, Central Laboratory for Light Measurements, Muenchen, Germany

E-mail: [email protected]



114

at the retina of the eye. Because of the imaging prop-erties of the eye, direct viewing of an intensive light source results to a more or less extended image on the retina. The retinal irradiance (exposure) is line-arly and directly related to the (time-integrated) ra-diance of the source looked at. Thus, in the retina-related spectral range of 300/380 nm to 1400 nm the emission limits in CIE S009 are provided in radi-ance/time integrated radiance respectively (in W/ m2

sr or J/ m2 sr).According to the restriction to “visible” LEDs,

the main hazards, which have to be considered, will fall into this category – retina-related and provid-ed as spatially averaged effective (“physiological”) radiance.

The following simplifi ed assessment bases on the determination of the safety-related radiance accord-ing to CIE S009, the so called alternative method, see Fig. 1.

As an alternative to the so called standard meas-urement method, the radiance of the source can be determined by dividing the irradiance E (in W/ m2) at the measurement aperture by the solid angle subtended by the applicable angle of acceptance γ (in Fig. 1 defi ned by the fi eld stop):

LW

m sr

EW

msrFOV

2

2⎡⎣⎢

⎤⎦⎥

=

⎡⎣⎢

⎤⎦⎥

[ ]Ω(1)

However, vice versa, the radiance can also be determined by measuring the radiant power Φe [W] passing through the measurement aperture stop where the solid collection angle ΩAP [sr] is defi ned by the diameter of the measurement aperture at the distance r and the area is defi ned by the area AFOV

light emitting LEDs it even turned out that within a certain range of tolerance, this transformation can be applied also to other types of white light sources like fl uorescent and incandescent lamps as well. Fur-thermore, having this conversion procedure once es-tablished and proofed, it is even possible to evaluate common data sheet values this way in terms of their possible photobiological hazards. On the other hand, taking the limit values in “physiological” radiances of the CIE standard as starting points it is possible to determine the corresponding photometric values of LEDs e.g. for comparison with the maximum ra-tings of the data sheets. Following this way, it is pos-sible for everybody to decide in a certain case if fur-ther actions or even detailed assessments are neces-sary or could even omitted.

2. USAGE OF STANDARD MEASUREMENT SETUP ACCORDING TO CIE 127

2.1. Requirements of CIE S009 to “visible” LEDs

This lamp safety standard prescribes requires measurement of source emission under specifi ed conditions followed by a comparison of the result with emission limits. These threshold limits of the optical safety standard are presented as irradiance/radiant exposure or (time integrated) radiance – de-pending on the particular hazard to the skin or hu-man eye. As the measurement conditions are directly associated with the emission limits, both are defi ned by the standard and have to be observed strictly.

Optical radiation at wavelengths in the visible and near infrared (between 380 nm and 1400 nm) is transmitted through the ocular media and absorbed

Fig. 1. So called alternative method for the determination of the “physiological” radiance in terms of CIE S009 (Fig. 5.3 of CIE S009)


115

3 GENERAL APPROACH FOR “VISIBLE” LEDS

Obviously, the determination of the radiance ac-cording to CIE S009 as well as the determination of the luminous/radiant intensity according to CIE 127 in fi rst order is based on the determination of the irradiance in a certain distance. The CIE-conditions involve the use of a detector with a circular entrance aperture having an area of 100 mm2 (corresponding to a diameter of 11,3 mm). The minimum aperture diameter applicable to measuring “physiological” ir-radiance and radiance according to the lamp safety requirements is 7 mm. However, larger measurement apertures may be used if irradiance is uniformly over the diameter of the measurement aperture. In most cases with common LEDs, this uniformity can be assumed at least within the given range of aperture diameters from 7 mm to 11,3 mm.

Due to the simplifi ed considerations stated above, the standard conditions of CIE 127 shall be applied in a fi rst step for the determination of a correspond-ing safety-related spatially averaged luminance. In a second step, this photometric value shall be trans-ferred to corresponding effective radiance which is comparable with the risk group-related limit val-ues of CIE S009.

3.1. Deduction of a “safety-related” spatially averaged luminance

3.1.1. General

Within the context of CIE 127, the luminance of the source can be determined by dividing the ir-radiance at the measurement aperture by the solid

in the source plane as determined by the acceptance angle γ (and in Fig. 1 realized by the circular fi eld stop in front of the detector:

LW

m sr

W

A m

sr

e

FOV

AP2

2⎡⎣⎢

⎤⎦⎥

=

[ ]⎡⎣

⎤⎦

[ ]

Φ

Ω

(2)

In general, the measurement strategy as well as the setup equals in some aspects the commonly used standard measurement procedure for the determi-nation of the averaged luminous intensity of LEDs in CIE 127, § 4.3 (Fig. 4.2). Thus, instead of setting up new equipment for safety-related measurements, it seems to be reasonable to use already existing data for this purpose – if possible.

2.2. Requirements of CIE 127

The measurement geometries in CIE 127, § 4.3 are designated as CIE Standard Conditions A and B for the intensity measurement of LEDs, see Fig. 2:

Usually, according to CIE 127, the detector has been calibrated for illuminance, and the Averaged Luminous Intensity of the LED will then be calcu-lated from the relation:

ILED = E · d2 Ω0, (3)

where the distances d (expressed in meters) are:for CIE Standard Condition A: dA = 0,316 m, for CIE Standard Condition B: dB = 0,1 m.It will be assumed that in any case an “Inverse

Square”-relation between luminous intensity and ir-radiance applies.

Fig. 2. Schematic diagram of CIE Standard Conditions for the measurement of Averaged LED Intensity (Fig 7.1 of CIE 127)


116

Thus, the spatially averaged luminance LFOV can simply be determined by dividing the measured lu-minous intensity ICIE (according to CIE 127 – also averaged) with the above determined value of AFOV: LFOV = ICIE / AFOV.

This is especially applicable, as long as the real source (chip-) dimension is below or equal to 2,2 mm. For the assessment of larger sources, in terms of the safety standard, a corresponding fi eld stop of this diameter would have to be applied in front of the source, see Fig. 1. Therewith, a part of the emission would be blocked and not considered in the course of the luminance determination. However, the applied detector of the setup for the determi-nation of the averaged luminous intensity accord-ing to CIE 127 considers the whole and unlimit-ed emission of the source. Thus, if the assessment with the above averaging area AFOV is too restric-tive for larger source (chip-) sizes, the ICIE – based luminance- values can even be reduced by a factor of (2,2/d)2 for comparison with the emission limits, where d denotes the present (apparent) source di-mension in mm.

The closer the LED spatial emission characteris-tics are due to an ideal Lambertian emission pattern the better and trustable will be the results of this ap-proach. Most suitable are sources that do not show an apparent source. This problem may possibly be avoided or solved by an appropriate treatment of the source (e.g. cutting off the integrated plastic lens) or by measuring the bare non-encapsulated chips at a technologically early stage in production process.

3.1.2. Non-ideal Lambertian sources

The above simplifi ed approach can also be ap-plied with some corrections to sources with narrower spatial emission characteristics. Sometimes, LEDs will be offered with built-in or attached beam-shap-ing optics – in order to realize a certain more or less directed light-emission – i.e. to increase the lumi-nous intensity in a selected direction respectively. Additionally, to characterize this directivity, the half intensity angle φHW of the spatial emission distribu-tion (the “beam width”) has also to be determined for documentation in the data sheet.

In these cases, it is recommended that these lu-minous intensity values should not be used with-out modifi cation for the above drafted simple way of source assessments since the result might be over restrictive. The measured luminous intensity ILED ac-

angle subtended by the applicable angle of accept-ance γ (in Fig. 1 defi ned by the fi eld stop in front of the source). However, the generally required measurement distance r in the lamp safety standard is defi ned by 0,2 m, which would require (“Inverse Square”-) corrections of the illuminance-readings. Usually, these measurement setups are designed and calibrated for a direct read out the luminous in-tensities I (cd) anyway, according to the (“Inverse Square”-) relationship above. Thus, it is even sim-pler to directly use these values for the determina-tion of the corresponding luminance L by the gen-eral relationship:

LdI

dA=

1 1cos,

θ (4)

where L is a luminance (cd/ m2), dA1 is a source area (m2), θ1 – angle between the surface normal and the direction of the “beam” (worst case: θ1 = 0°) inserted and averaged leading to L = I/A1.

Contrary to the first alternative by measuring the irradiance in Fig. 1, the solid angle implement-ed now already covered by the luminous intensity (as defi ned by measurement condition of CIE 127) and the area will thus be defi ned by the area of the source. Therewith, the real luminance can easily be determined via the measured luminous intensity and the usually known LED source (chip)-area. In terms of photobiological safety this would even corre-spond to a worst case evaluation. Such assessments, if any, are required only in CIE S009 for comparison with the highest emission limit LR for retinal ther-mal hazards at very short exposure durations. These conditions, however, are usually out of considera-tion in conjunction with LED- (and even most other conventional light-) sources. In these cases the spa-tially averaged safety-related luminance bases on the size of the area A1 as defi ned by the applicable angle of acceptance γ as indicated in Fig. 1 – the area AFOV of the fi eld stop in front of the source. For evalua-tions in terms of the highest possible risk groups and for comparison with the emission limit LB (retinal blue light hazard) in general, this angle most con-servatively amounts to γ = 11 mrad. With the appli-cable measurement distance r = 0,2 m the diameter dFOV of the measuring fi eld and therewith its area AFOV can be determined respectively:

dFOV = γ ∙ r = 2,2 mm, AFOV = dFOV

2 ∙ π / 4 = 3,8 ∙ 10–6 m2.


117

3.2. Deduction of a “safety-related” spatially averaged luminance

3.2.1. General

The above determined spatially averaged pho-tometric luminance LFOV bases on an assess-ment/ measurement where the spectral sensitivity of the detector was defi ned by the (CIE standard photometric observer-) V (λ)-function, see the ex-ample of a “pc-white” LED in Fig. 3. However, for comparison with the threshold limit values LB and LR of the safety standard in effective radiance, the spectral power distribution of the LED would have to be weighted with the R (λ)- and especially B (λ)-action function, see the same LED in Fig. 4. Based on related calculations and measurements, it can be stated, that the optical hazard evaluation in LED cas-es can be limited to the assessment of the blue light hazard only.

Due to the strong slopes of the LED spectra as well as of the applicable spectral action functions, the spectral weighting usually requires high-sophis-ticated (and expensive) measurement equipment. But with some simplifi cations and worst case-assump-tions in most cases a simplifi ed assessment is possi-ble and seems suffi cient.

3.2.2. Conversion factors for “pc-white” LEDs

In any case of “pc-white” LED sources, the spec-trum usually consists of a narrow blue and a broad-er more or less yellow peak, as to be seen in Fig. 3

cording to CIE 127 should be corrected to simulate corresponding ideal Lambertian conditions at which the source luminance remains unchanged.

Following from the above quoted general rela-tionship between luminance, intensity and source area the “Conservation of Radiance” (i.e. luminance) requires a defi ned relationship between the (e.g. real AChip and the projected ALED) source areas and the re-lated (luminous) intensities (e.g. the original – Lam-bertian – IChip and the modifi ed ILED):

ALED/AChip = ILED/ IChip. (5)A correction, based on a cosine (raised) approach

(H = H0 cosm) for the simulation of the spatial distri-bution of LEDs (see also CIE 127, chap. 4.1) can be applied to the measured value:

I = 2 x ILED/ (m+1), (6)

where ILED is a measured luminous intensity of a “narrow-beam”-LED according to CIE 127,

I is the corresponding luminous intensity of an ideal Lambertian source of the same luminance, which is required to proceed with the simplified assessment,

m is the modifi cation parameter which depends on the half intensity angle

φHWm = log 0,5/log (cosφHW) (for instance, m = 1 for φHW = 60° – ideal Lam-

bertian: I = ILED; m = 4,8 for φHW = 30°).The accordingly corrected value should be used

for the determination of the luminance (and no fur-ther correction – e.g. for the source area or -position are needed).

Fig. 3. Radiometric spectrum of a “pc-white” LED (grey line). In course of the assessment of the averaged luminous inten-sity according to CIE 127 the V (λ)-weighting (black line) was applied (result: black line dotted)


118

temperature of the combined white emission and the higher also the effectiveness of the B (λ)-weighting, compared to the V (λ)-weighting (see Fig. 5) will be.

Thus, between V (λ)- and B (λ)- weighted radi-ances there is relationship which in fi rst order de-pends on the colour temperature of the emitted white light. (Note: A similar consideration is also valid for R (λ)-weighting.)

Within the limited range of colour temperatures in Fig. 5, the transformation factor FB from V (λ)- (expressed in cd/cm2) to B (λ)- weighted radiances (still expressed in W/ m2 sr) depends even rather lin-early on TC, see Fig. 6. In order to keep these con-version factors in a “handy” order of magnitude we prefer to use the dimensions of the photometric lu-minance values in cd/cm2.

This transformation can even commonly be ap-plied also to other types of white light sources, even for an extended range of colour temperatures. Fig. 7 shows exemplarily a compilation of measured trans-

and 4. The colour temperature is usually adjusted between “warm” and “cold white” by the intensity-relationship of these two peaks. The higher the rela-tive blue part in the spectrum the higher the colour

Fig. 5. “pc-white” LED spectra for different colour temperatures

Fig. 4. For comparison with the blue light hazard-limit values, the B (λ)-weighting function (black line) has to be applied to the LED spectral distribution (result: black line dotted)

Fig. 6. Transformation factor FB for “pc-white” LEDs from V (λ)- (expressed in cd/cm2) to B (λ)- weighted radiances

(expressed in W/ m2 sr)


119

Usually in case of white light LED-data sheets the following main data are provided:

• The maximum luminous fl ux Φ (in lm as for conventional light sources);

• Colour temperature (in K).We take in attention that the spatial emission dis-

tribution is commonly ideal Lambertian (for broad-area illumination/perceptibility).

Based on this, the luminous fl ux of surface- emit-ting LEDs, can simply be transferred to correspond-ing luminous intensity by I = Φ / π and then be used as described previous (§ 2.1.1): divided by AFOV (3.8 x 10–2 m2) results the safety-related luminance in cd/cm2 (or even simpler: LFOV = Φ/0,12 in cd/cm2). As shown in 2.2.2, the conversion factor for effective radiance depends on the colour temperature. If the result is smaller or greater 1 x 104 W/ (m2 sr), the al-location is RG1 or RG2 respectively. As to be seen in Fig. 7, a very conservative worst case- transforma-tion factor would be: 12 (assuming very “cool white” light). Therewith, as “a rule of thumb”, for a very short assessment, the luminous fl ux in the date sheet can just be multiplied by 100 for comparison with the RG1-limit. A detailed consideration of the real colour temperature is necessary only if the RG1-lim-it would be exceeded.

4.2. An (even more) simplifi ed approach of doing a rough estimation

A reversed consideration can start by taking the limit values LB of the standard as starting point it is possible to determine the corresponding (photo-metric) values of LEDs for direct comparison with the data sheets without any measurement or calcu-lation. This way, it is rather easy at least to decide if

formation factors FB (for the transformation V (λ)→) which comprises besides LEDs also fluorescent lamps, Planckian radiators as well as other types of white light sources. Also in these cases, the re-lationship basically depends linearly on the colour temperature of the white light.

Obviously, this type of transformation can gener-ally be used for assessments of white light sources as an estimate rather than to perform high-sophisti-cated spectrally based measurements. Even common data sheet values can be evaluated this way in terms of their possible photobiological hazards. Since the technologies for LEDs of different manufacturers are more or less similar, this general approach should be generally applicable. However, additional safety fac-tors should be applied (according to the scattering of the measurement readings in Fig. 7).

A similar relationship between photometric val-ues and effective safety-related data as for white light source mentioned above can be established also for coloured LEDs. In this case, the relation again depends on the colour of the source. However, this relation is not at all linearly. From approximately 520 nm of peak wavelengths the limits for thermal hazards are in any case more restrictive than those for blue light. However, these limits are unreach-ably high.

4. EXAMPLES

4.1. Usage of data sheet values of LEDs

Instead of real measurements of the luminous intensities according to CIE 127 (i.e. by the manu-facturer), also the available data sheet values can be used in almost the same manner (i.e. by the user).

Fig. 7. Measured relationships between B (λ) – (expressed in W/ m2 sr) and V (λ)- (expressed in cd/cm2) weighted radi-ances of several types of white light sources


120

further considerations or even detailed assessments are necessary or not.

As mentioned above, in a certain case, according to the data sheet, the specifi c values of the luminous fl ux as well as of the colour temperature decide if a certain device will have to be allocated into RG 1 or RG 2. This is demonstrated in Fig. 8: if the combi-nation of these parameter-pairs is below the red line, the LED belongs to RG 1 if above, to RG 2. The re-lation in Fig. 8 is especially valid, as long as the real source (chip-)dimension is below or equal to 2,2 mm. However, since the maximum real luminance of the semiconductor material is more or less physi-cally limited, greater luminous fl uxes are commonly achieved by increased source (chip-)sizes. In such cases (assessment of larger sources) also the fi eld stop of 2,2 mm diameter would have to be applied in front of the source. As pointed out in 2.1.1, if the assessment with Fig. 8 is too restrictive for larger source (chip-)sizes, the fl ux-values can even be re-duced by a factor (2,2/d)2 prior to checking Fig. 8.

5 CONCLUSION

These simplified considerations of estimating potentially risks of light sources should be used as a fi rst step in getting familiar with a concrete risk as-sessment for a special light source, especially with simple LED sources.

It is not issued as a “hard” guideline to decide whether there is a potential risk or not. It should be rather used as a helpful tool to show whether detailed and more information is needed to do a declaration of conformity on possible eye related hazards of the source under consideration.

Werner Halbritter,graduated engineer of Mechatronics/Precision Engineering (University of Applied Science – Munich), since 1993 at OSRAM GmbH Munich, Germany - Central Laboratory for

Light Measurements, at present time Senior Key Expert for Spectroradiometry

Werner Horak,graduated physicist, Siemens AG, Munich, Corporate Offi ce for Radiation Safety, convener of IEC TC 76/WG 9 “non-laser sources”, chairman of CIE TC 6-55 “safety of LEDs”

Werner Jordan,Dr. graduated from the University of Munich. After some years of academicals work, he joined OSRAM GmbH in 1986, at present time he is in Central Laboratory for

Light Measurements where he is working on radiometry, photometry and lab accreditation in frame of Annex A: Summary of the requirements of CIE S009/IEC 62471)

Fig. 8. Combination of data-sheet values for creating a border line of a certain “pc-white” LED to decide between RG 1 (below the black line) and RG 2 (above black line)

121

naires suggests an increase of the mesopic lumi-nance at road levels (up to 2 cd/ m2 in European standards (EN 2004)) or in tunnel lighting (internal zone) where, in Italy, the great luminance required values is 3 cd/ m2 (UNI 2003). For these reasons the Italian standard for the selection of lighting classes gives the possibility to adopt a class with lower re-quirements in the maintained luminance or illumi-nance values if LED luminaires with CRI greater of 60 is adopted.

Unfortunately, in some installations drivers report higher sensation of glare if compared with tradition-al discharge lamps installations with practically the same calculated values of TI.

Also, the present state of knowledge doesn’t pro-vide a reliable quantitative description of some vis-ual perception parameters when LED light is in-volved: colour rendering, glare and contrast evalu-ation need improvements in their defi nitions (CIE 2007).

As a matter of fact, psychophysical quantitative parameters (e.g. CRI and TI), and even vision mod-els (e.g. STV), as defi ned in CIE technical reports or international and national standards, show low or doubtful correlation with the real world vision con-ditions when LEDs are involved.

Lighting experts are aware of the problem and several researches about LED and glare have been starting to be carried out (i.e. (EMRP 2009)).

Our work takes place in this kind of researches. It is divided into three phases: the fi rst two are car-ried out in laboratory conditions (infl uence of the source spectra, infl uence of the source dimensions). The last one (dynamic evaluation of visual perform-ances) will be carried out next year considering real

ABSTRACT

1The aim of this research is to compare the infl u-ence of two different luminous sources (incandescent light and LED light) on perceived contrast of ob-jects in a disability glare context. About 25 subjects were involved in an experiment in order to monitor the variations in their contrast perception threshold, with and without glare sources, using a calibrated computer LCD display. In this fi rst phase the spectral infl uence is investigate. Typical applications of this research results, could be road and tunnel lighting requirements on disability glare restrictions when LED sources are used.

Keywords: subjective experiment, contrast threshold, disability glare, LED

1. INTRODUCTION

1.1. LED

We chose to investigate disability glare due to solid state lighting and comparing it with a tradition-al source, because nowadays LEDs are a great prom-ise in lighting engineering. This is true, not only for energy saving, lifetime, fl exibility in luminaries de-sign and range of luminous fl ux control, but also for appearance and quality of lighting.

Considering road and tunnel lighting the spectral behaviour of the LED light improve the vision con-ditions, especially if peripheral vision in involved. The S/P (scotopic/photopic) ratio of these lumi-

* On basis of the report presented at the 27 th CIE Session in Sun City, South Africa, July 2011

A COMPARISON BETWEEN DIFFERENT LIGHT SOURCES INDUCED GLARE ON PERCEIVED CONTRAST*

Paola Iacomussi1, Giuseppe Rossi1, and Laura Rossi2

1 INRiM, Division of Optics, Torino, Italy 2 INRiM, Division of Thermodynamics, Torino, Italy

[email protected]



122

1.3. Models for vision

As said, glare is always a very negative matter in vision: discomfort glare is annoying while high levels of disability glare can totally hamper vision (e.g. this is the case of glare weapons).

The main goal of this research work, considering the three phases described above is to verify and cor-relate the STV (Small Target Visibility) model and the CIE disability glare model when LED luminaires are used for road lighting.

In night vision, especially during driving, glare can affect the safety of people. For this reason, dif-ferent methods have been proposed in order to gen-erate a model for night vision, especially concerning road lighting. The safety key issue of a road lighting system is the ability to make visible an obstacle at a distance of vision, conventionally considered equal at 83 m. STV method is an algorithm (ADRIAN, W. 1989), (ADRIAN, W. 1993) to quantify the ability of a lighting system to make visible a series of nor-malized obstacles, properly distributed on the illu-minated area. It was proposed as an alternative to the usual luminance design method (IES 2000), based on the achievement of minimum values of average luminance and uniformity. In Italy, a standard is un-der development: it permits the comparison between systems formally similar in terms of lighting class-es, but with different characteristics in the ability to make objects visible on the road.

The veiling luminance due to the light sources in peripheral vision is evaluated following the CIE General Disability Glare Equation (CIE 2002).

= + +

+

L

E

p

A

Veil

Glare

10 5 0 1

162 5

3 2

4

θ θ θ,

,

⎡⎣⎢

⎤⎦⎥

⋅

⋅ ⎛

⎝⎜

⎞

⎠⎟

⎡⎡

⎣⎢⎢

⎤

⎦⎥⎥

+ 0,025p,

(1)

where:Eglare is the illuminance in lux on the observer

eye due to the glare sources;q is the angle between the observation direction

and the glare source (in degree);P represents the eye pigmentation factor (0 for

black eyes, 0,5 for brown eyes, 1,0 for light eyes and 1,2 for very light-blue eyes);

A is the observer age in years.This model is applicable for angles q between

0,1° and 100°.

installations. The range of tested luminance is be-tween 0,5 to 20 cd / m2 for considering the last part of the transition zone in tunnel installations too.

1.2. Disability glare and contrast

Glare is a very crucial matter because it affects negatively our ability in vision, reducing contrast and for this reason, our ability to perceive objects and obstacles. For example, in road lighting, the vis-ibility of objects is the key condition for motorised traffi c safety, and their perception is strongly linked to contrast, glare and, in a lesser extent, colour ren-dering conditions. These problems are deeply con-sidered in CIE both in Division 1 and 4 where TC is involved in their study.

Usually subjective experiments concerning dis-ability glare have been performed considering light-ing sources of different intensity and angular exten-sion and their infl uences on perceived contrast. From these experiments results, several vision models have been proposed, in order to be applied in light-ing systems design and in defi ning maximum ac-ceptable values.

For road lighting, CIE suggests several design rules and parameters (average value of luminance or illuminance, uniformity, threshold increment, etc., CIE 2010) adopted in national and international standards as lighting classes (EN 2004), where the performance values are given according to road and traffi c conditions. These values, based on several decades of experiences in installations, are well fi tted to common discharge light sources usually adopted in lighting plants.

Only disability glare is considered and a maxi-mum level of TI is given, correlated to the average illuminance or luminance on the road surface. Instal-lations where TI values are lower then the required limit are considered without glare, installations with TI values greater then this limit are not allowed. This represent a simplify approach of a continuous vary-ing phenomena, where the increase of glare requires more light to maintain constant the traffi c safety con-ditions. To reduce the environmental impact of light-ing installations and save energy some more com-plex rules could be proposed (EMRP 2009) especial-ly for adaptive systems.

For these reason, glare and contrast evaluation needs to be reviewed and improved for characte-rising LED luminaries in road lighting installations.


123

sources more representative of road lighting will be used (i.e. sodium high pressure). The third and fi nal phase will consider a real installation and dynamic experiments.

The subject was seated in front of the calibrat-ed monitor at a distance of 0,6 m. The luminance of the screen was set at two nominal values (9 and 22 cd/ m2). Its uniformity was measured using a cal-ibrated ILMD (Imaging Luminance Measurement Device) placed at the observer position. The average values were 9.04 cd/ m2 and 20.83 cd/ m2 with a uni-formity (minimum value / maximum value) of 0.89 and 0.91 respectively.

The targets position on the display was defi ned using a pseudo-random algorithm for every screen luminance and glare conditions (no source, LED source, incandescent source) and was the same for every subject.

To improve the accuracy in the calculation of in-trinsic contrast, the luminance of target and lumi-nance of the background (a square area centred round the target with a side of three times the target side) were measured with the same ILMD at every random position of the target.

In front of both lamps two diffusing glasses were put in order to obtain two uniform sources of the same angular extended. Using this expedient only the spectrum infl uences on subject performance has been investigated.

Since it is very diffi cult to obtain high uniformi-ties of the source surfaces, their luminance were measured with the same ILMD, pixel by pixel and these values used to calculate the veiling luminance (eqn. 1) considering the correct angle between the

2. THE EXPERIMENT

2.1. Description

The research described in this paper is a prelimi-nary test carried out with the aim to gain knowledge about the infl uence of the different spectra between white LED and an incandescent source on glare in the vision of small virtual targets in order to de-sign both a static and a dynamic experiment in an ad hoc real installation, with signifi cant parameters in the more critical range.

This experimental session investigates the in-fl uence of different light sources on the perceived contrast, using self luminous images of a target on a monitor in a dark and quiet room (Fig. 1).

Behind a dark panel, on the two sides of the screen, two holes (100 mm diameter) accommodate the LED source on the right and the reference source on the left (Fig. 1). The angular displacement of the sources was variable with the random position of the target in the display (direction of observation) in the range from 14° to 46°. The emitted spectra of the two sources are diagrammed in Fig. 2.

This test considers as reference source an in-candescent lamp at 3000 K. Incandescent sourc-es are not representative of traditional sources for road lighting, but of sources used in experiments about glare and vision modelling. In the next phase, a second static experiment, to evaluate the infl u-ence of the source luminous surface (a large surface of uniform high luminance for traditional luminaires and a number of bright point uniformly distributed on a low luminance surface for LED luminaires),

Fig. 1. Experimental set-up


124

• With only the LED source on, providing glare and 503 lx on the subject’s eyes.

In this way it was possible to investigate:• The contribution of glare due to a specifi c light

source on perceived contrast (considering also the time needed to fulfi l the test and the average time from the detection to a target and its subsequent) in order to delineate the perceptual threshold, with particular attention to the specifi c infl uence of lights with different spectrum;

• The infl uence of glare on the ability of subject in discriminating differences between target with dif-ferent contrast;

• The validity of the Adrian model of STV for glare coming from LED sources.

2.2. Measurement accuracies

All the measurements were carried out with cali-brated instruments and the measurement uncertain-ty of 1.5 % for illuminance and 2 % for luminance.

The power light sources and display operated with stabilized power supply to guaranty high sta-bility of the photometric parameters for all the ex-periment time. The operating conditions were tested periodically during the experiment days.

Because the position of the head of the subject was weakly defi ned, depending on the observer head dimensions and small movements during the experi-ment, the Illuminance on the subject eye could have been only estimated. For this purpose the eyes illu-minace was measured in 15 points of a parallelepi-ped 20 mm x 40 mm x 60 mm centred at the nominal eye positions, Fig. 4. The uniformity of illuminance in this space was 0.63 for the incandescent source and 0.53 for the LED source. These non-uniformities are

target on the screen, the measured luminance of el-ementary surface framed by the pixel and converting this luminance value in the equivalent illuminance on the observer eye.

Twenty-two subjects with good visual capacities, distinguished by age (22 to 30: young group, 35 to 55 elderly group), iris colour (light and dark) and eyeglass, attended the experiment.

Targets dimension and source position satisfy the constraints of the CIE model for glare (CIE 2002). All targets have the same squared shape and dimen-sion (4 mm) and are surely supra-threshold (refer-ring to dimension).

The experiment was divided in two sections: in the fi rst part, a set of eighteen targets, with dif-ferent luminance, was generated and randomly po-sitioned on the screen. When the observer saw a tar-get, had to click with the mouse on it. Two different backgrounds luminance (see above) have been used, and presented alternatively to the subjects.

In the second part of the experiment, subjects were asked to put in order all targets they detected before, and displayed in the same place as in the fi rst experiment part, from the lower to the higher con-trast as shown in Fig. 3.

Both experiment parts were performed by all sub-jects in three different visual conditions:

• Dark;• With only the incandescent source on, provid-

ing glare and 479 lx on the subject’s eyes;

Fig. 2. Spectral density of the two sources used in the experiment

Fig. 3. Target sequence ordered by luminance contrast


125

gets. This means that the selected group of observers was suitable for this visual task.

Considering all subjects together, the contrast perception threshold shift due to the presence of glaring sources is visualized in Figs. 8 and 9.

This shift becomes signifi cant with a low lumi-nance background and an incandescent glare source in negative contrasts. With a high luminance back-ground the two glare source have practically the same infl uence.

The resolution in the luminance increments of the computer display and its graphic driver reduce the possibility to better investigation in the near under-threshold zone.

For what regards the second part of the test, the subjects were asked to align all the targets they se-lected before ordering them by contrast (Fig. 3).

Considered the ideal right alignment for each tar-get selection made by each subject, we then calcu-

the main source of uncertainty in the evaluation of the veiling luminance and of the perceived contrasts.

2.3. Data analysis and results

As data analysis, we fi rst calculated the contrast perception threshold for each environmental condi-tion (dark, incandescent light and LED), for each background used for the target visualization (low and high luminance) and for each group of subjects (group 1: age from 22 to 30, group 2: age from 35 to 55, Figs. 5–7).

For what regards the first condition (in dark-ness), in general we see that not so many differ-ences comes out from the age parameter as contrast perception thresholds are similar between the two groups (Fig. 5). On the other hand, the threshold shift is considerable if a lighter background is used instead of a darker one for the visualisation of tar-

Fig. 4. Measurement points of illuminance around observer’s eye

Fig. 5. Contrast perception threshold in the dark condition, using low luminance background (left) and high luminance background (right), for the group 1 (age from 22 to 30) and group 2 (age from 35 to 55)


126

have a small training to understand clearly the ex-periment behaviour.

At low luminance background, the two sources give practically the same result with a low accuracy in doing the visual task.

Experiments carried out at high luminance back-ground did not show the bias conditions because the subject now has experience in the task and, expect for the incandescent source show a great improve-ment in accuracy (Fig. 11).

3. CONCLUSIONS

The preliminary experiments described in this pa-per try to obtain data for compare the spectral infl u-ence on disability glare for application in road light-ing of LED luminaires.

In road lighting the safety traffi c condition is de-scribed by the capability to detect an obstacle on the carriageway at a distance of about 83 m, when the

lated all the mistakes committed by subjects doing this operation.

A mistake in positioning a target in the correct place considering its contrast, as regards the one be-fore and the one after, has been considered one er-ror and so on for all the sequence starting from the fi rst target at left.

Then we obtained a mistake index for each sub-ject, dividing the number of mistakes committed by the number of targets ordered (this number is dif-ferent for each subject because of course someone can detect more target than someone else).

Doing this operation we could compare the infl u-ence of glare in the capacity of the subjects in order-ing targets by contrast, that means distinguish very little luminance differences.

The test stated in dark conditions and low lumi-nance background (see Fig. 10) therefore the great number of mistake we recorded in the dark condi-tion can be biased by the necessity for the subject to

Fig. 6. Contrast perception threshold in the glare condition caused by incandescent light, using a low luminance background (left) and a high luminance background (right), for the group 1 (age from 22 to 30) and group 2 (age from 35 to 55)

Fig. 7. Contrast perception threshold in the glare condition caused by LED light, using a low luminance (left) and a high luminance background (right), for the group 1 (age from 22 to 30) and group 2 (age from 35 to 55)


127

Fig. 8. Comparison between the three contrast perception thresholds (dark condition, glare by incandescent light and glare by LED light) considering all the subjects and using a

low luminance background for the target visualization

Fig. 9. Comparison between the three contrast perception thresholds (dark condition, glare by incandescent light and glare by LED light) considering all the subjects and using a

high luminance background for the target visualization

Fig. 10. Comparison between the three testing condition (dark condition, glare by incandescent light and glare by LED light) considering all the subjects and using a low luminance background for the target visualization

Fig. 11. Comparison between the three testing condition (dark condition, glare by incandescent light and glare by LED light) considering all the subjects and using a high luminance background for the target visualization


128

national Symposium on Visibility and Luminance in Roadway Lighting October 26–27 1993 Orlando Florida.

3. CIE 2002. CIE 146:2002, CIE 147:2002, Collec-tion on Glare:, Vienna: CIE.

4. CIE 2007, CIE 177:2007, Colour Rendering of White LED Sources, Vienna: CIE.

5. CIE 2010, CIE 115:2010, Lighting of Road for Motor and Pedestrian Traffi c, 2 nd Edition, Vienna: CIE.

6. EINHORN, H.M., PAUL, B.M. 1999. Discom-fort glare from small light sources, Lighting Res. Tech-nol., 31 (4) 139–144.

7. EMRP 2009 Joint Research Project Protocol JRP 09 Metrology for Solid State Lighting 09, VSL 2010.

8. EN 2004. EN 13201–2:2004. Road lighting Part2: Performance Requirements.

9. IES 2000 IES RP-8–00 (R2005) – Roadway Lighting ANSI Approved, New York: Illuminating En-gineering Society.

10. UNI 2003. UNI 10095:2003. Tunnel Lighting, Milano UNI (in Italian).

11. UNI 2007. UNI 11248:2007. Road Lighting – Selection of Lighting Classes, Milano UNI (in Italian).

eye is adapted at the average road luminance and in the presence of disability glare due to luminaires.

The fi rst part of the experiment investigate this situation: the effect of glare is practically equiva-lent for the two source but at lower adaptation lumi-nance the LED gives better performances for nega-tive contrast.

The second safety condition is to clearly identify the obstacle or the lighted environments. This can be correlated to the ability to perceived and understand small luminance differences, i.e. the visual task stud-ies in the second part of the experiment.

The effect of glare is practically independent from the glare source type at higher adaptation lu-minances (22 cd m-2) while at lower adaptation lu-minances LED (9 cd m-2) LED sources give better performances then incandescent lamps.

REFERENCES

1. ADRIAN, W. 1989. Visibility of Targets: Model for Calculation. Lighting Res. Technol., 21, 181–188.

2. ADRIAN, W. 1993, The Physiological Basis of the Visibility Concept, Proceedings of 2 ndInter-

Laura Rossi, Doctor in Philosophy of Science and now she is Ph.D. student in metrology at INRiM, Department of Thermodynamics. She is an expert in soft metrology and her fi elds of interest are visual and auditory perception, subjective experiments design and cognitive psychology

Paola Iacomussi, Doctor in physics, is a senior researcher at INRiM, Department of Optics. She is the Italian member of CIE in Division 8 and expert in photometry, and works for art illumination

Giuseppe Rossi, Ph.D. in Metrology, is a senior researcher at INRiM, Department of Optics. He is the Italian member of CIE Division 4 and expert in photometry, optical characterization of materials and road lighting installation

129

soon have a huge market share. On the other hand, the optical characteristics of LED replacement prod-ucts are very different and in many cases, they suffer from an inferior quality due to the lack of standardi-sation and control.

In this study, the optical and electrical param-eters of twelve commercially available LED linear replacement lamps have been compared and the var-iation over time of these parameters is investigated. Distributors of LED tubes recommend their prod-ucts as a superior replacement of conventional T8 fl uorescent lamps. In most cases, the luminaire is not removed but retrofi tted. The existing ballast has to be bypassed and the starter must be removed. Some distributors argue that retrofi tted luminaires achieve equal or larger illumination levels on the task area, even though the lumen output of the LED replace-ment is lower, referring to the superior light output ratio (LOR) due to the directionality of LEDs. How-ever, manufacturer data of LED tubes are often lim-ited and incomplete and the overall light distribu-tion of the luminaire after replacement is unknown. Therefore, a case study is presented in which the fl u-orescent lamps in a small offi ce room have been re-placed by LED linear replacement lamps in order to compare the illuminance distribution on the task area.

2. DISCOMFORT GLARE PREDICTION

2.1. Luminance distributions

There are many optical differences between con-ventional T8 fl uorescent lamps and their LED lin-

ABSTRACT

1Many manufacturers and distributors of LED tubes claim energy savings of 50 % and more when replacing T8 fl uorescent tubes with LED linear re-placement lamps. Above, most distributors pretend that the same visual performance and comfort will be maintained after such replacement. Optical and elec-trical parameters of twelve commercially available LED tubes have been measured and compared and the evolution in time of these parameters has been monitored. The performance of these lamps with dis-tinctly different luminance distributions is investigat-ed. Additionally, a case study is presented in which the fl uorescent lamps in a small offi ce room were replaced by LED linear replacement lamps in order to compare the illuminance distribution on the work place. It has become clear that the impact of replacing a classical fl uorescent tube by a linear LED lamp on the visual performance must not be underestimated.

Keywords: LED linear replacement lamp, fl uo-rescent lamp

1. INTRODUCTION

Many manufacturers present a LED tube of lower power as a more energy effi cient replacement for the conventional fl uorescent tube. This causes a contro-versy in the lighting sector. Indeed, LEDs have a lot of advantages and luminaires based on LEDs will

* On the basis of the report presented at the 27th CIE Session in Sun City, South Africa, July 2011

PERFORMANCE OF LED LINEAR REPLACEMENT LAMPS*

Wouter R. A. Ryckaert 1,2, Inge A.A. Roelandts1, Mieke Van Gils1, Guy Durinck1,2, Stefaan Forment1,2, Jan Audenaert1,2, and Peter Hanselaer 1,2

1 Light and Lighting Laboratory, Catholic University College KAHO Sint – Lieven, Gent, Belgium 2 ESAT-ELECTA, KU Leuven, Heverlee, Belgium

[email protected]



130

LEDs. By means of reverse ray tracing, luminance maps of both light sources are constructed. A length-wise section of the luminance maps, in the plane of the long axis and through the centre of the SMDs, is shown in Fig. 2.

The fl uorescent lamp has a constant luminance of 11270 cd/ m² but the luminance distribution of the LED tube exhibits strong fl uctuations between zero and 3.42 x106 cd/ m². The fact that the luminance between the emitting surfaces of the SMDs is not always zero is caused by scattering in the direction of the observer of light refl ected back to the SMD carrier plate by the plastic cover.

2.2. GLARE

The radiation pattern of the fluorescent lamp is accurately determined by tracing a high number of rays (5 x 106 rays). From these data, and by con-sidering the light source as a luminaire, an Eulum-dat fi le is generated. Via this fi le format the light source is imported in the lighting planning software Relux®. Because the Eulumdat fi le format is not possible for the LED tube that consists of numer-ous spatially separated and very small light sources, each SMD is considered as a small luminaire. So the simulation results for a single SMD are import-ed in Relux®.

In Relux® a simple lighting situation is set up to compare the performances. The T8 lamp is posi-tioned against the ceiling in the centre of a 3 m by 4 m by 2.3 m room. A 1 m by 2 m table with the table-top 0.8 m above the fl oor is placed directly under the luminaire. The illuminance of all surfaces is calcu-lated in Relux® by means of radiosity. All surfaces in the room with the exception of the fl oor and the

ear replacement lamps but the most prominent one is in the luminance distributions. The conventional lamp presents a luminous surface with a spatially constant luminance while most LED replacement lamps present a strongly fl uctuating luminance dis-tribution because of the individual LEDs.

In this section, the luminance distributions and discomfort glare estimates of a conventional T8 fl uorescent tube of length 1200 mm emitting a fl ux of 3350 lm, and a LED linear replacement lamp with 32 SMD (surface mounted device) LEDs emitting a fl ux of 1500 lm are compared by means of computer simulations. Both devices are mod-elled in the ray tracing software package TracePro® (Fig. 1).The outer surface of the fl uorescent lamp and the light emitting, 2 mm by 2 mm, square sur-faces of the SMDs are modelled as Lambert emit-ters. A fl ux of 46.875 lm is assigned to each of the

Fig. 1. Detail of the LED linear replacement lamp as modelled in the raytracing software; the distance between the centre of a LED to the next one is 37 mm, there are 32

LEDs

Fig. 2. Detail of a lengthwise section of the luminance distributions of the conventional T8 lamp and the LED tube. The luminance is calculated at points with 1 mm interval, zero corresponds to the centre of the LED tube. Notice the logarith-

mic luminance scale on the vertical axis


131

of the table facing towards the centre of the room). We fi nd: Lb=29.6 cd/ m² and UGR=26. According to the European standard for workplace lighting (EN12464–1 2009) this high UGR implies that this simple lighting solution is only acceptable in walk through areas.

In this room model, leaving all parameters un-changed, the fl uorescent tube is then replaced by the LED linear replacement lamp, which is modelled as a row of 32 SMD luminaires. In the computer con-structed image of the lighting situation no obvious glare is present (Fig. 4). Because of the very small surface area of the luminous parts of the LED tube, 4 mm² per SMD and 128 mm² in total, the classic UGR expression can not be used. For luminous sur-face areas smaller than 0.005 m² a modifi ed UGR expression, taking into account the intensity I of the light source, has to be used (Eble-Hankins 2004) (CIE 2002):

UGRL

I

r pb

i

i ii

n

=⎛

⎝⎜⎜

⎞

⎠⎟⎟

=∑8

0 2520010

2

2 21

log.

,

where ri is the distance from the observer to lumi-nous area i.

For the UGR calculation for the LED tube we consider each SMD as a separate light source (n = 32). For the same observer position and analogous as in the conventional T8 case we fi nd: Lb =10.2 cd/ m² and UGR=21.6. This lower UGR value implies that the bare LED tube is suitable for more lighting tasks than the bare T8. However, one should be careful not to judge the quality of real world lighting appli-cations by UGR and illuminance values alone, e.g.: LED tubes of the type discussed here can cause ir-

ceiling are assumed to be 50 % refl ective. The re-fl ection coeffi cients of the ceiling and the fl oor are taken as 90 % and 20 %, respectively. All surfaces are assumed to be diffuse refl ectors. This situation is shown in Fig. 3. In this room the bare fl uorescent appears to be causing discomfort glare which can be quantifi ed by the Unifi ed Glare Rating (UGR). Because the T8 lamp has a large diffuse emitting surface the classic UGR formula can be used (CIE 1995) (Eble-Hankins 2004) (Murdoch 2003):

UGRL

L

pb

i i

ii

n

=⎛

⎝⎜⎜

⎞

⎠⎟⎟

=∑8

0 2510

2

21

log.

,ω

where n is the number of luminaires in the room, Lb is the average background luminance at the observ-er’s eye, Li is the average luminance of the luminous parts of luminaire, i, ωi is the solid angle of the lu-minous part of luminaire i as seen by the observer and pi is the Guth position index (EN12464–1 2009) for luminaire i.

The background luminance is determined by cal-culating the total illuminance and the direct illumi-nance at the observer’s eye in Relux®. The differ-ence between these values is the indirect illuminance Ei. The background luminance is then found by (CIE 1995) (Eble-Hankins 2004):

LE

bi=

π

We position the observer’s eye at a height of 1.75 m at a horizontal distance of 0.80 m from the centre of the table, facing the centre of the opposite wall (i.e.: we consider an adult person standing in front

Fig. 3. Model of the room lighted with the conventional T8 fl uorescent lamp

Fig. 4. Model of the room lighted with the 32 LED linear replacement lamps


132

fi eld goniophotometer type RIGO801 (TechnoTe-am®) which utilizes an image-resolving CCD meas-uring technique for determining ray data and far fi eld luminous intensity distributions was used to determine the luminous fl ux and radiation pattern of all lamps. The temperature and relative humid-ity in the room were controlled within narrow rang-es (25 °C ±1 °C and 32±5 % RH). The spectra and resulting CCT, CRI and MCRI values were deter-mined by using a telescopic measuring head cou-pled to a spectroradiometer (Oriel®) with an optical fi bre. A cooled CCD detector captured the spectral fl ux after a suitable calibration measurement with a spectral radiance standard. The electrical parameters were measured by a Yokogawa® WT3000 precision power analyser (basic power accuracy of 0.02 % rdg). The sinusoidal supply voltage with an RMS value of 230 V was delivered by a power source type Agilent® 7813 B.

The results are summarised in Table 1.

Comments:• The median value of the luminous flux

is 1479 lm, which is only 44 % of the luminous fl ux of a new conventional T8 36 W/830 (3350 lm) fl u-orescent tube of the same dimensions. The spread in luminous fl ux is ranging from 754 lm to 1774 lm.

• There is a strong variation in lamp power go-ing from 10,3 W to 31,6 W with a median value of 17,8 W. The lamp effi cacy varies between 50.8 lm/W and 89.6 lm/W. In comparison, the effi cacy of a T8 lamp/ballast combination varies between 75–95 lm/W, dependent on the power consumption of the (electromagnetic) ballast.

• It is remarkable that only two out of twelve lamps have a CRI higher than 80! According to the European standard EN-12464–1 ‘Light and light-ing – Lighting of work places – Part 1: Indoor work places’ lamps with a colour rendering index lower

ritating multiple shadows and many people simply dislike rows of bright points as general lighting. Fur-thermore, in most cases a T8 lamp is used in combi-nation with a luminaire that is designed to handle the omnidirectionality of the lamp. Simply replacing the T8 with a LED tube can change the radiation pattern of the luminaire drastically (Myer 2009).

3. OPTICAL AND ELECTRICAL PARAMETERS OF LED LINEAR REPLACEMENT LAMPS

3.1. Initial lamp characteristics

In September 2010, twelve LED tubes of 120 cm of brands distributed on the Belgian market were collected. These lamps are intended for replacement of conventional T8/36 W lamps. Initially, the dis-tributors were asked to deliver a lamp with a corre-lated colour temperature (CCT) between 3500 K and 4000 K. However, control measurements revealed that the spread in CCT is large, as can be seen in Ta-ble 1 (3194 K – 4733 K; cold white lamp of brand B not considered). Most distributors indeed offer only 3 types (warm white – neutral white and cool white), and the stated CCT range is usually large (f.i. 3800 K- 4800 K). The visual effect of the spread in CCT is illustrated in Fig. 5. This spread can be an issue if the lamps are not replaced in group or if a defective lamp must be replaced.

After lamp stabilization, all relevant initial opti-cal and electrical parameters were measured: lumi-nous fl ux, radiation pattern, spectrum, colour ren-dering index CRI, memory colour rendering index MCRI (Smet 2010), correlated colour temperature CCT, active power P, luminous effi cacy, power fac-tor PF and total harmonic distortion THD. A near-

Fig. 5. Picture of the lamps – variation in CCT

Fig. 6 Spectrum of LED tube (Brand A)


133

tors around 0,5! This implies that twice the current is needed to deliver the active power as compared to a lamp drawing a sinusoidal current in phase with the voltage for the same active power. Other lamp currents contain a lot of high frequency compo-nents which may cause electromagnetic interference. Some typical current waveforms are shown in Fig. 7.

As LEDs are Lambert light sources, the radiation pattern of the tube is also nearly lambertian, charac-terized with a full width at half maximum of 120 de-grees. This is illustrated in Figs. 8,9.

3.2. Lamp characteristics after 2000 hours

After determination of the initial lamp character-istics in September 2010, all lamps are being oper-ated at a burning cycle of 3 hours (2 h45 ‘on’ – 15 minutes ‘off’) for a at least one year. In January, April, July and October 2011, all optical and elec-trical parameters are measured again. As the dead-line for submitting papers was March 31, only the measurements of January are presented in this paper. In Table 2, the luminous fl ux variation is shown for all the lamps.

White through-hole LEDs are used in tubes of brands C, E, G, and J. It is well known that heat in a traditional through-hole LED cannot escape ef-fi ciently from the semiconductor element. This pos-sibly explains the (strong) decrease in luminous fl ux.

than 80 should not be used in interiors where people work or stay for longer periods (EN12464–1 2009). A typical phosphor white LED spectrum is shown in Fig. 6. In all lamps under study, blue LEDs with an individual phosphor layer are used.

• Up till now, there are no specifi c requirements for the current waveform and maximum harmonic components for LED tubes if the rated lamp power is less than or equal to 25 W [IEC 61000–3-2 2005]. However, harmonic distortion may cause problems when many fl uorescent tubes are replaced by LED tubes with high harmonic content. The total har-monic distortion THD is a measure for the harmonic current content. For a sinusoidal supply voltage, the power factor λ is related to the THD as:

λϕ

=+

cos.1

21 THD(1)

Hence, the power factor combines the phase an-gle φ1 between the fundamental current and voltage component and the harmonic current distortion. The lower the power factor, the higher the losses in the electrical installation and the higher the risk for har-monic-related problems.

Four lamps have a high power factor with values greater than 0,9. On the other hand, the THD of three LED tubes exceeds 100 %, resulting in power fac-

Table 1. Initial lamp parameters

BrandLum.Flux[lm]

P[W]

Effi c.[lm/W]

CCT[K] CRI MCRI PF THD

A 1650 22,8 72,4 4186 90 96 0,97 14 %

B 1535 23,6 65,0 6876 72 74 0,45 192 %

C 1595 17,8 89,6 3709 76 84 0,82 56 %

D 1774 21,2 83,7 4016 69 74 0,66 90 %

E 754 10,3 73,4 4207 76 86 0,48 55 %

F 1707 20,9 81,6 3194 65 72 0,93 17 %

G 1036 15,2 68,2 3307 71 78 0,51 162 %

H 1437 17,7 81,1 3853 77 86 0,96 16 %

I 1605 31,6 50,8 3365 88 95 0,53 135 %

J 920 14,5 63,4 3678 78 89 0,84 59 %

K 1479 18,3 80,8 4733 65 63 0,78 54 %

L 1185 17,6 67,3 5329 73 80 0,91 22 %

Median 1479 17,8 73,4 3853 76 84 0,82 55 %


134

luminance distribution, will change the luminous in-tensity distribution of the original luminaire and thus the illuminance on the task area.

The luminous intensity distribution of the lumi-naire with four different lamp types were measured under controlled conditions (25 °C, sinusoidal volt-age) with a near-fi eld goniophotometer (see § 3.1). The lamps considered are:

• A new fl uorescent lamp type T8_36 W/840 – radiation pattern in Fig. 11-d. The measured active power of the luminaire is 51 W. Hence, the old elec-tromagnetic ballast consumes 15 W. The measured luminaire effi ciency LOR is 77 % (CCT=4000 K).

• LED tube brand A: tube with a diffuser – ra-diation pattern shown in Fig. 11-a. The active pow-er of the luminaire is 23,5 W; the LOR is 85 % (CCT=4186 K).

Moreover, the active power dissipated by the Brand C LED tube was also reduced by 19 %, which par-tially explains the decrease in luminous fl ux. The measured active power of all other lamps remained almost constant (change between –2 % and +1 %), as well as all other parameters considered (CCT, CRI, PF and THD).

4. RETROFIT OF A FLUORESCENT LUMINAIRE: A CASE STUDY

In Fig. 10 a small offi ce room used by the stu-dents’ union of the Catholic University College KAHO Sint – Lieven is shown. The room is 4 m by 6 m and 2.7 m high. Three old T8 luminaires are installed, each with one T8–36 W/840 fl uorescent lamp.

Replacing a conventional fl uorescent lamp by a LED tube with a hemispherical, yet quite different

Table 2. Luminous fl ux variation

Brand Ф, lmSept. 2010

Ф, lmJan. 2011

Differ.

A 1650 1659 +0,6 %

B 1535 1503 -2,1 %

C 1595 978 -38,7 %

D 1774 1792 +1,0 %

E 754 645 -14,4 %

F 1707 1711 +0,2 %

G 1036 894 -13,7 %

H 1437 1519 +5,7 %

I 1605 1643 +2,4 %

J 920 823 -10,5 %

K 1479 1584 +7,1 %

L 1185 1224 +3,3 %

Fig. 7 Current (black line) and voltage (dashed line) waveform (left: Brand A – middle: Brand D – right: Brand B)

Fig. 8. Radiation pattern (3 D)


135

will decrease over time, especially if through-hole LEDs have been used. The uniformity is in the same range. The light distribution on the task area is shown in Fig. 12 before and after the retrofi t.

The original lighting installation should have been designed to comply with the lighting specifi -cations [EN12464–1 2009] or legal requirements. It is clear that in most cases the modifi ed lighting installation will not provide the same level of illu-minance as the original lighting installation (Fig. 12) and the specifi ed requirements for lighting solutions for work places will not be met anymore.

To investigate the visual appearance of the small offi ce room (Fig. 10), 44 subjects were asked to fill out a questionnaire. The subjects, divided in groups of 4–6 persons, had to evaluate all four lamp types. During each lamp replacement, the sub-jects had to leave the room for several minutes. The reduced illuminance values in the room were no-ticed by nearly all respondents. Obviously, the room is perceived darker when the fl uorescent lamps are replaced by LED tubes. Adaptation cannot compen-sate for those large differences in illuminance and luminance.

• LED tube brand I: tube with only 32, but high intensity LEDs – radiation pattern shown in Fig. 11-b. The active power of the luminaire is 32 W, the LOR (Light Output Ratio) is 86 % (CCT=3365 K).

• LED tube brand K: tube with 360 SMD LEDs – radiation pattern shown in Fig. 11-c. The active power of the luminaire is 19,7 W (CCT=4733 K).

While a fl uorescent lamp emits light in all direc-tions, a LED tube is a Lambertian directional light source emitting light in a downward-hemispheri-cal arrangement (Fig. 8). Hence, the linear replace-ment lamps cannot use the luminaire refl ector de-sign in the same way as fl uorescent lamps do. The hemispherical radiation of LED linear replacement lamps results in less light losses within the lumi-naire. Hence, the luminaire effi ciency will increase after relamping (in our case, the LOR increases from 77 % to about 85 %). On the other hand, the radia-tion pattern of the luminaire can change consider-ably (Fig. 11).

For each luminaire – lamp combination, Eulum-dat fi les (.ldt) were generated and imported in the lighting planning software DIALux® to simulate the light distribution in the room under considera-tion. The calculated mean illuminance values Eavg and uniformity values g1 (i.e. the minimum divid-ed by the mean illuminance) on the horizontal task area (0.8 m height) are given in Table 3. A wall zone of 0.3 m was used.

Replacing all fl uorescent tubes by LED tubes will decrease the power consumption substantially, with energy savings up to 70 % (installation and mainte-nance costs not considered). However, the mean il-luminance will be reduced with about 50 % to an un-acceptable value! The illuminance values in Table 3 are initial values (depreciation factor equals 1) and

Table 3. Work plane illuminance

Pinstalled, W Eavg, lx g1

T8 153,6 278 0,21

Brand A 70,5 149 0,21

Brand I 95,7 146 0,15

Brand K 59,1 160 0,19

Fig. 9. Radiation pattern (C0–180 & C90–270)

Fig. 10. Small offi ce room used by the students’ union of KAHO


136

safety standards (EN 60598–1) and CE mark labe-ling. In fact, the organization modifying a luminaire should take over the full future responsibility for the luminaire with respect to all aspects (safety, EMC, photometric…);

• The lamp manufacturer’s warranty will be void if the adapted luminaire does not comply with lamp safety and performance standards.

6. CONCLUSIONS

Luminaires based on LEDs are emerging into the market. Low energy consumption, long life, dimma-

5. OTHER REMARKS

All commercial, economical and juridical aspects have not been considered in this study. Neverthe-less, there may be safety and other concerns [Cel-ma 2010]:

• It is known that an unsafe situation (elec-tric shock risks) can occur when LED modules are installed;

• The lamp holders of the original luminaire may be overstressed by the weight of the LED tube;

• Probably, the converted luminaires will not comply anymore with requirements of the luminaire

Fig. 11. Experimental radiation patterns


137

Nevertheless, there are applications in which the LED tubes may be the best solution, especially in case of cold-temperature conditions.

ACKNOWLEDGEMENTS

The authors appreciate the financial support of the Hercules Foundation – project AKUL035 “Near-Field Imaging Luminance Goniometer” and the fi nancial support of the Institute for the Pro-motion of Innovation by Science and Technolo-gy in Flanders (IWT-Vlaanderen)- project IWT 070488 – Groen Licht Vlaanderen/ Green Light in Flanders and IWT SB/0991442 – Jan Audenaert..

REFERENCES

1. CELMA 2010 CELMA position paper – T5 and T8 Fluorescent Lamp and LED Lamp/ module Adaptors “Retro-fi t Conversion Units” for T8, T10 & T12 Luminaires (2010).

2. CIE 1995 Commission Internationale de l’Eclairage (CIE) 117–1995, Discomfort glare in in-terior lighting.

3. CIE 2002 Commission Internationale de l’Eclairage (CIE) 2002, TC 3–01 Report, Glare from small, large and complex sources, Vienna.

bility and variability, compactness and negligible heat transfer in the light beam make solid state lighting an attractive alternative to traditional light sources. On the other hand, there is a lack of standardisation to evaluate products and a lot of products have an in-ferior quality and/or provide confusing or false per-formance claims. In this study, twelve LED tubes in-tended as replacements for T8 fl uorescent lamps were investigated. Bare lamp tests as well as performance tests in a typical luminaire with a parabolic refl ector have been done. The luminous fl ux values of the LED tubes tested are low and could result in unacceptable low illumination levels, even though the luminaire ef-fi ciency increases slightly when replacing a T8 lamp by a LED counterpart. Furthermore, many LED tubes have a bad colour rendering. Visual tests confi rm that the performance of LED tubes is still insuffi cient to replace fl uorescent lamps, especially in offi ce light-ing. To improve the energy effi ciency of a lighting installation, the most effective way is still by the in-stallation of a new high effi cient lighting installation. In that case, the lighting quality can be guaranteed and can be combined with reduced operating costs. When old-style fl uorescent lamps with poor color rendering indices (50–60) were installed, a relamp-ing with new fl uorescent tubes will also improve the lighting quality and reduce the energy consumption.

Fig. 12. Illumination distribution on the task areaLeft: luminaire with T8 lamp – Right: luminaire with LED tube brand I


138

ond Edition, 2003, Visions Communications, ISBN: 1–885750–05–6.

8. Myer 2009 Myer, M.A., Paget, M.L., Lingard, R.D., CALiPER Benckmark Report, Performance of T12 and T8 Fluorescent Lamps and Troffers and LED Linear Replacement Lamps, 2009 (US Depart-ment of Energy, http://www1.eere.energy.gov/build-ings/ssl/benchmark.html).

9. Smet 2010 Smet, K., Ryckaert, W., Deconinck, G., Michael, P., Hanselaer, P. Memory colours and colour quality evaluation of conventional and solid-state lamps. Optics Express, 18 (25), 26229–26244 (2010).

4. Elbe-Hankins 2004 Eble-Hankins, M.L., Wa-ters, C.E., VCP and UGR Glare Evaluation Sys-tems: A Look Back and A Way Forward, Leukos 1 (2) (2004) 7–38.

5. EN 12464–1 2009 EN 12464–1 Light and lighting – Lighting of work places – Part 1 : Indoor work places (2009).

6. IEC 61000–3-2 2005 IEC 61000–3-2 Electro-magnetic compatibility (EMC) – Part 3–2: Limits – Limits for harmonic current emissions (equipment input current ≤ 16 A per phase) (2005).

7. Murdoch 2003 Murdoch, J.B. Illuminating Engineering, From Edison’s Lamp to the LED, Sec-

Wouter R.A. Ryckaert received the Master of Engineering degree in electrical engineering from KAHO Sint-Lieven in 1998 and his M.Sc. degree in electrical and mechanical engineering from Ghent University in 2001. He received his Ph.D. from Ghent University in 2006 with a dissertation that explored the topic of the reduction of harmonic distortion in distribution networks with grid-coupled converters. Since September 2006, he is lecturer at the Catholic University College Sint–Lieven (KAHO Sint-Lieven) and responsible for the topic ‘Interior lighting and energy effi ciency’ in the Light and Lighting Laboratory of KAHO. Since September 2011, he is associated lecturer at the ELECTA research group of the ESAT department of K.U.Leuven

Inge A.A. Roelandts received the Master of Engineering degree in construction engineering from KAHO Sint-Lieven in June 2011. The subject of her master proof was “Energy effi cient lighting – relighting of municipal buildings, performance of LED linear replacement lamps and impact of absence/presence detector and daylight sensor”. Since September 2011, she studies the Master of Science in Industrial Engineering and Operations Research at Ghent University. Her fi elds of interest are lighting, methods engineering, work measurements and lean management

Mieke Van Gils received the Master of Engineering degree in construction engineering from KAHO Sint-Lieven on June 2011. The subject of her master proof was “Energy effi cient lighting – relighting of municipal buildings, performance of LED linear replacement lamps and impact of absence/presence detector and daylight sensor”. Since July 2011, she is an industrial engineer at AVG Consulting in Hamme. Her fi elds of interest are energy saving, methods engineering and safety on construction sites


139

Guy Durinck received a Master of Science in Physics from University Ghent (Belgium) in 1989 and a Ph.D. from Katholieke University Leuven (Belgium) in 1999. Since 2000 and 2010 respectively he is ‘docent’ (lecturer) at Katholieke Hogeschool Sint-Lieven (Catholic University College Sint-Lieven) and Hogeschool Universiteit Brussel. He teaches mathematics and physics to students in engineering and physical optics – to students in optometry. His current research interest is in the design of lighting and illumination optics and in optical modeling by raytracing

Stefaan Forment received a Master of Science in Physics from Ghent University in 1997. He received his Ph.D. from Ghent University in 2005. Since May 2006, he is active as a technological consultant and as postdoctoral researcher at the Catholic University College Sint – Lieven (KAHO Sint – Lieven) in the Light and Lighting Laboratory. His current research interests are spectral radiometric characterisation and near fi eld goniophotometry of solid state light sources (LED, OLED) )

Jan Audenaert graduated as a Master in Information and Communication Technology and started a Ph. D. research project funded by the Agency for Innovation by Science and Technology in Flanders (IWT) at the Light and Lighting Laboratory (Ghent, Belgium). His research focuses on raytracing for non-imaging optics

Peter Hanselaer received his Ph.D. in Physics at the University of Ghent (B) in 1986. Peter is professor at the Catholic University College Sint-Lieven and associate professor at the K.U.Leuven, dept. EAST/ELECTA. In September 2011, he became Vice president of the department of industrial engineering, responsible for research and international cooperation. He is board member of the Belgian Institute on Illumination and Belgian delegate in the CIE, division1.In 1997, Peter founded the Light & Lighting Laboratory which was supported by IWT Flanders and several industrial and scientifi c partners. Actually, the Laboratory is hosting 15 people. Scientifi c Ph. D. research activities are combined with consultancy activities towards the Flemish industry The main research areas are spectral radiometric and photometric characterization of light sources (fl ux, colour temperature) and materials (refl ectance, scattering, fl uorescence), the development of LED and OLED applications, optical design, energy effi cient lighting and stand alone photovoltaic systems


140

CONTENTS

VOLUME 20 NUMBER 2 2012

LIGHT & ENGINEERING(SVETOTEKHNIKA)

Jiří Habel and Petr Žák

The Present and Future of Lighting Engineering

T. Gerloff, M. Meyer, M. Taddeo, and A. Sperling

Measurements on OLEDs: Optical Properties, Reference Standards, Stability

D. Gasparovsky, A. Smola, M. Mácha, and P. Janiga

New Approach to Determination of Luminaire Maintenance Factor Curves for Various Conditions

Tor Mjøs and Pål Larsen

Indoor lighting – Energy Friendly Installations

S. Onaygil, Ö. Güler, and E. Erkin

Cost Analyses of Led Luminaires in Road Lighting

S.A. Fotios, Á. Logadóttir, C. Cheal, and J. Christoffersen

Using Adjustment to Defi ne Preferred Illuminances: Do the Results Have Any Value?

L. Halonen and M. Puolakka

Development of Mesopic Photometry – The New CIE Recommended System

Alexei A. Korobko, Vladimir M. Pyatigorskiy, and Anatoly Sh. Chernyak

The Measurement of Road Luminance Irrespective of Pavement Condition

P.1. Bodrogi, N.1. Wolf, and T. Q. Khanh,

Spectral Sensitivity and Additivity of Discomfort Glare Under Street and Automotive Lighting Conditions

U. Krüger, B. Ruggaber, and F. Schmidt

Spectral Properties of Imaging Luminance Measuring Devices Considering the Angular Dependence of the Spectral Transmission of Filters

Tony Bergen

A Practical Method of Comparing Luminous Intensity Distributions

Alexei A. Bartsev and Raisa I. Stolyarevskaya

VNISI Testing Centre for Solid State Lighting Application

Albert A. Ashryatov, Svetlana A. Mikaeva, and Anatoly S. Fedorenko

On Further Increase of Luminous Effi cacy of Low Pressure Fluorescent Lamps

light & engineering

Documents