ED 217 515

TITLEINSTITUTION

REPORT NOPUB DATENOTE

AVAILABLE FROM

EDRS PRICEDESCRIPTORS

IDENTIFIERS

,DOCUMENT RESUME

EA 014 496

Lighting for Education.Ontario Dept. of Education, Toronto.; OntarioMinistry of Colleges ands.Universities, Toronto.ISBN-0-7743-6275-881

42p.; Prepared bY\Architectural Services inconjunction with ,Ernest Wotton.Government of Ontario Bookstore, 880 Bay Street,Toronto, Ontario, M7A 1N8, 'Canada '($4.00).

mF01/pc02 Plus Postage: -

Color; *Design Requirements; *Educational FacilitiesDesign; Electronic Control; Elementary SecondaryEducation; Energy Conservation; Flexible LightingDesign; Higher Education; Human-Factors Engineering;*Lighting; *Visual Environment; Windowless Rooms;Windows*Fluorescent Lighting

ABSTRACTSome of the qualities and quantities that must be

juggled to produce good lighting for` educational facilities areanalyzed with photographs, tables, and drawings. The three categoriesof laMps used for school lighting (incandescent, fluorescent, andhigh intensity discharge) are described; a lamp selection guide-givesthe design characteristics of lamps in each category. Theeffectiveness of windows, the advantages and disadvantages ofwindowless schools, and screening devices for controlling glare andsunlight are discussed along with Aspects of daylighting design fornew schools. Scientific findings applicatle to light in schools aresummarized. Cohsiderations,met by the provision of manual switching,.time, photoelectric, and other light controls are described.Guidelines are offered for appraising an ex;sting lighting system.Related improvement procedures are categorized as those that requireno capital investment, those with a short pay-back period, and thosewith a long pay-back period. Exterior lighting and forecasts oflighting in the year 2000 conclude the booklist. A glossary is,provided and the appendices contain drawings and descriptions of theprincipal lamp types used, as well as two statements from official

,sources concerning the effects of ultraviolet radiation. (MLF)

*********************************************************************** Reproductions supplied by EDRS are.the best that can be made ** from the original document. *

***********************************************************************

ri

I

81-05113BN 0-7743-6275-8

Prepared by Architectural Servicesin conjunction withErnest Wotton; P.Eng., C.Eng., FIEE,FCIBS, FIESLighting Consultant and Designer, Ontario

The author acknowledges with thanks com-ments and information from:

'Mr. John BirettAlalton County RCSS; Mr.George Clark, GTE Sylvania Incorporated;Dr. Paul D. Clarke, M.D., Ontario Medical'Association; Mr: C. L. Crouch, IlluminatingEngineering Research Institute; ProfessorJohn Flynn, Pennsylvania State University;Mr. James Groves, Etobicoke School Board;Mr. Walter Jungwirth, Dufferin CouhtYSchool Board; Dr. A.M. Muc, Ph.D., OntarioMinistry of Labour; Mr. John C. Rankin, On-tario Ministry of Education; Dr. Gordon J.Stopps, M.B., Univergity of Toronto; Mr. Mur-ray Young, Dufferin County School Board;and the Illuminating Engineering Society ofNorth America.

Contents

1. Introduction 3 5. Some Effects of Lighting

Lighting and Health- Ultraviolet radiation and health- HID lamps in schools ,

5 Lamp colour rendering and the choiceof illuminanceLoss of muscle strength under variouslight sourcesLamp flickerFluorescent ligbting and hyperactivechildrenStudent tiredness

- The windowless school

Special Lighting Requirements- Lighting for areas with visual display

units (VDUs)- Lighting for plant growth

2. Creating a Good Visual Environment:Basic Considerations I 5

Illuminance- The effect on visual performance- Illuminance recommendations

The advantages of providing morethan one illuminanceEquivalent Sphere Illumination (ESI)Lighting for visually impaired students

Glare 8

Brightness Balance 10

Colour 11- Colour appearante, colour rendering,

and the choice of colours usedin decoratingColour rendering and the choice oflarfips for school activities

3. Lamps and Ballasts

LampsIncandescent lampsFluorescent lampsDummy lampsHigttiplensity discharge (HID) lamps:

Mercury lampsMetal halide lampsHigh pressure sodium lampsLow pressure sodium lamps

Operating characteristics of HID lamps

Ballasts

13

21

21

23--

6. Switching and Energy Managementin Lighting Systems 25

Manual Switching .

13 Trme Controls

Photoelectric ControlsOh -off photoelectric controlTop-up photoelectric control

Occupancy Detector Control

Monitoring Lighting Use: TheElapsed-time Meter

7. Making the Most of Your Lighting16 System: Basic Considerations

4. Windows and Daylight 19

The Effectiveness of Windows

Screening Devices for ControllingGlare and Sunlight

Aspects of Daylighting Design

19

19

20

- Procedures requiring no capitalinvestment

- Procedures with short pay-back period-.Procedures with long pay-back period

8. Exterior Lighting

9. Lighting in the Year 2000 .

Bibliography

Glossary

Appendix 1: Principal Lamp TypesUsed

25

25

25

26

26

27

29

31

33

34 \?

35

Appendix 2: The Effects ofUltraviolet Radiation 39

t.

I ilqi,1101,0 I 11114

IL

LIP

*q',Iidhiildic, ,

a

I 1. introduction a

f

The general appearance of a classroom-the way the classroom iooks and the qualityof the lighting has an immediate andmarked impact on both teachers and stu-dents. It affects their enjoyment of the class-room: one that is gloomy and monotonous,and in which neither teachers nor studentslook their best, will not be a pleasant place inwhich to work. It also affects the students'ability to learn. School work makes excep-tional demands on the vision of the students.If the visual environment is such that theycannot see well, then they will notjearn well.

When we that we like the lighting in aparticular classroom, we are usually referringto more than 1st the lampi and luminairesused to produce the effect. What we arereally talking about is the total result - thevisual environment.

The visual environment is the most so-phisticated and best documented aspect ofenvironmental design. It involves windowsand thearnount of daylight they let in; thequality of the electric lighting system; the col-ours and finishes used on walls, ceilings, andother surfaces; and the design and layout ofthe furniture - to name some of the majorconsiderations: It alSo Involves considerationof energy costs and consumption.

Improvements in lighting technologysince the turn of the century have greatly in-creased our ability to control the visual envi-ronment, In the early 1900s, there was noelectric lighting in schools, and classroonia

Cambrian College,BartydowneGartipus, Sudbury

Townend StefuraBaleshta & Phister,Architects

had to have high ceilings and tall,Windows_tOlet In sufficient daylight. lñdowIOcàtknidetermined_the-frontof-eachfoionitineestudents-_wereseAted_So that Iightwover their left aboUlders.TheyWel_'4_vallat-4-::,sumed to beright-nanded.WhatertifiCial-lighting there was came fronicoal;bilorgas-=---,lamps.--

A Majorldevelopnient in-SON:01_1 -hit--occurred just after the Firat-World-WacwasAhe introJiactiOn for generatilse-_- of thecandescent tungsten filament limp the aO___of lamp found in every_ tOday;Tgbeglnwith, this forni-ofelectitcjightingiaerelyauplemented daylight: it waSnotused.tere=,.--_-_-_:place daylight toanygreatextent BUt asthese lamps beoarne more efflelent and eltric energy became cheaperfdaylightWas

_

considered less as the source of light forf=--clasSrooms._-_-

The tubular fluorescentduced after the_Second World War arei _

brought about a radical change in the think--ing on classroom lighting. With this high-effi-ciency lamp (it ienow three to four times as--efficient as the incandescent lamp), many ot_the previous restrictions on classroom light='ing were overcome. It now became cheapenough to make it possible to have virtuallyas much light as we wanted whenever wewanted it. Consequently, the fluorescentlamp was adopted as the principal means ofclassroom lighting, The classroom wassOmetimes provided with large windows,sometimes with small windows; and occa-sionally with no windows at all. Even whenthe classroom had large windows, daylightwas usually considered as a bonus addi-

-tional to the electric light.

=

Lighting in universities followed a similarpattern of development. Many of the univer-sities with older buildings Were slow to adoptthe fluorescent [woes_ a light source foraesthetic reasons. 'Theirattitude changed;however, as new buildings were constructedincorporating advanced fontiof lighting sys-tems. There was, as might be expected, a -Marked contrast in lighting conditions be-tween the old andnew buildings. As a result,there has been econlipuingopgradingef thevisual environment ftoughout the post-sec-ondary system. /

Colleges of applied arts and technology,a comparatively recent development-on theeducation scene, did not have to ma..e thistransition: They are almost all served by new-

. or extensively renovated buildings; and theirstandards of Illumination reflect the design,4';12: phy of tRe1960s and 1970s.

last quarter of the twentieth centurypresents a new challenge. Advances in light-ing technology must now be reconciled withthe increasingly vital considerations of en-ergy costs and conservation. Although light-ing of all kinds consumes only about 5 percent of the total national consumption of en-ergy, it is a particularly conspicuous user ofenergy. Inevitably, as the cost of energy ris-

es, there will be pretsure-____ aamount of energy lco_rtsu

systems._

Th0-001600k10114-00$01-0;meanheeUin fact! -Atte 'Citianusible energy-saving OM!

tnteitielightinIuminanceS.40cOrinehrttid_

= the ute-S1-#SYIsity_datgtr:more-effeCtiiiajOnlinairek-_-p fifes-with o_ing, will

Georgian College,Barrie Campus,BarriePage & Steele - Salter& Allison, Architects

0

4

11041040.-r til_reSS

weigh one design OlO!ieffectpnithe:

awareelements.--bles ot figures and codified : piocedUres arepossibly-noteS reliable a guide aewat _oncethought .= it is experiencelhat-MUSt counow. , s ,

This booklet sets out to be nOtacr Muth a- rigid guide 'or checklist but rather to bonSider

some of the qualities and quantitiesthat must:. be juggled to produce good lighting for edu-cation.

01=

N

2. Creating a Good VisualEnvironment: BasicConsiderations

I order to create a good visual environment,we must be concerned with:

1. producing enough light;There must be adequate illuminance to en-able us to do what we want to do quickly andaccurately.

2. limiting glare and extreme contrasts inbrightness;Both these factors ma' not only cause visualdiscomfort but actually reduce our ability tosee,

3. producing the right kind of light;. The major factors here are colour and colour- rendering. Colour rendering is concerned

with the way coloured objects look under dif-ferent types of light sources. If colours usedin the room are such that they are-distortedby the light source, the general effect will bedispleasing.

. A further overall concern is of course theneed to keep down costs and conserveenergy.

urninance

The effect on visual performanceHow much light is required to see what weneed to see quickly and accurately? (For ex-ample, we might be reading a book or verni-er, inspecting a weld, or making a drawing.)Or, to put it in the usual jargon, what illumi-nance is required to perform a given visualtask?

Visual performande depends on manyhuman and physical factors. The human fac-tors include the basic variability from oneperson to another, the effects of age and fa-tigue, and the degree of training or experi-ence a person has in what he or she is doing.A skilled person needs less light than onewho is unskilled, all other things being equal.

The physic t actors affecting visual per-formance are ncerned with what we wishto see - that is, the visual task - and its back-ground or the visual field in which it liesThese factors include the nature of the taskand its inherent difficulty.

Visual performance,is improved if:

- there is good contrast between the taskand its background

- there are no reflections - called veilingreflections - from the task

-there is no severe glare from the luminairesor other light sources (including the sun)

- there is an increase in the illuniinance pro-Vided

Illuminanece recommendationsA person with Formal sight benefits from in-creased illurnihance: he or she sees fasterand more accurately. Thus any illumjnancerecommended for a particular visual task isalmost always a compromise between thebenefits fromincreased illuminance on theone hand and the amount (and cost) of en-ergy to provide that illuminance on the other.Such a compromise is not by4ny meanseasy, and there is a further, complicating fac-tor: in everyday situations, as opposed tolaboratory situations, it is practically impossi-ble to determine the effect on visual perform-

: ance resulting from varying the illuminancewithin quite wide tolerances around its pre-scribed value (i.e., the recommended illumi-nance that applies the time).

The Illuminating EngineeringSociety ofNorth America (IES) publishes recom-mended illuminances for different visualtasks. These reconimendations have beangenerally accepted inCanada. The IES com-mittees responsible for making the resin-mendations are made up of lighting special-ists from'both Canada and the USA.

Illuminance recommendations are notimmutable. The IES iscurrently (1981) in theprocess of issuing new recommendationswhich are to supersede those that have been-in use for the past twenty years.

Continuing research has shown that thefundamental system used to establish illumi-nance recommendaiions cannot be com-pletely and scientifically established. A newsystem based on research is ten years away.Consequently, an alternative approach hasbeen adopted by the IES. It is based not onresearch but on consensus - a majorityagreement of individuals knowledgeable inlighting research and application. Thatagreement was reached after a review of

5

Table 1 /Recommended Illuminances

ClassroomKindergarten' 350 luxGeneral classrooms, seminar rooms 500 luxVisually impaired students 1500 lux

Large group instruction areas(including lecture theatres)2

AuditoriumAssemblies only2When used for study or

examinations

CafeteriaGeneralServeryKitchenWhen used for study or

examinations

DraftingGeneralOn bokds3

Graphic artsGeneralOn boards3

Gymnasium(Gymnatorium arid Cafetorium)GeneralWhen used for study or

examinations

LaboratoriesGeneralDemonstrators' desks

250 lux

250 lux

500 lux

250 lux350 lux500 lux

500 lux

500 lux',INK) lux

500 lux1000 lux

LibraryStacks (with little local reading) 250 luxReading, note taking, cataloguing 500 luxSeminar rooms 500 lux

Offices

Music room

Sewing 3

ShOps3

Typing,

500 lux

500 lux

500 lux

500 lux

500 lux

Washrooms and locker areas 250 lux

CorridorsWith lockers 250 luxWithout lockers3 100 lux

Mechanical and service areas 250 lux

Storage areas 100 lux

Notes:1. The illuminance-of 350 lux has been selectedfor the kindergarten since the visual tasks are notsevere.

2. It should be possible by simple switching tovary the illuminance by 50 per cent eitherup or

250 lux down to accommodate note taking or audio-visualpresentations. Total control by areas is desirable.

500 lux 3. Local lighting may be used to supplement thegeneral lighting where fine work is to be carriedout or feature displayearelocated. Supplemen-tary lighting is frequently provided for chalk-

750 lux boards. Since study carrels cut off light from theoverhead general lighting system, a local lightingluminaire in each carrel is frequently desirable.

500 lux

available vision research and of lighting prac-tices in other countries. The consensus ap-proach of the IES has been used as the ba-sis for the illuminance recommendations inthese guidelines. (See Table 1.)

The previous method of allocating one il-luminance as the only one appropriate for aparticular visual task has been dropped. In-stead, visual tasks are grouped into a seriesor categories with a range of three illumi-nances applicable to each category. The ap-propriate Illuminance is selected from therange of three by means of a system ofweighting factors These factors recognizethat working speed and accuracy are im-proved with more light; that more light is

needed if contrast in the visual field is poor,and that more light is needed as one getsolder. This last point is important if class-rooms are to be used by adults as well as byyoung students.

These weighting factors have been con-sidered in selecting the illuminances forschool spaces shown :n Table 1.

'I

The advantages of providing more thanone illuminanceThe choice of illuminances to be used has adirect effect On the electric power (i.e., thewatts) drawn by the lighting system and onthe energy consumed. The choice is thus ofgreat importance in these days of energy,management and high energy costs. If an il-luminance of, let us say, 300 lux can be usedinstead of 500 lux, then there is straightawaya potential reduction of one-third in theamount and cost of the energy consumed,

The choice of illuminance will depend(among other things) upon the visual task.The moredifficult the task, the greater the il-luminance required.

It is not always possible, however, todesignate classroom use at the time thelighting systems are designed, and thus todesign each system to suit a particular use.Accordingly, the practice has been to providea general lighting system with one fairly highilluminance throughout the general class-rooms and spaces. This takes care of mostdifficult visual tasks such as reading a paper-back book or pencil handwriting. But the factis that a kindergarten class, with no difficultvisual tasks, can operate perfectly well with alower illuminance. With such a general light-ing system, therefore, it is inevitable thatsome classrooms will be provided with ahigher illuminance than they need. As a re:suit, more energy will be consumed than isnecessary.

A way around this problem is to considera lighting system capable of providing twolevels of illuminance. This can usually beachieved by simple switching, by the use ofspecial ballasts which provide two light out-puts from fluorescent lamps, or by the use oflamp-dimming circuits. These techniques arediscussed in sections 3 and 6.

Equivalent Sphere Illumination (ESI)Providing the recommended illuminance fora particular visual task is no guarantee thatthe task will be seen well. This may seem anastonishing statement to make it raises thequestion of why there should be recom-mended illuniir Ices at all. But we have allnoticed the eff_ct of a patch of sunlight on achalkboard the reflected glare makes it im-possible to read the writing on the boardeven though the light level is more thanenough.

A similar, but much less obvious, phe-nomenon occurs where there is reflection ofa large luminous area in the print or on thesurface of a page, even where the text i6printed on dull or matt paper. The effect -very possibly undetectable by the naked eye

is as though a luminousypil were superim-posed over the visual task. Hence the name"veiling reflections". Veiling reflections cre:ate the most insidious problems on semi-glossy pages with print or pencil writing.They reduce the -contrast-betweemthe-pri htor writing and the page, making readingmore difficult.

Hbw well one sees a visual task dependson more than lust providing the right illumi-nance. It depends on contrast. This, in turn,depends upon ne reflection characteristicsof the visual task, the locations of the lumi-naires and the observer with respect to thevisual task, and the distribution of light fromthe luminaires.

The above considerations have led tothe concept of Equivalent Sphere Illumina-tion (ESI). ESI is a method of specifying illu-minance in such a way that it takes contrastinto account. Since, as we have seen, con-trast affects visibility, the method has the po-tential of being a more effective way of indi-cating how well a lighting system enablesone to see a particular visual task than amethod in which only the illuminance (mea-sured in lux) is specified.

In recent years ESI has been used tospecify the illuminance for a number of visualtasks. This booklet, however, does not rec-ommend this practice:Pie main reason forthis is that insufficient experience has beenobtained with ESI to permit the inclusibn ofESI values in the consensus approach to illu-minance recommendations.

Further, in order fo obtain the maximumESI on a visual task, that task must be care-fully iocated with respect to the luminaires inthe lighting system and to the windovA..Inmany classrooms the working positions arechanged frequently as the work and the com-position of the class are changed. So there is

guarantee that the resultant ESI on thevisual task is at its desired level.

Finally ESI requires a computer pro-gram for its design calculations. It also re-quires a very special photometer and a sec-ond (but smaller) computer program toconvert the photometric measurements intoESL

7

c

8

Lighting for visually impaired studentsAbout one-half of school-age children whoare legally blind, as registered with the Amer-ican Printing House for the Blind, have usefulvision at the near point and are able to useprinted material as their means of learning.Thus many a legally blind child wilt become amore effective and efficient reader of thatmaterial if attention is-paid to the visual envi-ronment in which he or she works.

It is generally accepted that most visuallyimpaired students benefit from increased illu-minances, although exceptions to this arerecognized - for example, people with achro-rmaopsia and cataracts. An illuminance of1500 lux has been recommended by the IESfor classrooms for the visually impaired, al-though higher illuminances have been used.This illuminance about three times thatadopted-for many general classroomsmight be from a general overhead lightingsystem or from a combination of gerierallighting plus supplementary lighting on thework.

Some visually impaired students readwith their eyes close to the printed page. Thelighting system for these students must bedesigned so that the light is,,not severely ob-structed by the student's head: the light mtistfall on the page.

SUMMARY..

- Deiign the lighting system to maintain theappropriate illuminances on the visualtasks. Refer to the recommended illumi-nances shown in Table 1.

,Try to ensure that visual tasks'of good con-trast are used; if this is not possible, locatethose tasks so that bne does not seereflections off them.

Consider a lighting system capable of pro-viding more than one level of illuminance ifthe space is to be used by adults as well aschildren.

The use of Equivalent Sphere Illumination(ESI) i4ot recommended.

A higher illuminance than normal is recom-mended for students with impaired sight.The lighting design must also take into ac-&writ the possiliilitP that these studentsmay obstruct the light by sitting with theireyes close to the visual task.

Glare

Glare occurs when the luminance - the"brightness" - of some parts of the field ofview is substantially greater than that in otherparts.

Glare can take two forms} First, it can re-duce the ability to see. This is known asdisability glare. It commonly occurs in build-ings when a bright light is reflected off a sur-face. For example, the sun shining on achalkboard can cause disability glare fromcertain parts of the classroom it may be im-possible to see anything that is writtentrn theboard. Teachers may be unaware of thisprobm. They place themselves in a positionwhere they can see well, little realizing thevisual difficulties the students areexperiencing.

Secondly, it can cause discomfort.Discomfort glare can be presAt without re-ducing the ability to see. This is particularlytrue in very bright surroundings which resultfrom a liberal 'display of bright luminaires. Lit-tle or no disability may be caused, but thediscomfort may be marked. Dispomfortglarecan also be caused by daylighting if the win-

,dows provide a clear view of the sky.

Discomfort glare is a function of the lumi-nance and the size of the luminaire, and:of,.the number of luminaires in the field of view.The brighter the luminaire, orthe bigger theluminaire, or the greater the number of lumi-naires seen, the greater the discomfort glarethat will be caused. The influence of the lumi-nance of the luminaire is particularly signifi-cant. If the,luminance is doubled, but allother factors are held constant, the glare ef-fect will be more than trebled.

The coloul of die surroundings is also acontributing factor. Discomfort glare can bereduced if light colours are used. A luminairethat can cause it dense discomfort in a spacewith dark walls may be quite innocuous if thewalls are painted a light cplizyThis is a sim-ple and effective way of reducing discomfortglare, but there is one possible drawback:

'Care must be taken to ensure that the wallsthemselves do not become too bright mg,thus cause discomfort glare. There have, forinstance, beeqomplaints from typing stu-dents-who have had to sit facing a brightwall.

A

1, 4--_

_

Finally, discomfort glare is influenced bythe positions of the luminaires relative to theline of sight. A luminaire right in the line ofsight can cause cohsiderable discomfortglare. A luminaire out of the line of sight will,of course, cause no discomfort glare.

Glare Lan be cauied by any form of lig' s.ing, including indirect lighting. Both direct rand indirect H'D lighting systems, in particu-lar, should be studied for their potential tocause glare:.

sr

Indirect lighting has a soft, diffuse quali-ty. The lamp is completely screened from -

° view, and all the light is directed onto the ceil-ing which reflects it down to the desks. Nev-ertheless, indirect lighting can Cause glare..Unless the lumina -es are well chosen andcorrectly spaced apart, very bright soots mayappear on the.ceiling directlY above the lamp

particularly if HID lamps are used:Thesebright spots may cause glare`

A number of methddsliave been devel-ped to s ify the discoMfort glare charab-

eristics o a lighting s tem. In North Amer-ica the Visual Comf rt Probability (VCP)method it used.

Suppose t I fi uniform g&eral.systprn is des' ned to give a particular illumi-nance usin a particular luminaire in a roomwith part lar,decoration. VCR data pro-Vided fo hat luminaire will show the percent-'age of ple who, seated in the most uncle-

Nsira e location (i.e., with the most glare),w Jld find the lighting system acceptable.

When first designed, the system mightshow a VCP of only 50 per cent. This is prob-ably too low: a figure of 70 per cent is fre-quently Used, although the adoption of leis

0, value is arbitrary. So the system would haveto be redesigned, using, possibly, luminairesof 17er luminance, in order to increase theVC to 70 per cent.

, -The limitations of VCP are that it appliesonly to (1) general overhead lighting systemsilluminating horizontal surfaces, such asdesk tops; and (2) the location in the roomwith themorst discomfort glare effect. Thus, ifthe work surface is not horizontal, if the lay-out of the working positions changes fromtime to time, or if the uniform general lightingsystem is not used, VCP has little or novalue.

The is, however, a simple way of ap-praising an existing lighting system for dis-comfort glare. Select typical locations in theroom and shade the eyes with the hand heldhorizontally. lf, as a result, the lighting sud-

.denly becomes me comfortable, then thelighting system is likely to be too glaring. Ob-viously this method cannot determine withany precision by how much the discomfortglare should be reduced, but it is still a validtest and one that is not subject to the liniita-tions of VCP that chmoe been mentioned.

1 2

SUMMARY

- Glare can take two forms disability glareand discomfort glare.

- Disability glare reduces the ability to see. Itcommonly occurs when a bright light isreflected off a surface. The sun's rays, forexample, can easily cause disability glare.

- Discomfort glare is a function of the lumi-nance and size of the luminaires, and of thenumber of luminaires in the field of view.

- Glare can be causes by any form of light-ing, including, perhaps surprisingly, indirectlighting. Study both direct and indirect HIDlighting systems, in particular, for their po-tential4o cause glare.

- Visual Comfort )bability (VCP) is amethod of spearying the discomfort glarecharacteriStics of a lighting system used inNorth America. It has little or no value if thework surface is not horizontal, if the layoutof the working positionschanges frequent-ly, orif a uniform general overhead lightingsystem is not used. .

- Study a lighting system by shielding theeyes with the hand. If, as,a result, the light-ing suddenly becomes more comfortable,then the system is likely to be too glaring.

The appropriate illuminance alone is noguarantee of a classroom that is a cheerfuland pleasant place in which to learn. The col-our and brightness of the principal surfaces -ceiling, walls, floor; and furniture - also playtheir part in creating the classroom environ-ment. Brightness, in particular, determineshow comfortable a classroom is visually. Aclassroom war darkwalls-and dark-fah-au re-,for example, will be gloomy whatever the illu-minance.

A proper balance of brightness levelsmust be maintained. If there is a substantialdifference in brightness between the visualtask and the area immediately surrounding it,or between large adjacent areas in the Stu-dents' field of view, then they may well expe-rience visual discomfort.

The luminance of the principal surfacesis determined by the lighting system and by

what are called the reflectances of thosesur-faces. Recommended reflectances for theprincipal surfaces are shown in Table 2.Their use will ensure that there is an appro-priate balance of brightness levels and willthus help the designer to create a space thatis visually comfortable. Desk tops, for exam-ple, should have a reflectance of betweenabout 35 and 50 per cent. They will then belower in brightness but not much lower -than the brightness.of most school tasks. Avery dark or a glossy white desktop shouldnot be used.

The recommended reflectances are suffi-ciently high to ensure good utilization of thelight falling on them. Much of this light will bereflected, contributing to the illuminance inthe classroom.

Table 2/Recommended Reflectance of Sur-faces in Learning and Administrative Areas

Ceiling 70-90%40-60%

Floor 30-50%Desk top 35-50%Chalkboard up to 20%

Note:1. The reflectance of feature and special walls candeviate from this range. The consequence.of thismust be carefully considered to ensure that thedesired overall environment is retained.

Although extreme brightness contrasts inthe classroom must be minimized, the use 4colour contrasts is recommended. These willensure that a lively, rather than a bland, envi-ronment is produced.

SUMMA_RY

- The Iiimiiiance of the principal surfacesgreatly influences the visual comfort of aspace. The luminance is determined by thelighting system and by what are called thereflectances of the principal surfaces. Thevalues of reflectance in Table 2 should beused: they will help to create a space that isvisually comfortable.

- Although extreme brightness contrastsmust be minimized, colour contrasts arerecommended.

Figure 1 'Comparison of Spectra from Cool White and Deluxe Cool White Fluorescent Lamps

250

E

200

o150

U)

g 100

250

COOL WHITEDELUXE COOL WHITE

300 400 500 600 700

Ultraviolet Bfue Green R ed Infrared

Wavelength inNanometers (nm)

If the radiant power is measured in the spectrum from a lamp, a diagram such as the one above can bedrawn. This one shows the radiant power emitted by a Cool White (CW) fluorescent lamp and a DeluxeCool White (CWX) fluorescent lamp. In the CWX lamp, the power in the red part of the spectrum is greaterthan that in the CW lamp, while the power in the yellow part is reduced. This results in the better colour ren-dering properties of the CWX lamp: reds seen under the CWX lamp are not greyed as when seen underthe CW lamp.

Colour appearance, colour rendering, andthe choice of coloursused in decoratingWe know from our everyday experience thatthe way we react to a school space isinfluenced by the colours used in its decora-tion. Green interiors - a relic of institutions forthe Lonaon poor in Victorian times - havebeen superseded by a wide range of lively,colourful interiors. Yet these colours do notalways produce the expected effect. Onecause for complaint is thatthe colours useddo not work well with the particular lightsources under which they are seen. We musttherefore try to ensure that the colours wechoose are compatible with the light source.

The appearance of the light source is nohelp in this choice. The colour of the lamp isno indication of how coloured objects will ap-pear under light from that lamp. Moreover,two lamps may look the same colour thatis, they may have the same colour appear-ance and yet make the same coloured sur-faces look different. For example, Cool White(CW) and Deluxe Cool White (CWX)fluorescent lamps look much the same. Butunder the CW lamp, yellows become particu-larly vivid and reds are greyed: under theCWX lamp these effects do not occur.

11

The effect of light on coloured objects isknown as colour rendering. The Colour Ren-dering Index (Ra) gives an indication of howclosely a given light source matches the col-our rendering ability of a reference source.The reference source always has an Ravalue of 100; the Fla of the given light sourcevaries, for example, from a value in excess of90 for fluorescent lamps with very good col-our rendering, to about 70 for Cool Whitefluorescent lamps, and down to about 20 forhigh pressure sodium lamps.

There has been a tendenCy to use IR, asa figure of merit in competing one lamp withanother; instead, it must be used with cautionand the following points borne in mind:

Ra will be meaningful as a comparisbn onlyif it is used with lamps that have substan-tially the same colour appearance that is,with lamps that look substantially the same.Ra will not be meaningful as a comparisonbetween, say, Cool White and Warm Whitefluorescent lamps: these lamps do nothave the same colour appearance.

If the Ra is less than about 90, then thecomparison of one Ra with another is ofdoubtful value.

The Ra is not a particularly useful guideto the selection of lamps for a specific light-ing purpose. The only effective way of goingabout this is to see the lamp in use and tonote how well it renders colours.

Similarly, the only effective way of select-ing a colour for decoration is to see it underthe light source that will be used to illuminatethat decoration. Unpleasant-looking interiorscan result if this simple procedure is not fol-lowed. Ye the procedure can easily be over-looked when repainting a space.

Special care must be taken in the selec-tion of the same colours to be used in adja-cent spaces illuminated by different lightsources. In these conditions, the differencein the appearance of the colours will be par-ticularly noticeable.

Colour rendering and the choice of lampsfor school activitiesGood colour rendering lamps are essentialfor school activities in which the effect of col-our has a major impact on the student's re-sponse. Art classes are an obviJus example.The same applies to cooking classes and tothe cafeteria: even thelastiest food can lookunappetizing under the wrong lamps.

12

0

The use of good colour rendering lampsis also particularly important in the lowergrades where children are taught the recog-nition of colours and their relationship to ob-jects'in the world. The children must see col-ours as they are, otherwise it is possible thattheir learning will be hindered.

The Lamp Selection Guide (Table 3)summarizes the colour rendering propertiesof typical lamps suitable for school use.

SUMMARY

The colour of the lamp is no indication ofhow coloured objects will appear underlight from that lamp.

'The effect of light on coloured objects isknown as colour rendering. The ColourRendering Index (Ra) gives an indication ofthe colour rendering properties of lightsources.

There has been a tendency to use the Col-our Rendering Index as a figure of merit incomparing one lamp with another. Instead,;t must be used with caution: it is not a par-ticularly useful guide in the selection of alight source for its colour rendering proper-ties.

The effective way of selecting a colour fordecoration is under the light from thesource that will be used to illuminate thatdecoration. Unpleasant-looking interiorscan result if this procedure is not carriedout.

If thesame colours are to be used to deco-rate adjacent spaces illuminated by differ-ent light sources, then the difference in theappearance of those co_urs under the dif-ferent sources will be particularly notice-able.

Good colour rendering lamps are essentialfor a number of school activities art, cook-ing, and teaching the lower grades and inthe cafeteria.

3. Lamps and Ballasts

Lamps

Three categories of lamps are used forschool lighting: incandescent, fluorescent,and high intensity discharge. The Lamp Se-lection Guide (Table 3) gives the designcharacteristics of lamps in eachcategory.

Incandescent lampsThe incandescent lamp is the one in com-mon household use. It consists of a filamentof fine, coiled tungsten wire enclosed in aglass bulb which is usually filled with a mix-ture of nitrogen and argon. Electric currentpassing through the filament heats it until itgives off light.

One variation of the incandescent lampis the tungsten-halogen lamp. Here the fila-ment is usually in an atmosphere of iodinevapour instead of the nitrogen and argonmixture. The effect of the iodine vapour istwo-fold: it prevents the lamp bulb from dark-ening as quickly as in other incandescentlamps, and it prolongs the filament life. Con-sequently, the light output is maintained bet-ter and the lamp life is longer than in a stan-dard incandescent lamp. The tungsten-halogen lamp may be used for display pur-poses. Ho Wever, it costs much more than astandard incandescent lamp of approxi-mately the same wattage.

The efficacy of the incandescent lamp-that is, the lumens of light emitted per watt of

"power - increases as the lamp wattage in-creases. Also, the efficacy and the lamp lifeaffect one another. Efficacy can be increasedif a shorter lamp life is acceptable; a longerlife can be achieved by a reduction in effi-cacy. For example, a standard 100 W lamphas a rated life of 750 hours and the ex-tended-life 100 W lamp a rated life of 2500hours. The increase in life is achieved at theexpense of a drop in efficacy from 17.5 ImANto 15 ImAN.

The rated life of an incandescent lampdoes not mean that every lamp, when oper-ated correctly, will burn for that length of time- for example, 750 hours in the case of thestandard 100 W lamp. Some lamps will burnfor a longer time than the rated life and somefor a shorter time. At the rated life, 50 percent of the lamps in a large installation willhave failed. This is a significant point to bearin mind when considering lamp replacement.

All too frequently it has been assumed thatall the lamps will burn for the rated life, andthe lamp replacement program has been-based on that assumption. In addition to theproblems caused by- incorrect estimates oflamp replacement costs, this results in burnt-out lamps appearing in an installation. Thesewill spoil the look of the installation and alsoreduce its effectiveness.

Some luminaires may be olierated withmore than one type of lamp and possiblymuch more efficiently with one lamp typethan with another. For example, a baffledownlight will produce about the samenance with a 75 W ellipscidal reflector lampas with a 150 W reflector flood lamp.

Fluorescent lampsThe fluorescentlf- ip consists of a glass tubecontaining mercury vapour and with elec-trodes at each end. The inside of the tube iscoated with a powder (i.e., the phosphors).An arc flashes to and fro between the elec-trodes, producing invisible ultraviolet radia-tion together with a little visible blue light.The radiation falls on the phosphors coatingthe tube, causing them to fluoresce and giveoff light. Most of the ultraviolet radiation isthus absorbed by the phosphors and theglass tube; little is emitted by most lamps.

Light of almost any colour can be pro-duced by using the appropriate phosphors:something like three dozen "white" lampsalone appear in the catalogues. It has al-ready been mentioned that two fluorescentlamps may look the same colour yet have dif-ferent colour rendering properties. Usually,but not always, the lamp output falls as thecolour rendering improv"s.

Fluorescent lamps range in size from4 W to 215 W. Reduced wattage lamps havebeen introduced in the past few years. Theyuse 10 to 20 per cent less power than thecorresponding standard fluorescent lamps,depending on size. One such lamp is the 4-foot 35 W Rapid Start lamp which can besimply substituted for the common 4-foot40 W Rapid Start lamp without change ofballast.

A fluorescent lamp cannot be operatedwithout a ballast - usually two lamps fromone ballast. Special ballasts are available bywhich fluorescent lamps can be dimmed.There are also multi-level ballasts which en-able the light output to be reduced by a sim-ple reconnection of the ballast leads.

13

Table 3 LampSelection Guide

Notes.(a) According

to type, 3hdurs perswitching.

(b) Accordingto type.

(c) Accordingto type, 10hours perswitching.

(d) 5 hoursperswitching.

14

LLamp Name

SunlightSiMulating andColourMatchingFluorescent

.

Cool White.

Deluxe ( :

Cool White I

FluorescentWarm WhiteFluorescent

Approx.CorrelatedColour Temp.(K)

5000- 5500 4400 4000 3100

Lamp ColourAppearance Bluish- White White

,

_White Yellowish - White.

ColoursStrengthened

AllNearlyEqual

OrangeYellowBlue

AllNearlyEqual

OrangeYellow

ColoursGreyed

NoneAppreciably Red

NoneAppreciably

RedGreenBlue ,

Approx. InitialLuminousEfficacy (Im/W)

55-72 80 - 55-72 80'

Approx. RatedLife (Hours)

(a)

12 000- 20 000(a)12 000- 20 000

(a)12 000- 20 000

(a)12 000. 20 000

Remarks Very good colourrendering. Lightappears cold atlow illuminances.

.

Acceptable formost applicationswhere colourrendering notimportant. Lightappears cold atlow illuminances.

Use where goodcolour renderingimportant. Lightappears cold atlow illuminances.

Acceptable formost applicationswhere colourrendering notimportant. Lightappears warm.

..

The life of a fluorescent lamp - unlikethat of an incandescent lamp - is affected bythe number of times it is switched on and off.The life is rated by the number of hours thelamp is allowed to remain alight after beingswitched on. the longer it remains alight thelonger the life. For example, the life may berated at 20 000 hours, assuming that thelamp operates for three hours per switching.As with incandescent lamps, at the rated life,50 per cent of the lamps in a large installationwill remain alight

As a general rule, if fluorescent lampsare not to be used for five minutes or more.they should be switched off.

Dummy lampsA dummy lamp looks much like a standardfluorescent lamp but it gives off no light. It isa device which can be substituted for one ofthe two lamps (usually 4-foot 40 W RapidStart lamps) operated by one ballast: the

two-lamp ballast remains without alterationto the circuit, but only one lamp emits light.The use of a dummy lamp results in a sub-stantial drop - of about one-third - in the lightoutput from the other lamp. There is also adrop of about three-quarters in the powerdrawn by the two-lamp ballast.

High intensity discharge (HID) lampsThere are four types of HID lamps: mercury,metal halide, high pressure sodium, and logypressure sodium.

HID lamps, with the exception of the lowpressure sodium lamp, have been found ac-ceptable for school locations where colourrendering is not important in gymnasia andworkshops, for example.

1-.

1 4

4

DeluxeWarm WhiteFluorescent Incandescent ...

.

DeluxeMercury

PhosphorCoatedMetalHalide

HighPressureSodium

.

LowPressure .:,Sodium ,

3000

.

3000 - 3400 3500 2200

-

Yellowish- White Yellowish- White Purplish- White Purplish-White Yellowish Strong Orange

Red, GreenOrangeYellow

RedOrangeYellow

RadYellowBlue ,

RedYellowBlue "

YellowOrangeGreen

YellowOrange

Blue.

Blue Green, GreenRedBlue

All Except yellowand Orange goaBrown-Bla ck

55-72 20

.,

55 853

120 180

', (a)

12 000- 20 000(b)

750-2500(c)

16 000- 24 006(c)

7500-20 000(c)24 000

.(d)18 000

Colour renderingnot quite goodenough Where .

colour renderingimportant. Lightappears warm.

. .

Good colourrendering.

6

Similar torema or coolwhit luorescentlard .

Similar toremarktior coolwhite fluorescentlamp.'

Similar toremarks for warmwhite fluorescentlamp, but lamp

1 with poorercolour rendenng.

_ -,

Poor colourrendering makesuse unlikely forschool interiors,but may be usedfor exteriors.

Mercury lampsA mercury lamp consists of two envelopes(the outer glass bulb and the inner (and muchsmaller) arc tube. The arc tube contains mer-cury vapour which emits light when an elec-tric current passes through it forming an arc.Ultraviolet radiation is also emitted. This radi-ation falls on phosphors coating the inside ofthe outer bulb, causing them to emit light.The light from a mercury lamp is thus a com-bination of that from both the arc tube andthe phosphors.

Metal halide lampsThe metal halide lamp is similar in construc-tion to the mercury lamp. The main differ-ence is that its arc tube contains metal hal-ides in addition to the mercury vapour. Thesechange the colour of the light emitted fromthe arc tube from the blue of the mercury arc

to white. The colou further improved insome lamps - thos that would probably beused in schools py coating the inside of theouter bulb wi osphors, as in the mercurylamps..

The efficacy of the metal halide lamp is11/2 to 2 times that of the mercury lamp.

High pressure sodium lampsThe high pressure sodium lamp lias two en-velopes. The outer is of glass and the innerarc tube, which contains sodium, is of alumi-na, a material that is much more resistant tosodium attack at high temperatures thanglass. The light from this lamp is usually de-scribed as being "golden-white".

15

Low pressure sodium lampsThe low pressure sodium lamp has two en-velopes. The inner the arc tuba is of spe-cial non-staining glass. It contains some met-allic sodium and a gas, usually neon. Theouter envelope is a vacuum jacket, similar inprinciple to a Thermos flask. Its purpose is tokeep the discharge in the arc tube at its opti-mum operating temperature.

The low pressure sodium lamp is notsuitable for school lighting. It gives off yellowlight which turns every colour it falls on eitheryellow or muddy.

Operating characteristics of HID lampsA ballast is required to operate HID lamps.Ballasts are available which enable somelamps to be dimmed.

With the exception of the low pressuresodium lamp, all HID lamps have similar op-erating characteristics. Each lamp takes afew minutes (from one to seven according tolamp type) to come,up to full light output afterbeing switched on. Then, it power isswitched off- momentarily andswitched on

_again, it will take from two to fifteen minutesbefore the lamp gives off its full light output.Thus HID lamps must not be used wherelight is to be provided immediately at the flickof a switch. Where HID lamps are used, afew incandescent or fluorescent lamps,which en-lit their full light output immediatelyon switching, are frequently included in theinstallation. These provide some light duringthe period in which the HID lamps are build-ing up to their full light output.

The life of an HIDdlamp is much more af-fected by switching than that of a fluorescentlamp. with HID lamps, in particular, the lessthey are switchea on and off, the longer theirlife. The life is rated by the number of hoursthe lamp is allowed to remain alight after be-ing switched on. For example, the life of amercury lamp may be rated at 24 000 hours,assuming that the lamp operates for tenhours per switching. At the rated life, 50 percent of the lamps in a large installation willremain alight.

As a general rule. if HID lamps (exclud-ing low pressure sodium lamps) are not to beused for one hour or more. they should beswitched off.

SUMMARY

At the rated life of a lamp, only 50 per centnot 100 per gent of the lamps in,a large

installation will remain alight.

The life of fluorescent and high intensity Ndischarge (HID) lamps is affected by thenumber of times they are switched on andoff.

The life of HID lamps is much more afrfected by switching than that of fluorescentlamps: with HID lamps, in particular, theless they are switched, the longer their life.Switch off fluorescent lamps if they are not'going to be used for five minutes and HIDlamps if they are not going to be used forone hour. Incandescent lamps may beswitched at will: their life is not affected byswitching.

HID lamps take a few minutes to come upto full light output after being switched on.They must not be used alone wherd light isneeded immediately at the flick of a switch.A few incandescent or fluorescent lamps inan installation of HID lamps Will providelight while the latter lamps are building up

.their light output.

Refer to the Lamp Selection Guide (Table3) when choosing a lamp for a particularschool area.

Ballasts

A ballast is required to operate bothfluorescent and HID lamps. The device hastwo functions. First, it limits the flow of cur-rent through the lamp.: without the ballast thecurrent will increase and very quickly destroythe lamp. Secondly, it provides a voltage suffi-ciently high to start and maintain the arc dis-charge.

The heart of a present-day ballast is acore made of thin steel stampings sur-rounded by a coil of copper wire, similar to atransformer. It is anticipated, however, thatwithin the foreseeable future, solid state bal-lasts will supersede the present type of 41-last.

Ballasts consume energy. For example,there is a loss of about 16 W in a standardballast for two 40 W Rapid Start fluorescentlamps: the total power drawn by the ballastplus lamps is thus 96 W..

New types of ballast, which offercertainadvantages over the standard ballast; arenow available. Some of them are as follows:

- A ballast for two 40 W Rapid Startfluorescent lamps which has a loss ofabout 9 W instead of the 16 W of the stan-dard ballast: the total wattage drawn by theballast plus lamps is thus 89 W. There is noloss of lamp light output with this ballast,but it costs more than the standard ballast.

- A ballast for two 40 W Rapid Startfluorescent lamps which, with the lamps,

. draws total power of about 67 W. There is adrop in lamp light output of about 30 percent.",

- A ballast for two 40 W Rapid Startfluorescent lamps which can provide either100 per cent or about 60 per cent of thelight output from the lamps by a simple re-connection of the ballast leads. The powerdrawn by the ballast plus lamps will beabout 96 W or 57 W respectively.

Ballasts for.use in dimming circuits forfluorescent, mercury, and metal halidelamps.

Solid state fluorescent lamp ballasts willbe available in the foreseeable future. Thesewill have the following advantages comparedwith present-day ballasts:

- much lower wattage loss

- the ability to operate the lamps at high fre-quency so that they give off more light - thelight output from a fluorescent lamp in-creases with the lamp operating frequency

- no audible hum

All present-day ballasts hum to a greateror lesser extent. The hum is usually, al-though not always, louder as the lamp wat-tage increases, and may be amplified by theway in which the ballast is mounted in the lu-minaire.

Ballast hum can be disturbing, althougha hum that disturbs the quiet of a library willnot be heard in a woodworking shop. Guid-ance is given by individual manufacturers onthe suitability of their.ballasts for use in vari-ous types of locations, some noisy, somequiet. There is, however, no industry stan-dard indicating the degree of hum made bydifferent types of ballast.

SUMMARY

A ballast is required to operate fluorescentand.FilD lamps. In addition tofthose for nor-mal lamp operation, special ballasts areavailable which reduce the wattage drawnby (and thus the light output from)fluorescent lamps, and which permitfluorescent and HID lamps to be dimmed.

Ballasts consume energy.

- Present-day ballasts hum to a greater orlesser extent. The hum can be distracting.

17

PHOTO, OMER 1URZOZUS

4. Windows and Daylight

Windows have usually been' considered anadvantage in a school. They let light in andenable students and staff to see out. There isan instinctive desire for contact with the out-side. Yet, by contrast, teachers in a wittdow-less school have referred approvingly to "nodirect sunshine; no glare; no outside disturb-ances noise, weather conditions".

The effectiveness of Windows 'How effective are windows in modernschools in providing light by which to workand study? There is a great need to collectdata on the amount of daylight available forinterior lighting and the extent to which it canreduce lighting and HVAC" costs. There isalso a need to improve the techniques bywhich interior daylighting is predicted: anyprediction using existing techniques will be oflimited accuracy.

With these comments in mind, considera traditional cellular classroom about 9.1 m ,

(30 ft) by 7.6 m (25 ft) by 2.7 m (9 ft), withwindows 2 m (6 ft) high from sill to ceilingrunning along one wall. For many normal-length working days, there will bwenoughdaylight by which to work over about a 3 m(10 ft) width of the classroom measured infrom the windows. Thus if, as is common, theclassroom is illuminated by three rows offluorescent luminaires parallel to the win-dows, the row nearest the windows can beswitched off much of the time.

Windows are capable of producing both disa-bility and discomfort glare. The former, whichresults in reduced visibility and visual per-formance. is caused by direct sunlight or thereflection of that sunlight. The latter may becaused by a view of clouds or the blue sky -often referred to as "sky glare". Attempts toalleviate sky glare by reducing the windowsize may result in an increase in discomfortdepending upon the location of the wirtdows.Blinds and drapes are usually a more effec-tive means of controlling that glare.

Heating, ventilating, and air conditioning.

Direct sunlight also causes another prob-lem: a build-up of heat in the classroom. Onesquare metre of unshaded window can trans-mit solar radiation equivalent in heatingpower to one square metre of steam radiator.Moreover, a change in the characteristics ofthe solar radiation, which occurs when it fallsupon interior surfaces, traps much of the ra-diation inside the room - it is unable to es-cape back through the window. As a result,the temperature in the classroom rises, aneffect known as the "greenhouse effect". (Onthe other hand, some of the heat actuallypro ,uced in the classroom - by the HVACsystem, by the lighting system, by the peoplein the room - will escape to the outside un-less something is done-to prevent it.)

Sunlight mayi be controlled by externallouvres, by interior drapes, and by devicessandwiched between double glazing. Lou-vres are the most effective: it is always bestto control the Iun before it reaches the glass.The possibility of maintenance problemsfrom ice and snow must be borne in mindwhen considering the use of externallouvres.

In the northern hemisphere, both glareand heat build-up may be expected fromsouth- and west-facing windows. Heat fromthe latter windows can be of unexpected se-verity since the rays of the sun pass throughthe glass with only little loss by reflection atits surface. Horizontal louvres can be de-signed for south-facing windows so that thesun is screened in summer (when high in thesky), but is allowed to reach the windowsnwinter (when at a low angle). Vertical shad-ing-devices work-best-en-windows-facing-ap-proximately east and west.

Inadequate window design and the lackof screening devices could mean that day-light cannot be used, simply because of thesun glare. Drapes will have to be drawn forvisual and thermal comfort, and electric light-ing used instead. ,

Any screening device - particularly inte-rior drapes must be so chosen that it doesnot become too bright and cause glare whenthe sun shines on it. On the other hand, atnight the device must not appear dark, other-wise it will create excessive contrast with theother surroundings in the room.

19

S.

Aspects of Day lighting Design

In a new school, windows%and daylight will beconsidered as an element in lighting designadditional to the electric lighting. Good day-lighting design does not simply mean largewindows. It must be approached both quanti-tatively and qualitatively on broad and sensi-tive design terms.

The quantptive design of windows isstraightforward. It involves:

- building orientation - the direction the win-dows face

- the size, shape, and glazing of windows

- sunlight control _

- the trade-off between energy saved byusing daylight instead of electric light andthe heat loss and heat gain through thewindows

- an assessment of the cost of vandalism

20

Life-cycle costing may be applied to thewindow design so that the windows providethe best return on investment. Such costingwill make it clear that daylight is not free. .

In contrast to the quantitative design ofwindows, tlfe qualitative design is complex.Consideration must be given to matters forwhich there are no firm design guidelines. Sofar the significant conclusion of studiesaimed at evaluating whether or not schoolsshould have windows is that there is rio sig-nificant conclusion.

The advantages of the windowlesVschool are that it eliminates window break-age and outside distractions-and reducesvandalism, excessive glare, and wasted wellspace. 'ts disadvantages are viewed as pri-marily psychological. Ithas been percep-tively pointed out that people bring their atti-tudes and prejudices to a building. Their pastexperience is an essential part of the percep-tual translation and one ignores it at one'speril. Why should we want to ignore thesedeeply held beliefs?

If staff and students believe that roomssho'ild have windows, then it seems likelythat rooms will have to have windows. How-ever, considerations of both the qualitativeand quantitative aspects of design may showthat the windows can be small and providedas a pleasant amenity rather than as a maincontnbutor to the lighting of the room.

5. Some Effects of ighting 0. a

Newton's sketch show-ing the formation of aspectrum.

Reproduced by kindpermission of theBodleian Library,Oxford

Studies on the effect of light on humans areinconclusive. Although experiments on'ani-mals have received considerable scientificdiscussion, it is difficult to understand howsignificantly these findings can be applied tohuman beings. Our understanding of boththe healthful and harmful effects of light hasbeen influenced by conjectural and specula-tive reporting. The comments below outlinethe present state of knowledge on some ofthe effects of light on health.

But first, a general comment on the na-ture of electric light and sunlight: As Newtonfound in his famous experiment of 1666, arainbow or spectrum is formed when sunlightis passed througtta prism. The wavelengthsof the visible light forming that spectrum ex-tend from 380 nanometers (blue light) to 760nanometers (red light). Ultraviolet radiationand infrared radiation are also found in sun-light. It is important to recognize that all of thewavelengths of any lamp, both old and new,are present in sunlight. Not every lamp emitslight in all the wavelengths present in sun-light, but no lamp emits light of a wavelengththat is not in sumlight,1

Ultraviolet radiation and healthUltraviolet (UV) radiation affects our health(and also causes suntan). It is present in thesun's rays as well as in many commonlyused lamps, including fluorescent, HID, andsome tungsten-halogen lamps. Not all lampsemit the same UV radiation per watt of pow-er. There is even considerable variation inthe UV radiation from the various types of thecommon 4-foot 40 W fluorescent lamp. (SeeAppendix 2 for additional statements onultraviolet and health.)

The design of a luminaire and the materi-als used in its construction determine theamount of UV radiation it emits. For example,anodized aluminum is a good reflector of theradiation, whereas glass and many paintsabsorb it. Thus an open luminaire with an an-odized aluminum reflector will emit much ofthe UV radiation from the lamp; a luminairewith a painted reflector and 4a glass front willemit little of the radiation.

The absence of UV radiation can beharmful to health, since it can cause a defici-ency in vitamin D, the vitamin needed to de-velop strong bones. In severe cases, bonediseases such as rickets may result. Ricketsis in fact the earliest air pollution disease: it

'I A*

21

22

was first described in England in 1650. It hasbeen directly attributed to smoke from thecoal fires in Europe's industrial cities whichblocked the UV radiation from the sun. Now-adays, however, many foods are fortifiedwith vitamin D additives, and a normal dietwill provide enough vitamin D without theneed for UV radiation.

Too much UV radiation, as well as too lit-tle, can be harmful. If the outer bulb of a mer-cury or metal halide lamp is broken, the lampmay not be extinguished: it may continue togive off both light and UV radiation from thearc tube. If the lamp, withsits broken bulb, ishoused in an open reflector particularly anopen aluminum reflector then that reflectormay concentrate the UV radiation. The con-

t., centration of that radiation may well harm theeyes of someone who looks directly at the lu-minaire.

Special types of mercury and metal I- ',t-ide lamps are available which extinguishthemselves automatically if the lamp bulb isbroken. It is recommended that these beused unless the lamp is protected frombreakage.

HID lamps in schoolsMany successful school lighting installationsuse HID lamps. However, from time to timethere has been adverse reaction to theselamps, particularly to high pressure sodiumlamps in general use at the time of writing.The lamps have been criticized on thegrounds of

1. the colour of the light from the lamps;Accurate colour discrimination is impossibleunder HID lamps. The colour of paint usedfor art-is-distortedrflesh_tones_are_cbanged_____and food can look unappetizing.

2. the glare they cause;Glare is not unique to HID lighting. It cantbecaused by any form of lighting.

3. the time the lamps take to btirn brightlythat is, tie lamp striking.time;HID lamps are not suitable where light isneeded instantly at the flick of a switch.

4. their possible health effects.

The phenomena of glare and colour ren-dering, and the operating characteristics ofHID lamps, are well known. That they in par-ticular give rise to complaints points to the in-

\ con'ettose-of-the_lamps andemphasizes theneed for good lighting design. it is also worthrememberingithat lamps are constantly im-proved in colour, efficacy, and performance.It may be found that an improved lamp canbe used in a location previously denied toearlier lamps of the same type.

As far as the health effects of HID lampsare concerned, the studies conducted havebeen inconclusive. Laboratory studies, spec-ifically with high pressure sodium lamps,have led to the conclusion that there may beunexpected effects on the user impression ofclarity with these lamps. But there is no clearunderstanding of this phenomenon or of theextent of the problem.

Lamp colour rendering and the choice ofilluminance'It is sometimes stated that when one usesgood colour rendering lamps, the illuminancein a working space can safely be less thanthe illuminance needed when lamps ofpoorer colour rendering are used. This state-ment is incorrect. It has resulted from lack ofunderstanding of a classic experiment on"visual clarity" by Aston and Bellchambers.There is no evidence to show that the illumi-nance for a visual task depends on the col-our rendering properties of the lamps used tolight that task.

Loss of muscle strength under variouslight sourcesAlthough there are suggestions to the contra-

_ ____rythere_is_no_good.scientific evideace_tbatmuscle strength is affected by the visible lightspectrum and the presence or absence of ul-traviolet radiation. In other words, a student'smuscle strength is not affected by the type oflamp used to illuminate the school.

Lamp flickerFluorescent and HID lamps tend to flicker asthey.get old. Children are much more sensi-tive to flicker than adults. The onset of flickermay well be the signal to change a lampeven though it has 'not reached the end of itsrated life.

Fluorescent lighting and hYpe'ractivechildrenPresent data are inconclusive on any con-nection betweenhyperactiyity in children andthe use of fluorescent lighting in classrooms.

Student tiredness t' IThere is some evidence to suggest that stu-dents working under fluorescent lighting withvery good colour rendering become less vis-ually fatigued than those working under'some other forms of fluorescent lighting.There is no evidence-on,how this reduction invisual fatigue affects the general fatigue andwork performance of students.

The windowless schoolThe windowless school is a radical departurefrom the conventionally designed school. Ac-cordingly, a number of researchers have in-vestigated the, reaction of staff and studentsto their windowless environments. The re-sults have been summarized as follows:

tif any single conclusion is to be reached fromthe studies of windowless schools it is thatthe absence of windows neither improves norimpairs [scholastic] performance. Althoughsome students like the [windowlessenvironment], others, possibly a majority,would prefer to have windows. The moststriking conclusion seems to be absence csignificant findings, either for or againstWhether this is the fault of the experimentaldesigns-used, or a reflection of the real ab-sence of any pronounced effect, is notknown 'Yet the absence of enthusiasm for

'the windowless school suggests that it shouldbe used with some caution, particularly sinceits long-term effects are unknown.*

Lighting for areas with visual displayunits (VDUs)The viewer screen of a VDU behaves some-thing like a migtr..Peflections of bright sur-faces (a lumina:re or a window, for example)on the screen will dilute, or even obscure, the

'Belinda L Collins. Review of the'PsychologicalReaction to Windows," Lighttng Research andTechnology, vol 8. no 2 (1976), p 81

images of the d a appearing on'it. Thismeans that the operator must make an extriseeing effort or adopt an awkward posture in ,

order to read the data presented on thescreen. The lighting must therefore be de-signed so that reflections of bright surfacesare not seen in the viewer screen.

It is also necessary to avoid too high abrightness contrast between the viewerscreen and other surfaces (such as a windowor desk top), otherwise the eye will tire as it'adapts first to one brightness and then to theother.

Consequently, the location of the VDUmust be chosen so that no reflections ofbright surfaces (luminaires, walls, windows)are seen in the viewer screen. Furthermore,the VDU operator must not.have a directview of a bright surface, particularly a win-dow.

In many situations the VDU operatorlookt first at the source documents and thenat the viewer screen. In order to keep downthe brightness contrast between the screenand the other surfaces at the work station, 4general illuminance of about 250 lux shouldbe provided at the work station, and an illu -'minance of 500 lux on the source documents(probably by a local light). The'top of thework station should have a reflectance ofabout 30 to 40 per cent.

Lighting for plant growthMany types of incandescent, fluorescent,.and HID lamps are used to grow plants in -'doors. The lamps used, either alone or incombination, must produce light with aproper colour balance if normal plant growthand shape is to result. Particular attentionmust be paid to ensuring that thKe is enoughenergy in the red and_blue parts of the spec-trum. A wide spectrum fluoreScent lamp ismade to meet the special requirements ofplant growing.

23

Sai

.4

/14,14.

:'''!!.11 it,

;,}11;t1,1w

;1,51.111'

6. SWitching and EnergyManagement in LightingSystems

It is quite common to find that luminaires areleft on in an unoccupied space or in placeswhere daylight provides adequate light. Infact, overuse of lighting systems mostfre-quently occurs because luminaires are notswitched off, and not because they areswitched on unnecessarily.

Switching is a vital element in lightingSystem design. If the provision of switches isskimped, all the lighting in a large space maybe controlled by one switch. This means thatall the lighting will be left on even though partof the space may receive adequate daylightor even though only part of the space may bein use. The consequences in'; rms of energycosts are obvious.

Manual Switching

Manual switching should at least permit sep-arate control of the rows of luminaires paral-lel to the windows, since these can beswitched off for much of the time during nor-mal school hours.

It is relatively expensive to provide addi-tional manual wall switches to an existinglighting system. It may be possible, however, Ito rearrange existing switching so as to pro- jvide more flexible control of the lighting.

Pull cord switches may be put to use in,say, a school office. They can be provided 1

less expensively than wall switches.

C ntrcis

If a school ceases to be occupied at a fixedtime each day. it may be worthwhile installinga time control to switch off most of the lightsshortly after that time. The time controlshould have an override device to'permit thelighting to be switched on should it be re-quired after the usual fixed time.

A low-cost electronic timer has been de-veloped for direct fitting into a standard wallbox to replace a conventional mane 'al wallswitch in individual classrooms. ThL timer isprogrammed for 15, 30 or 60 minutes' light.

Photoelectric Controls

Daylight entering through large windows can ,

provide enough light in part or all of a class-room for part of the day. Photoelectric controlcan ensure that the electtic lighting -can nei-ther be switched on nor remain on if daylightprovides the required illuminance. A photo-electric sensor measures the illuminance'provided by daylight at a chosen location orlocations in the classroom. When that illumi-nance reaches a predetermined level, thecontrol switches the electric lighting either onor off.

Photoelectric control-is not suitable forswitching high, intensity discharge (HID)lamps because of thei, starting characteris-tics. It may, however, be used to controlthese lamps on a continuously variable diM-ming circuit.

Where photoelectric control is used, anoverride switch should be provided to ensurethat no lights are left on when the classroomis unoccupied.

On-off photoelectric controlThe control switches off the electric lightingwhen the illuminance produced by daylight isabout equal to that produced by the electriclighting system. Similarly, when the illumi-nance from the daylight falls below a certainlevel, the control switches the electric,lightingback on. Consequently, when switching oc-curs, the illuminance in the space is suddenlyjust about halved ordoubled, depending onwhether the illuminance from the electric lightis subtracted from or added to the illumi-nance from the daylight.

There is considerable evidence that theoccupants of a space find this sudden

\change in illuminance of 50 or 100 per centunsatisfactory when large areas are con-trolled in this way. However, it appears to bemore acceptable if, for example, photoelec-tric control is applied to the row of luminairesnearest the window. This is mainly becauseswitching this row will be infrequent.

25

Top-up photoelectric controlEquipment is available which will automati-cally vary the light output from and thus theilluminance produced by incandescent,fluorescent, and HID lamps. This equipmentis controlled bya photoelectric sensor so thatthe illuminance from the electric lighting topsup that from daylight when thedaylight fails'to reach the desired illuminance level. As thedaylight fades,the electriclighting increases;as the daylight gets brighter, the elp6triClighting is reduced.

This method of control uses less energythan the on-off photoelectrib control.

Special ballasts are required for thefluorescent and HID lamps.

Occupancy Detector Coritrol

An occupancy detector senses the presenceof people in a space. If nobody is there, thelamps are either dimmed or switched off, de-pending on the circuit used. There is usuallya time delay between the last person leavingthe space and the lamp switching. However,vhen a person enters the space, the lampst,Jme on immediately.

26

Monitoring Lighting Use: ThqElapsed-time Meter

The elapsed-time meter records the numberof hours an electric lighting system has beenin use over an extended period of time. It canthus be used to calculate the energy con-sumed by the lighting syStem over that perkod.

The record of the lighting system usemust not of course include other uses ofelectric energy, such as office equipment,tools, and machinery. The usual approachhas been to wire a separate meter into each

. lighting circuit being studied. But this is ex-,pensive and inconvenient.

An alternative elapsed-time meter hasbeen developed by the National ResearchCouncil of Canada. This meter uses a pho-toelectric device to record lighting use. Itdoes not have to be wired into an electriccircuit: instead, it is simply mounted on aconvenient surface near the luminaires beingstudied. The meter will sum the-number ofhours the electric lighting has been provided.

7. Making the Most of YourLighting Systerrl: BasicConsiderations

Before attempting to make your present light-ing system more effective and energy effi-cient, ask yourself this question: Is the light-ing system a good one and appropriate tothe space? If the system is neither effectivenor in good shape, consider replacing it.

The first step in making an appraisal ofthe present system is to measure the illumi-nance,- the light level in the variousspaces. A light meter measuring to an accu-racy of about ± 10 per cent will be adequate.(This accuracy will not be achieved with thecheapest light meters.) The illuminance in aspace will vary from place to place: the de-gree of variation depends on the luminairelight distribution, the luminaire spacing, andthe mounting height. -

Compare the measured illuminanceswiththose recommended in Table 1. There isno need for the two sets of illuminances toline up exactly since the eyes perforni satis-factorily over a widerange of conditions.

Other aspects of the lighting system thatshould be investigated are as follows:

the amount of glare caused by the lighting;Does the lighting suddenly become morecomfortable if you shade your eyes withyour hand? If so, there is too much glare

the colour rendering characteristics of thelamps;Are some colours accentuated or made tolook dull or otherwise unacceptably distort-ed? Is there a need for lamps with particu-larly good colour rendering (e.g., in an artclassroom)?

the general condition of the luminaires:Have the luminaires been poorlymaintained, with cracked, discoloured ormissing lenses, and reflectors with ground-in dirt?

- the location of the working positions withrespect to the luminaires and windows.

If you have decided that the presentlighting system is worth retaining, numberof procedures can be adopted which will en-able you to save energy consumption andcosts as well as increase the effectiveness ofthe system. Some of these procedures oaf)be carried out without any capital investment:others will require some additional outlay, butthey can be expected to pay for themselveswithin a fairly short time.

Whatever you do, remember that themost efficient way of saving electrical energy- and the cost of that energy - is to switch off

the lamps when they are not required.Incandescent lamps should be switched

off as soon as they are not required: their lifeis not affected by the number of times theyare switched. Switching does, however, af-fect the life of flUorescent and-high intensitydischarge (HID) lamps. Accordingly, a com-promise must be reached between the en-ergy saved by switching and the reduction inlamp life - with the consequent earlier lampreplacement resulting from the switching.Rememberthat as a general rule, iffluorescent lamps are not to be used for fiveminutes or more they shOuld be switched off;if HID lamps are not to be used for an hour ormore they should be switched off.

Procedures requiring no capitalinvestment

1. Remove unnecessary luminaires.

It is recommended that the luminaires beremoved rather than just the lamps (as is fre-quently suggested) because a number of lu-minaires without lamps will look unsightlyand could spoil the appearance of the spaceconcerned. A lighting system that looksmakeshift today will go on looking makeshiftfor the next 'fifteen years unless something isdone about it.

However, it may not be possible to re-move luminaires, but only lamps. If these arefluorescent or WIC) lamps, the ballasts in theluminaires should be disconnected sincethey will continue to consume energy even ifno lamps are connected to them. For exam-ple, a commonly-used ballast for two 40 WRapid Start fluorescent lamps will continue todraw about 16 W.

If the luminaires are in rows and a uni-form illuminance is required, remove alter-nate luminaires in each row rather than allthe luminaires in one row. Check the illumi-nance once the luminaires have been re-moved.

2. Replace amps which are beyond theiruseful life.

A badly blackened fluorescent lamp; forexample, will produce juste fraction of itsrated lighfoutput.

3. Keep the luminaires clean and replacediscoloured lenses.

4. Locate the work positions to take advan-tage of daylight.

27

28

Procedures with short pay-back period

1. Convert to more efficient lamps.

In general, it care be said that with theever increasing cost of energy, the most effi-cient lamp will be the least expensive in thelong run. However, more than efficiencyalone must be considered before a conver-sion program is started. Other factors in-clude:clude:

the cost of the lamp sand its ballast if one isrequired)

the colour appearance and colour render-ing properties of the lamp

the ability of the luminaire to accept thenew lamp, or the cost of modifying the lumi-naire to permit it to do so, or the cost of anew luminaire

the light output of the lamp in relation to theilluminance required

2. Use more efficient ballasts.

New ballast designs have resulted inmore efficient ballasts. For example, a lumi-naire for two 40 W Rapid Start fluorescentlamps that uses a standard ballast will drawa total of 96 W or more; using the new bal-last, it will draw only 89 W.

Solid state ballasts for fluorescent lampswill become available in the foreseeable fu-ture. They will consume less energy and op-erate the lamps more efficiently than anypresent-day ballasts.

Carefully review the pay-back periodwhen converting to the more efficient bal-lasts.

3. Use dummy fluorescent lamps.

A dummy ("phantom") lamp may beused to replace one ref the two 40 W RapidStart fluorescent lamps operated by a singleballast. The dummy lamp gives off no lightand consumes no energy. Its use reducesthe power drawn by the luminaire from about96 W to about 36 W.

4. Replace fluorescent lamps with lamps oflower wattage.

The 4-foot 40 W Rapid Start fluorescentlamp may be replaced by the 4-foot 35 WRapid Start fluorescent lamp in most (but notall) luminaires without change of ballast. Thiswill result in a reduction in light output (andthus of illuminance) of about 10 per cent. The

)

k-power drawn by the luminaire, using a stan-.dard ballast, will be reduced from about 96 Wto about 86 W.

5. Improve switching.

The simpler forms of switching systemsdiscussed in section 6 could be considered.

6. Repaint school spaces.

Paint finishes should have thereflectances specified in Table 2. This willensure that much of the light will be reflectedby the principal surfaces (thus contributing tothe illuminance), and will help create'a spacethat is visually comfortable. The finishesshould be chosen under the light from thelamps which will normally be used. If this isnot done, uneYpected (and unpleasant)changes in the appearance of the finishesmay result.

7. Provide light coloured blinds to reflect thesun's heat back through the windows.

The blinds must be light in colour andmust be kept clean, otherwise the sun's heatwill be absorbed and then re-radiated into thespace and warm it.

Procedures with long pay-back periodProviding a sophisticated lighting control sys-tem would entail a longer pay-back periodthan the procedures discussed so far. Thissystem would be integrated with an auto-matic energy control system. The lightingcontrols should probably top up the daylightin spaces with windows rather than provideonly on-off control. (These controls are dis-cussed in section 6.)

Finally, a new lighting system could beinstalled. The new lighting system should betightly designed so that it conforms closely tothe design criteria. This will require more de-sign skill than much of that displayed up tonow: it has not been unusual, for example, tofind illuminances 25 per cent higher thantheir design value, which means that 25 percent more energy than necessary is beingused.

Skilful design will require accurate infor-mation on the lamp light output, luminaire effi-ciency, the finish of surfaces, and space lay-out. It may also require ceiling systems thatpermit more precise location of luminairesthan many existing ceiling systems. The de-signer should pay as much attention to thequality of the lighting and the ambience itproduces as to the technical aspects of thedesign.

Fanshawe College,Main Campus,London

Craig Zeidler Strong,Architects

Exter16'

Exterior lighting is primarily Intended fornight-time access: pathways to the school(and people on and near the-pathways) areclearly distinguished; car perking isfacilitat-ed. The lighting can be.provkied by_lumkmires mounted high upon the schOalbuild-ing and directed butwards. These.may besupplemented (or, sometimes, supplanted)by pole-mounted luminaires illuminating thepaths and their immediate surroundings andthe parking area. The illuminances andthelamps recommended for good contemporaryresidential street lighting are appropriate forschool exterior lighting. The lighting mustclearly identify steps and changes of level.

Lighting fora landsuaped campus mustdo more than meet the requirements of ac-cess. It must also bein harmony With thelandscape. A landscaped campus presentsmany interesting problems to the lighting de-signer. The *pus sets out-to achieveanoutdoor environment that lends itself natu-rally to student life. Kis designed to be at har-mony with itself and with the surroundingneighbourhood. That harmony must not bedisturbed by the lighting, which must not betoo loud and which willbe unwelcome if it in-trudes oh a neighbouring space.

_ -

The1iht1ngmuMIent purposes_ Ofthe_Vcarnpus;$0m0-

,othertlargti=pulighti(10 fii-..riffir$70*,from small lumi rEw4Oetilitirkbillitt.cloeingwallt-4poltSi

hotot sourcolnot:care ft 110E4 give)4sklight whIchIsconpatllwith liiuà1natedTprobable:that phosphor toa _ometal halide-tempt provide the best commise between therieedfordering and thtneed for

7

Luminaires -used outdoorsbléth

dirtladert areattthe- light: athlcklayerofdeadiflsectslscomMonly bend inside a luMinaire:- Consequent-ly, the either be tightly

a breather.PlasticOncluding polycarbonate, have

replaced glass bowls and lenses In some =outdoor luminaires. Polycarbonate has ahigh impact strength and, although it discol-ours somewhat,is frequently selected for lu-minaires where vandalism might occur.

-

=

Znr

e'cr,

32 29

_

$

.r.r

re't

Use of ReflectingBlind to Project Sun-light to Interior ofSpace

9:Lighting in The Year 2000

It is anticipated that over the next twentyyears lighting research will be aimed at:

- achieving a solid scientific base forillumi-nance recommendations;.

, .- reducing the electrical energy consumed,.

by lighting systems;This will include work on the more effectiveuse of daylight and the development of .more efficient lamps of good colour render-ing.

studying the effectof light on health andwell-being.

Listed below are a number of specific ap-plications of the above research, togetherwith some other design criteria whiCh are ex-pected to be-used in the year 2000.

All lighting design will be carried out by alighting specialist who will be skilled incombining daylight and electric light to pro-duce the most effective lighting system. Itwill have been found that the commonpresent practice of making one specialistresponsible for daylighting and another re-sponsible for electric lighting does not re-sult in the best lighting design.

- Sunlight will be collected from outdoorsand then distributed throughout a school by"light pipes".

Beamed sunlight will be used to provide u §-able illumination in a school. A device suchas reflecting venetian blinds will be used toproject the sunlight onto the ceiling where itwill be diffusely reflected onto the work.

- Guidelines will be developed to optimizethe design of windows. These guidelineswill take into account psichological consid-erations such as the response of people towindows of different shapes and sizes, aswell as parameters such as latitude, cli-mate, and orientation.

Glazing will be developed which will re-spond to changing daylight conditions. Forexample, the glazing may cloud over andblock the hot summer sun, but let the sun'srays pass throUgh during the winter towarm and illuminate the interior space.

Greenhouses will be built along the walls ofa school. They will capture heat from thesun, and this will be used to warm theschool. Shutters over the glass will retainthe heat at night. The greenhouses willgrow food for consumption in the cafeteria.

-The present IES illuminance recommenda-tions, which are based on consensus, willbe replaced by recommendations based onscientific study.

.

31

m

32

Guidelines will be developed on the effectof light and lighting systems on health andwell-being.

.Lamps of good colour rendering and ofgreater efficacy will be developed. The the-oretical maximum efficacy of a white lightsource js about 300 ImNV. The maximumefficacy of a preserit-day white light sourceis about one-quarter of this value: there is alot of room for improvement.

HID lamps will be instant starting and ofbetter colour rendering.

Low watt .ige HID lamps that do not requirea separate ballast will t .s a direct, high-effi-ciency replacement for incandescentlamps.

Lighting systems will be specially designedfor use in spaces containing visual teach-ing aids (such as visual display units).

Life-cycle costing will be used instead,of in-itiafcost in drawing Op the budget forschool building and operating-costs, includ-ing the budget for the lighting system.

Life-cycle costing assesses the eco-nomic consequences of various design al-ternatives as they relate not only to thecapital cost but also to the-future costs of alighting system, such as the cost of energy,replacements, and maintenance. Thismethod of costing minimizesThe impact ofvariables which are beyond the control ofthe lighting designer. One particular varia-ble is the future cost of energy.

Power load (in kW) will no longer be used, as a figure of merit with which to compare

the energy-conserving properties of lightingsystems..Instead, the comparison will beOn the basis of actual energy (in kWh),which will take into account the switchingand dimming characteristics of the sys-tems.

. The electrical energy used for lightingis the product of the lighting power load (inkW) and the time (in hours)'that load is inuse. Thus

Energy = Power x Time(kWh) (kW) (h)

At present only the power load is gen-erally considered in designing alightingsystem: the smaller that load for a givenarea and use, the more energy- efficient thesystem is considered to be. This statement,however; is not accurate, since it does nottake into account the number of hours thesystem will be switched on. Moreover, itdoes not take into accountreductions in thepower load resulting from dimming systemscoupled to daylight: as daylight increases,the dimmiog systems will reduce the elec-tricdighting and thus the power drawn bythat lighting.

Panel 'arm will be developed which willnot have to be directly connected to asource of electrical power.

The lighting and other requirements forelectricity will be met by on-site power gen-eration. When needed, this will be supple-mented by power from the local utility. Amicro- processor will monitor the system.

Bibliography(Principal references only)

Aston, S.M., and Bel !chambers, H.E. "Illumi-nation, Colour Rendering and Visual Clarity."Lighting Research and Technology, vol. 1,no. 4 (1969), pp. 259-61.

Collins, Belinda L. "Review of the Psycholog-ical Reaction to Windows." Lighting Re-search and Technology, vol. 8, no. 2 (1976),pp. 80-88.

Illuminating Engineering Society of NorthAmerica. IES Lighting Handbook. 5th-and 6theditions. New York: IES, 1972 and 1980.

Mayron, Lewis W., et al. "Light, Radiationand Academic Behavior."- AcademicTherapy, vol. X, no. 1 (Fall 1974), pp. 33-47.

O'Leary, K. Daniel, et al. "Children's Hyper-activity Behavior Under Two Lighting Condi-tions." In Proceedings: 5th Annual Meeting, .

American Society for Photobiology. NewYork: ASP, 1977.

Wotton, Ernest. "Some Considerations Af-fecting the Inclusion of Windows in Office Fa-cades." Lighting Design and Application, vol.6, no. 2 (1976), pp. 32-40.

33

Glossary

Brightness (also called subjective bright-ness): A sensation resulting from viewingsurfaces or spaces from which light is emit-ted. Thus, when we say, "The wall is brighterthan the ceiling" or "That luminaire is toobright", we are describing our subjective re-action as opposed to a measured quantity.

(Electric) Energy: The unit is the watt-hour-or, more often, the kilowatt-hour (kWh).Thus, the daily 24-hour energy consumptionby two 500 W lamps is 24 kWh.

(Electric) Power: The unit of power is thewatt (W). Thus, when we say that a lampcircuit draws 97 watts, we are referring to thepower drawn and not the energy consumed.

Equivalent Sphere Illumination (ESI): Amethod of specifying illuminance which takescontrast (and hence visibility) into account.

Illuminance (sometimes also referred to asillumination; level of illumination; lighting lev-el): The concentration of luminous flux on asurface.

The unit used in Canada has been thefootcandle (fc) which is equivalent to the illu-minance of 1 lumen uniformly distributed-over a surface area of one square foot.

The SI (or metric) unit is the lux (SIsymbol: lx). One lux is the illuminance of 1lumen uniformly distributed over a surfacearea of one square metre (1 m2).

Thus 1 fc = 10.76 lx or 1.076 dalx

Lumen (Im): The unit of luminous flux thestuff we call light. For example, a new 100 Wlamp may emit 1700 Im.

34

Luminaire: The complete lighting unit con-sisting of the lamp or lamps, together withthe parts designed to distribute the light, toposition and protectthe lamps, and to con-nect the lamps to the power supply. Formerlyknown as "lighting fixture" and in Great Brit-ain "light fitting".

Luminance: The concentration of luminousflux emitted from a luminous or reflecting sur-face.

The unit used in Canada has been thefootlambert (fL) which is equivalent to1 lumen transmitted through or reflected byone square foot of surface.

The SI unit is the nit (nt) which is equal to1 candela per square metre (cd/m2).(Thecandela [cd] is the SI unit of luminous intensi-ty.)

Thus 1 nt = 3.43 fL

In everyday speech the term "bright-ness" is used loosely instead of "luminance".Thus, one might say that "the brightness of -that wall is 10 cd/m2" when, to be correct, theterm "luminance" should have been used.Luminance is a measured quantity; bright-ness is not

Luminous efficacy (or, in short, efficacy):The quotient of the total luminous flux (lu-mens) emitted by the total lamp power'(watts) input. It is expressed in lumens perwatt (Im/W).

Thus, if a 100 W lamp emits 1700 Im, itsefficacy is 17 Im/W.

The term "luminous efficiency" has, inthe past, been used extensively for this con-cept.

Reflectance: The ratio of the reflected flux tothe incident flux that is, the percentage oflight falling on a surface that is reflected.

) 2,4

Appendix 1: Principal Lamp,Types Used

Incandescent lamps

Lower wattage lamps (uSually 150 W or less)most commonly used for general lightinghave Type A bulb and, usually, MediumScrew base. (Lamps are also available with

- Candelabra Screw, Intermediate Screw, and.D.C. Bayonet bases.)

..

Spherical globe -shape lamp with Type Gbulb and Medium Screw'base.

Reflector lamp with Type R bulb and MediuniScrew base. An aluminum coating Is vapor-ized on the insidee of the bulb to act as areflector. Control of the beam spread is de-termined by the bulb shape and the degreeof hosting of the inner bulb surface. Spotlamps are-lightly frosted; flood lamps areheavily frosted. These lamps are suitable forindoor use only since the hot bulb will break ifrain or snow falls on it. ,7

Reflector lamp with Type PAR bulb and Me-dium Skirted Screw base. The bulb shape issimilar to the R bulb except that it is made oftwo pieces of pressed glass instead of a sin-gle piece of blown glass. A lens system ismoulded on the lamp fce. The beam controlis more accurate than that of an R bulb. Thehot bulb will not break if rain or snow falls onit, so the'lamp can be used outdoors as wellas indoors.

The reflecting surface may be either astandard aluminum coating or a dichroicfilter. The latter is a many-layer coating ofmetallic salts. It reflects light into the beambut allows heat from the lamp to escapethrough the reflector sides. Thus the radiantheat in the lamp beam is reduced.

A luminaire approved by the CanadianStandards Association for use with a PARlamp with a standard aluminum reflectorinaynot be approved for use with a PAR lamp ofthe same voltage with a dichroic filterreflector. A luminaire must be specifically'ap-

' proved for use with the latter lamp. .,

35

.c

.

36

Reflector lamp with Type ER bulb and Me7dium Screw base. The bulb is similar to theR bulb but designed to focus the light beamat a spot a few centimetres ahead of thelamp. The lamp works more efficiently than .the R lamp in many deep-baffled daylights.

41

Tubular lamp with Type T bulb. The lampscome with a variety of types of bases includ-ing Mediurri Screw, Intermediate Screw,Candela4raScrew,D ,.C..Bayonet and Disc. .

,..,

.1.

41,

Higher wattage lamps (usually of more than150 W) most commonly used for generallighting have Type PS bulb. Up to and includ-ing the 300 W size, most lamps have the Me-dium Screw base; the Mogul Screw base isused on most higher wattage lamps.

, .

Hign Intensity Discharge (HID)tamps

Most mercury and metal halide lamps of400 W or less and high pressure sodiumlamps of 1000 W or less have the Type E.bulb. In the smaller diameter lamps (E18, forexample) the bulge in the bulb is much lesspronounced than that illustrated here. Lowwattage lamps have the Medium Screwbase; medium and high wattage lamps havethe Mogul Screw base.

2

ff

p

Most mercury and metal halide lamp ofmore than 400 W have the Type3T bulb lidthe Mogul Screw base.

4

Low pressure sodium lamps have the tubularType T bulb. Usually the lump has either asingle Bayonet base or else a Bi-pin base ateach end.

FiCiores6ent lamps

Rapid Start lamps (such as the common 4-foot 40 W lamp) and some Ihstant Startlatnps have the Medium Bi-pin base.

Most Slimline (Instant Start) lamps haye theSingle Pin base.

800mA., 1000mA., and 1500mA lamps havethe Recesseclouble Contact base.

Notes:1. The maximum bulb diameter in eighths of aninch is given in the lamp description appearing inthe manufacturers' catalogues. For example, thestandard 4-foot 40 W fludrescent lamp has a 112bulb a tubular bulb twelve-eighths of an inch, or11/2 inches, in diameter.

2. Most lamp bulbs are of common lime glass.However, low expansion "hard" glass, similar:tothat used to make oven-ware, is used where thelamp may be subjected to thermal shock. A typicallamp with a "hard" glass bulb is the 150 W PAR38reflector lamp. Ittuan be used out of doors where itis liable to thermal shock from rain and snow.

37

Typical lamp bases forincandescent and HID tamps(excluding low pressure sodiumlamps)

Principal uses

Low wattage HID lamps and most generalpurpose incandescent lamps up to and including the 300 W size.Medium Screw 11/16 in. diameter (d)*

Medium aruthigh wattage HID lamps andmost general purpose incandescent lamps ofmore than 300 W.

'Mogul Screw 1 19/32 in. d

PAR reflector lamps.Medium Skirted Screw 11/16 in. d

Low wattage incandescent lamps of typesnot used for general lighting.purposes (suchas some tubular lamps).Intermediate Screw 9/8 in. d

Candelabra Screw Y2 in. d

Bayonet 9/8 in. d

Disc 7/8 in. d

Note:For many years brass was almost the only mate-rial used for screw lamp bases. Now, however,most incandegcent lamps up to and including the300 W size, and some of lamps, have aluminum bases. Aluminum cos:, less than brass, it iseasier to work, and it is a better conductor of elec-tricity. Screw bases of brass are still recom-mended for damp locations: in these locations thealuminum base may bind with the lampholdermaking it diffic .1 to remove the lamp

The production orlamp bases in metric units is not yetscheduled by the manufacturers.

ti

Appendbq 2: The Effects ofUltraviolet Radiation

The following is a statement by the (British)Chartered Institution of Building Services(CIBS) Lighting Division:

. . considerable publicity has been-given to-the claim that exposure to light from conven-tional fluorescent lighting can have deleteri-ous effects on people's health, strength andwell-being. Further, it has been asserted thatone particular light source, called a "full-spectrum" fluorescent !amp can alleviatethese problems. The purpose of this state-ment is to set down the opinion of the CIBSLighting Division Technical Committee onthese claims.

The first point td note about these claims isthat they do not refer to the ability to see.What is being claimed is that in some un-specified way radiation from conventionalfluorescent lamps carthave general,. non-vis-uafeffects. This is riot an unreasonable prop-osition given that some non-visual effects oflight are well established. For example, hu-.man circadian rhythms are known to belinked to the naturally occurring day-night cy-cle. The second point to be noted aboutthese claims- is that they identify the factor re-sponsible for the effect, namely the relativelack'of near ultra-violet radiation in the spec-

, tral emission of conventional fluorescent lightsources. The full-spectrum fluorescent lamp'emits much more near ultra-violet radiationthan most conventional fluorescent lampsand in this respect is much more like daylightoutdoors.

The validity of these claim- can be consid-ered at two levels: first the level of practicalexperience, second the level of scientific ex-periment. At the level of practical experienceit is important to realise that it is not justfluorescent lamps which have relatively littlenear ultra-violet emission. All light sourcesused for conventional interior lighting, eg in-candescent lamps, fluorescent lamps, highpressure discharge lamas, and daylight ad-mitted through windows, have a reducednear ultra-violet component compared withthat available outdoors. Yet all these lightsources have been used for many years forinterior lighting. In this situation it does notseem unreasonable tq expect that if the,claims for near ultra-violet were true late dif-ferences in the health, strength and well-be-ing of those who work indoors and outdoorswould be apparent. No such differences areapparent. This may be because, for mostpeople, exposure to intent, lighting occursfor only part of the day. At other times peopleare exposed to unfiltered dayii;ht and henceto high levels of near ultra-violet radiation. Itmay be that when people have no access to 'unfiltered daylight for very long periods, asfor example when working for long hours in-

., ,)

doors during winter in northern latitudes,then exposure to near ultra-violet radiationfrom light sources might be beneficial. How-'ever, such restricted exposure is rare. Forthe more usual situation many years of expe-,-hence with wideTange of different lightsources, including daylight, in interiors doesnot suggest that an absence of near ultra-violet radiation is detrimental to people'shealth, strength and well-being.

As for the level of scientific experiment,here the picture is more confused. Numerousexperiments have been reported dealing withthe effect of near ultra-violet radiation onpeople, animals and plants. Mar)), of theseexperiments have been poorlyVesigned andinadequately reported. Where clear effectshave been found the direction of the effect-has sometimes been different for differentcreatures. Frequently the assumption hasbeen made tbat the results from animals andplants are applicable to people but onlyrarely has the influence of intervening varia-bles been considered. Overall, there is noconsistent pattern of results which indicatesa clear beneficial effect of near ultra-violet ra-diatiun cn people's health, strength and well-being.

Nonetheless, there are enough data tosuggest that the spectrum of radiation towhich people are exposed might be impor-tant in more ways than simply allowing for vi-sion tdoccur, but at the moment the case isnot proven. What is required now is carefullyconducted research which seeks to identifythe effects of.different forms of radiation onpeople under realistic conditions. Resultsfrom such research would be more convinc-ing than sensational assertions based onanything from the behaviour of hyperactivechildren through tooth decay in hamsters tothe feeding habits of snakes. Until more reli-able evidence is available, the scientific sta-tus of the claimstabout the benefit of near .

ultra-violet radiation from light sources mustremain in doubt.

Building Services, vol. 3. no. 3 (1981), pp. 36-37.

The following statement has been made bythe Ontario Ministry of Health:

The quantities of ultraviolet radiation avail-able from artificial sources are found to bevery low. Eight hours exposure to 500 foot-candles is the approximate equivalent offifteen minutes spent outdoors on a sunnyday. Consequently, the use of artificial illumi-nation sources which simulate the solarspectrum will bd of very limited use until suchtime sufficiently documented scientific evi-dence proves otherwise.

Illumination Systems in Hospitals. Ministry ofHealth, Ontario, 1979, p. 5.1.

39