Design and performance of pedestrian subway lighting systems

10
Design and performance of pedestrian subway lighting systems John Burnett * , Alex Yik-him Pang Department of Building Services Engineering, Hong Kong Polytechnic University, Kowloon, Hong Kong SAR Received 2 December 2003; accepted 2 March 2004 Available online 10 April 2004 Abstract Pedestrian subways can provide for safe and convenient movement of pedestrians across busy roads and similar obstacles. However, fear for personal safety may deter use, particularly at night. Good lighting is an important part of subway design, yet design guides and standards specify only limited criteria for performance, e.g. maintained horizontal illuminance. Various types and arrangements of luminaires are in common use, but as-installed lighting performance is variable. Lighting performance also de- teriorates over time, not least due to prevailing environmental conditions. The Hong Kong Government is responsible for over 330 pedestrian subways distributed throughout the territory, making maintenance a significant ongoing commitment. A survey of subway lighting systems has been undertaken to evaluate the performance of the systems against design standards, and to ascertain the opinions of subway users on perceived lighting quality, with a view to identifying the most successful design solutions. Factors affecting the life cycle costs of the most common lighting systems installed in Government maintained subways are also discussed. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Pedestrian subways; Lighting systems; Performance evaluation 1. Introduction Pedestrian subways, unless part of a labyrinth within a major transportation complex, tend to be rather modest facilities in terms of the provisions for engi- neering services. However, they form an important part of the urban infrastructure, especially in dense built up areas where preference is given to the movement of road vehicles. Subways and footbridges provide for safe crossing and access to parks, gardens, buildings, and other facilities that are adjacent to busy roads, highways and major transport interchanges. A disadvantage of footbridges, which in Hong Kong are often very con- servatively designed in heavy steel or reinforced con- crete, is that they blight the landscape and may reduce privacy and obstruct daylight and views of adjacent buildings. The relatively shallow depth of a subway or underpass can allow for the inclusion of compact ramps that facilitate the movement of the personal possessions and other goods transported by a variety of conveyances. For most pedestrians using a subway would seem preferable to the alternative of scaling the steps of an overhead walkway. However, fear for personal safety may deter subway use, particularly at night, and even in daytime if the subway is of such dimensions that the exit is not clearly visible from the entrance. Whilst it cannot be expected that improvements to pedestrian lighting will address crime and fear problems in all circum- stances, evidence suggests it is an important factor (Painter, 1996). Certainly, poor lighting does tend to increase the level of anxiety felt by pedestrians, and poorly designed, uninviting pedestrian subways or un- derpasses tend to deter use, impacting on accessibility and deterring the free movement of pedestrians (Kosk- ela and Pain, 2000). Improving the lighting in a subway seems to be re- garded as a sound strategy when confronted by com- plaints from users. As far back as 1980 the Hong Kong Government’s response to a question in the Legislative Council as to the measures that would be taken to im- prove security in pedestrian subways, especially those in * Corresponding author. Tel.: +852-2766-5111; fax: +852-2764-3374. E-mail address: [email protected] (J. Burnett). 0886-7798/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.tust.2004.03.001 Tunnelling and Underground Space Technology 19 (2004) 619–628 Tunnelling and Underground Space Technology incorporating Trenchless Technology Research www.elsevier.com/locate/tust

Transcript of Design and performance of pedestrian subway lighting systems

Page 1: Design and performance of pedestrian subway lighting systems

Tunnelling andUnderground SpaceTechnology

Tunnelling and Underground Space Technology 19 (2004) 619–628

incorporating Trenchless

Technology Research

www.elsevier.com/locate/tust

Design and performance of pedestrian subway lighting systems

John Burnett *, Alex Yik-him Pang

Department of Building Services Engineering, Hong Kong Polytechnic University, Kowloon, Hong Kong SAR

Received 2 December 2003; accepted 2 March 2004

Available online 10 April 2004

Abstract

Pedestrian subways can provide for safe and convenient movement of pedestrians across busy roads and similar obstacles.

However, fear for personal safety may deter use, particularly at night. Good lighting is an important part of subway design, yet

design guides and standards specify only limited criteria for performance, e.g. maintained horizontal illuminance. Various types and

arrangements of luminaires are in common use, but as-installed lighting performance is variable. Lighting performance also de-

teriorates over time, not least due to prevailing environmental conditions. The Hong Kong Government is responsible for over 330

pedestrian subways distributed throughout the territory, making maintenance a significant ongoing commitment. A survey of

subway lighting systems has been undertaken to evaluate the performance of the systems against design standards, and to ascertain

the opinions of subway users on perceived lighting quality, with a view to identifying the most successful design solutions. Factors

affecting the life cycle costs of the most common lighting systems installed in Government maintained subways are also discussed.

� 2004 Elsevier Ltd. All rights reserved.

Keywords: Pedestrian subways; Lighting systems; Performance evaluation

1. Introduction

Pedestrian subways, unless part of a labyrinth within

a major transportation complex, tend to be rathermodest facilities in terms of the provisions for engi-

neering services. However, they form an important part

of the urban infrastructure, especially in dense built up

areas where preference is given to the movement of road

vehicles. Subways and footbridges provide for safe

crossing and access to parks, gardens, buildings, and

other facilities that are adjacent to busy roads, highways

and major transport interchanges. A disadvantage offootbridges, which in Hong Kong are often very con-

servatively designed in heavy steel or reinforced con-

crete, is that they blight the landscape and may reduce

privacy and obstruct daylight and views of adjacent

buildings. The relatively shallow depth of a subway or

underpass can allow for the inclusion of compact

ramps that facilitate the movement of the personal

* Corresponding author. Tel.: +852-2766-5111; fax: +852-2764-3374.

E-mail address: [email protected] (J. Burnett).

0886-7798/$ - see front matter � 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.tust.2004.03.001

possessions and other goods transported by a variety of

conveyances.

For most pedestrians using a subway would seem

preferable to the alternative of scaling the steps of anoverhead walkway. However, fear for personal safety

may deter subway use, particularly at night, and even in

daytime if the subway is of such dimensions that the exit

is not clearly visible from the entrance. Whilst it cannot

be expected that improvements to pedestrian lighting

will address crime and fear problems in all circum-

stances, evidence suggests it is an important factor

(Painter, 1996). Certainly, poor lighting does tend toincrease the level of anxiety felt by pedestrians, and

poorly designed, uninviting pedestrian subways or un-

derpasses tend to deter use, impacting on accessibility

and deterring the free movement of pedestrians (Kosk-

ela and Pain, 2000).

Improving the lighting in a subway seems to be re-

garded as a sound strategy when confronted by com-

plaints from users. As far back as 1980 the Hong KongGovernment’s response to a question in the Legislative

Council as to the measures that would be taken to im-

prove security in pedestrian subways, especially those in

Page 2: Design and performance of pedestrian subway lighting systems

620 J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628

areas more susceptible to crime, was that frequent visits

by police patrols and adequate lighting would help to

keep subways safe for lawful users (Legislative Council,

1980). In a study undertaken for the UK’s Department

for Transport (2000) it was reported pedestrians, andmore so female pedestrians, identified subways or un-

derpasses as unsafe places to walk, citing a sense of

isolation and vulnerability to crime. It was concluded

that improving subways through improved lighting and

interior decoration can be effective in reducing fear,

provided the subways are maintained, cleaned and kept

free from crime. In a report by Bracknell Forest Bor-

ough Council (2002) it was observed that ‘‘for manyelderly, young and disabled people, the existing under-

passes can be a mental and physical barrier to accessing

the town centre despite lighting and decorative im-

provements. The result is that many examples have been

reported of elderly and young people crossing very busy

dual carriageways to get from the estates to the town

centre. More controlled crossings at-grade, along with

lighting and visibility improvements to existing under-passes should help those who feel excluded from the

town centre currently’’. From these and other similar

reports it may be concluded that a well designed subway

particular in terms of lighting performance, will en-

courage subway use, whereas poor lighting discourages

use, tending to reduce the freedom of movement of pe-

destrians, and may result in preference to use inappro-

priate alternatives.In Hong Kong besides those managed by the Mass

transit Railway, the Kowloon Canton Railway and a

number of private sector companies, over 330 subways

fall under the responsibility of the Highways Depart-

ment (HyD), with maintenance carried out by the

Electrical & Mechanical Services Department (EMSD).

The number of luminaires installed in the Govern-

ment operated subways is in excess of fifteen thou-sand, with over eleven thousand in subways spread

throughout various towns in the New Territories.

Lighting performance in these subways depends on the

choice, layout and installation arrangements of lumi-

naires and the reflectance of wall and floor surfaces.

Given the capital outlay and cost of maintenance,

environmental factors that affect the performance of

subway lighting systems are important design consider-ations.

This paper reports on a survey of Government

maintained subway lighting systems in Hong Kong. This

included an assessment of actual lighting performance

against prevailing design standards, but taking into ac-

count depreciation of light output due to pollution. A

survey of subway users has also been undertaken to

determine whether local design standards are adequateand which lighting arrangements are the most preferred

by users. Issues impacting on the life cycle cost of sub-

way lighting systems are also discussed.

2. Subway lighting systems

2.1. Design specifications

As stated in a number of authoritative standards andguides on the matter, the main objective of pedestrian

subway lighting is to provide for the safety and security

of users, and the lighting quality should be such as to

reduce fear of use. In Hong Kong the Highways De-

partment’s Public Lighting Design Manual (1996) gives

specific requirements for the design of lighting for cov-

ered pedestrian routes, as well as public transport in-

terchanges, tunnels and high mast lighting forgovernment projects. Apart from stating the required

lighting levels emphasis is also placed on other aspects of

visual comfort, e.g. uniformity and avoidance of glare.

The design recommendations for subway lighting in-

clude maintained illuminance and uniformity within

subways, and maintained illuminance for subway ramps

and stairs (Table 1). A maintenance factor of 0.65 based

on the initial lumen (100 h) of lamp output is used indesign calculations.

The Hyd manual specifies that energy-saving tubular

fluorescent lamps (18 W–58 W MCF) shall be used

generally and mounted on the ceiling wherever possible.

To maximise available light luminaires should be surface

mounted or partially recessed, except wall-mounted lu-

minaires installed less than 1.8 m above floor level shall

preferably be fully recessed. Subway luminaires arelongitudinal mounted in order to reduce glare. It also

specifies that spacing between adjacent luminaires shall

not exceed 5 m to avoid dark spots in the event of lamp

failure. Provisions for daylight savings at ends of sub-

ways and emergency lighting in long subways are also

covered in the Manual. Given the levels of humidity and

air pollution found in Hong Kong the index of protec-

tion specified is at least IP65, that is, dust tight, andwater projected in jets from any direction shall have no

harmful effects (BSI, 1992). Externally smooth, vandal-

resistant polycarbonate diffusers are specified for fluo-

rescent luminaires.

Additional guidance on the design of subway lighting

systems is available from various authoritative sources,

which historically for Hong Kong has usually meant

UK sources, although some designers may use guidesfrom international organisations (listed in the HyD

Manual). The Institution of Lighting Engineers’(ILE)

technical report, (1997) makes recommendations for

levels of illumination and lighting objectives for sub-

ways, ramps, stairs and approaches for the safety and

security of people, and draws attention to desirable

features of the complete lighting system for effective

operation and economic maintenance. Performancerecommendations are given in terms maintained average

horizontal illuminance and minimum horizontal illumi-

nance lighting during day and night (Table 1). The (CIE,

Page 3: Design and performance of pedestrian subway lighting systems

Table 1

Lighting performance specifications for ‘enclosed’ subways and staircase/ramps

Authority> HyD (1996) ILE (1997) CIE (2000) CIBSE (1992) BSI (1996)

Subway

EHðAVEÞ (lx) 100* 350 d 100 d – 350 d

150 n 30 n 100 n

EHðMINÞ (lx) 150 d 50 d 150 s 100 d

100 n 15 n 300 l 50 n

Minimum uniformity 0.5 – 0.6 – –

Reflectance

Wall – – 0.5 – Light coloured

surfaces

Ceiling 0.5

Floor –

Stair ramp

EHðAVEÞ (lx) 50* r 350 d – 300 350 d

100* st 150 n 150 n

EHðMINÞ (lx) 150 d 100 d

100 n 50 n

Notes: Enclosed subways are those where the exit cannot be seen from the entrance, or where daylight penetration does not make significant

contribution to the minimum lighting (ILE, 1997). EHðAVEÞ and EHðMINÞ refer to maintained average and minimum horizontal illuminance, re-

spectively (the illuminance on the reference surface ensured by appropriate lamp renewals and maintenance). d¼ day, n¼ night, s¼ short, l¼ long,

r¼ ramp, st¼ stairs, * indicates the document is not specific (illuminance) uniformity¼ ratio of minimum illuminance to average illuminance on a

surface.

J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628 621

2000) also limits its recommendations to illuminance

and uniformity, but includes recommendations in re-

spect of reflectance’s of interior surfaces, as does the

CIBSE (1992) and a British Standard (BSI, 1996), thekey performance criteria for which are also summarised

in Table 1.

It is clear that the local design standard is very similar

to UK standards as far as illumination of the interior

zones of subways is concerned, but there is no provision

for adaptation at exits/entrances during daytime, and no

specific criteria for reflectance of subway surfaces.

Clearly, UK recommendations for illumination of stairsand ramps during daytime are much higher than the

values specified in Hong Kong.

2.2. Lighting quality

It is important that the lighting is of sufficient quan-

tity and quality to permit recognition of approaching

pedestrians. However, it needs to be borne in mind thatin reality subways, especially those with ramps, will of-

ten be used by cyclists, skateboarders, and even the

occasional motorcyclist. Subjective responses to an il-

luminated enclosed space depends on more than simply

the amount of light (illuminance) and peoples reactions

to entering such spaces may vary from ‘bright’ or ‘well-

lit’ to ‘gloomy’ or ‘poorly-lit’. It is the spatial distribu-

tion of light, particularly on vertical surfaces thatinfluences adaptation and visual performance and con-

sequently the degree of user satisfaction.

Other than illuminance levels (quantity of light) the

guides and standards referenced are not precise in de-

fining the quality of light. According to CIBSE (1992) it

is important that lighting of subways and stairways re-veal the presence of other people moving in the distance

and be of sufficient quality and quantity to permit some

facial recognition as pedestrian approach one another. It

is also recommended that vertical surfaces be well illu-

minated, and be as light coloured as practicable. It is

also recognised that the speed of adaptation and visual

orientation to enable people to recognise objects and

others depends, amongst other factors, on the conditionof the eyes, which can vary significantly with age and

physical condition. The so-called Black Hole Effect, a

major issue in the design of road tunnel lighting systems,

can also apply in the case of pedestrian subways, sug-

gesting that higher values of illumination be provided at

the ends of subway tunnels (Table 1). Doubling the il-

lumination over a distance of 6m during daytime is

suggested by ILE (1997), who also consider it essentialto use a light source which emits radiation over the

whole colour spectrum (full spectrum lighting) to im-

prove colour rendering and thereby aid vision.

2.3. Lighting equipment

An important issue for the selection of lighting

equipment for subways is vandalism and accidentaldamage. CIBSE (1992) advises that luminaires should

if possible be recessed into the wall or ceiling, with

Page 4: Design and performance of pedestrian subway lighting systems

622 J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628

diffusers of impact resistant material, such as toughened

glass or polycarbonate, secured by vandal resistant fas-

tenings. However, recessed ceiling-mounted luminaires

that fail to provide sufficient illuminance on ceiling

surfaces may make a subway appear gloomy. The CI-BSE guide states that Cornice recessed or wall-mounted

luminaires will solve this problem, and this view is en-

dorsed by both BSI (1996) and ILE (1997).

Recessed ceiling and wall mounted fittings are the

most common installations in the newer subways de-

signed for the Highways Department, but refurbished

subways are likely to utilize surface mounted fittings

(Fig. 2). The majority of luminaires are fitted with T8fluorescent tubes controlled by electromagnetic ballasts

or electronic ballasts. As the HyD Manual does not

specify the reflectance of surfaces in subways the sur-

faces of existing subways were often finished in darker

colours. Whilst this means that surfaces do not discolour

so easily the installed lighting level needs to be increased

to compensate. However, in recent years HYD has also

designed and renovated subways with light colouredsurfaces. Furthermore, it is understood that the

Department is revising the maintained illuminance to

150 lx.

2.4. Maintained light output

The main performance specification in all guides and

standards is the maintained illuminance. Maintainedhorizontal illuminance is the illuminance on the refer-

ence surface (e.g., subway floor) at the time maintenance

has to be carried out by cleaning equipment and subway

surfaces and replacing lamps. Light output falls due to

accumulation of dirt on light sources and optics, aging

of components, premature failure on one or more lamps,

temperature and voltage variations, etc. (CIE, 1977).

The maintained illuminance is usually taken as the ini-tial light output multiplied by all the depreciation fac-

tors (maintenance factor – MF), that is:

Maintenance factor ðMFÞ¼ lamp lumen ðoutputÞMF

� lamp failure factor � luminaire MF

� room surface MF: ð1Þ

On this basis, according to the HyD specification theinitial illuminance in a subway should be no less than

155 lx (100/0.65). For sealed units (IP65 or better) op-

erating continuously depreciation due to dirt on lamps

and temperature variations are likely to be much less

than that through aging or dirt accumulation on lumi-

naire surfaces, which depends on the extent to which dirt

is present in the atmosphere. The rate of fall-off of lu-

men output and lamp failure rates varies with lamp type,and should be predictable from manufacturer’s data.

Both factors are likely to be not less than 0.95 for the

fluorescent lamps most commonly used in government

subways. Therefore, based on a MF of 0.65, the allow-

ance for depreciation of light output due to dirt on

diffusers and subway surfaces over the recommendedmaintenance (cleaning) period of one year is expected to

be around 25%. In practice, the depreciation due to

soiling and dirt will vary with the local conditions, in

particular the extent of air pollution.

3. Evaluation of subway lighting performance

3.1. Maintained illuminance

Site measurements were undertaken to evaluate the

performance of the subway lighting systems maintained

by EMSD against the design criteria given in Table 1.

The selection of the 20 subways surveyed was based

firstly on a representative sample of lighting systems that

had been installed or refurbished within the previous fiveyears, i.e., after the issue of the revised HyD Manual in

1996.

Without the resources to monitor air pollution or

dust levels at individual subways so to take pollution

into account the second criteria for selecting the sub-

ways was their location in relation to the air quality

monitoring stations operated by the Environmental

Protection Department (EPD, 2003). These stations aresituated in a number of urban areas in the territory to

continuously monitor five pollutants; Total Suspended

Particulate (TSP), nitrogen dioxide (NO2), ozone (O3),

sulphur dioxide (SO2) and carbon monoxide (CO). Most

are positioned on buildings at heights of 13–21 m, with

several located near street level.

The subways surveyed are located close to one of the

nine stations to see if there was a correlation betweendepreciation in light output and the level of air pollution

in the vicinity. It was of interest to see if the depreciation

correlated with readings of TSP. Levels of SO2 and NO2

might also correlate with depreciation as a consequence

of corrosion and discolouration. The main features of

the lighting systems and nature of their location are gi-

ven in Table 2.

For each subway surveyed the change in illuminancewas measured over an interval of two months during the

period February to April 2003. Horizontal illuminance

at floor level was measured in accordance with CIBSE

recommendations (CIBSE, 1992), using a Testo model

545 luxmeter. Air pollution data recorded at the nearby

monitoring station was obtained from EPD’s website.

Table 2 summarises the results of the measurements.

From Table 2 it can be seen that based on the ‘spot’measurements, i.e., not taking into account when light-

ing systems had been subject to annual maintenance or

cleaning, a number failed to meet the criteria for

Page 5: Design and performance of pedestrian subway lighting systems

y = 0.1147x–6.0471

0.00

1.00

2.00

3.00

4.00

5.00

6.00

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

(µg/m3)

(%)

Fig. 1. Depreciation of light output (%) against SO2 (left), NO2 (centre) and TSP (right).

Table 2

Maintained horizontal illuminance, levels of pollutants and depreciation of light output

Subway District Area Lighting

system

Average

maintained

illuminance (lx)

TSP SO2 NO2 Depreciation

(two months) (%)

1 Central b R, E 85 78.4 18 54.6 3.3

2 b R, M 109 3.2

3 c R, E 130 2.8

4 Kwai Chun d R, E 83 75.6 21 61 3.4

5 a S, E 123 1.9

6 Kwun Tong e S, E 89 83.6 15.8 72.8 3.6

7 f R, E 110 3.3

8 Sham Shui Po e S, M 132 84.6 18.8 66.8 4.3

9 Tsuen Wan a R, E 106 79.8 17.8 64.2 3.6

10 f R, E 114 3.5

11 e S, M 125 3.5

12 Sha Tin a R, M 88 66.8 13 48.2 1.6

13 e R, E 140 1.5

14 e S, E 126 1.3

15 Tai Po e S, E 109 70.4 14 50.4 2.1

16 e S, E 130 1.7

17 e S, M 145 1.5

18 Yuen Long d S, E 126 98.2 19.2 56.6 4.8

19 d S, E 110 4.4

20 d S, E 78 5.6

R¼ recessed c/w prismatic diffuser, S¼ surface mounted luminaire, E¼ electronic ballast, M¼magnetic ballast, a¼ residential, b¼ commercial,

c¼mixed residential/commercial, d¼highway, e¼ densely populated residential and f¼ industrial .

J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628 623

maintained average illuminance of 100 lx. Fig. 1 shows

that depreciation in light output is closely related to the

prevailing levels of TSP, with differences ranging be-

tween 5.6% and 1.3% depreciation measured over the

two months. There is less correlation between depreci-

ation of light output with NO2, although there is likelyto be some impact on system failures given that NO2

reacts with humid air to form nitrous acid that can ox-

idise metallic components of the lighting system. There

is no correlation with SO2 as it varies only slightly be-

tween locations. Whilst the polycarbonate diffusers

of the luminaires are resistant to corrosion, ‘yellowing’

of diffusers with age does contribute to depreciation oflight output.

Page 6: Design and performance of pedestrian subway lighting systems

Fig. 2. Five types of lighting arrangements (A–E).

Table 3

Lighting measurements in five subways with lighting systems A to E

Type A Type

Subway section

Horizontal illuminance at floor (lx)

Maximum 132 160

Minimum 86 80

Average 109 112

Uniformity 0.79 0.7

Vertical illuminance at eye level (lx)

Maximum 136 293

Minimum 88 75

Average 108 168

Uniformity 0.82 0.45

Ramp/staircase

Horizontal illuminance at floor (lx)

Maximum 210 150

Minimum 18 15

Average 55.4 48.5

Uniformity 0.32 0.3

Vertical illuminance at eye level (lx)

Maximum 150 135

Minimum 10 8

Average 75 48.7

Uniformity 0.13 0.16

624 J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628

3.2. Vertical illuminance

The installed lighting in a subway is designed to

achieve a particular level of horizontal illuminance at

floor level. However, the vertical illuminance measurednear eye level is an important factor in the ability to

recognize the faces of other users, and is likely to in-

fluence the subjective opinions of users as to the lighting

quality. In this connection, five subways of similar size;

about 2.5 m height, 3.5 m width and 45 m in length, each

using a different lighting system (A–E) as shown in

Fig. 2 were selected for further investigation. Both the

horizontal illuminance at floor level and the vertical il-luminance at eye level were measured in accordance with

CIBSE recommendations (CIBSE, 1994) using the Testo

luxmeter. The vertical illuminance was measured at 1.5

m above the floor, with the average of the readings taken

in each direction recorded. Both subway core and ramp/

staircase sections were also measured for maximum and

minimum illuminance and uniformity (Table 3).

The average horizontal illuminance in all five sub-ways (109–126.8 lx) was slightly above the design stan-

dard for maintained illuminance (100 lx), with

uniformity close to or exceeding the design standard.

The vertical illuminance varied significantly between

maximum and minimum. The low uniformity for types

B and D is attributable to the recessed mounting

method, which impairs the intensity distribution of

luminaries.

B Type C Type D Type E

150 198 168

79 79 78

119 117.4 126.8

0.66 0.67 0.61

330 385 266

102 48 83

189.2 162.6 130.3

0.54 0.3 0.64

310 180 250

19 17 21

70.8 50.7 80.1

0.27 0.33 0.26

610 121 482

27 10 29

158.2 62.5 112

0.17 0.21 0.26

Page 7: Design and performance of pedestrian subway lighting systems

J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628 625

4. User opinions on subway lighting

Lighting level, types of luminaire, glare, colour tem-

perature, etc., can influence user’s perceptions of subway

lighting quality. To evaluate user’s opinions on lightingperformance users were interviewed (in Cantonese) at

the five selected subways using a short questionnaire

containing eight questions. The person-to-person ap-

proach ensured that the respondents could understand

the questions and thereby achieve more accurate re-

sponses with a high rate of return. The vocabulary used

was felt to be that best understood by lay persons,

avoiding as far as possible technical terms. The Englishversions of the questions are given in Table 4.

The questions sought to obtain psychological re-

sponse to lighting quality, overall satisfaction with

subway lighting design, and feelings associated with

safety and security. In order to obtain more or less

spontaneous responses the order of questions was mixed

to obtain response about the subway, its exit/entrance

and more general feelings about safety and security. Thequestionnaire used a five-point scale, scoring from )2 to

Table 4

Survey questions and aggregates of scores for responses

1. What is your opinion on the brightness of the subway lighting? 2.

lig

� Too dark )2 �

� Dark )1 �

� Fair (satisfactory) 0 �

� Bright +1 �

� Very bright +2 �

3. Does the different light level between subway and its exit or

entrance cause you to feel uncomfortable?

4.

co

� Very Serious )2 �

� Serious )1 �

� Fair 0 �

� Acceptable +1 �

� Very acceptable +2 �

5. Can you distinguish people’s faces in the subway? 6.

� Very difficult )2 �

� Difficult )1 �

� Fair 0 �

� Easy +1 �

� Very easy +2 �

7. Do you feel that good lighting of subways can deter loiterers and

suspicious users from theft?

8.

� Strongly disagree )2 �

� Disagree )1 �

� Fair 0 �

� Agree +1 �

� Strongly agree +2 �

Type Q1 Q2 Q3 Q4 Q5 Q8

A 11 29 7 8 28 25

B 12 32 )10 )16 )12 0

C 18 22 )16 )13 )2 1

D 3 22 )2 4 13 5

E 20 22 9 28 35 28

+2 to indicate the extent of either negative or positive

responses to the questions. At each of the five selected

subways the questionnaire was administered to 50 users

chosen at random, with their ages recorded in five cat-

egories – elderly, middle age, young, teenage and child.Given the simplicity of the questionnaire and limited

sample size the data can only be taken as indicative of

user opinions on each of type of lighting system sur-

veyed. The maximum ‘score’ for any question is 100 (50

users scoring þ2 to �2). As Table 3 and Fig. 3 shows

none of the scores for Q1 to Q5 and Q8 which deal with

performance are particularly high, indicating a fairly

neutral response lighting quality as judged by the users.The responses to question 1 (Q1 – brightness) seem to

indicate that the majority of users were not too dissat-

isfied with the amount of light provided in the body of

subways, suggesting the local requirement of 100 lx is

not unreasonable. Colour temperature (appearance) of

the fluorescent lighting seems to be regarded as satis-

factory for most users (Q2). For the ramp/staircase the

average lighting level of all types was just above 50 lx,marginally fulfilling the CIE standard, with great vari-

Do you like the colour temperature (appearance) of the subway

hting?

Very unsatisfactory )2Unsatisfactory )1Fair 0

Satisfactory +1

Very satisfactory +2

Does the direct glare of the subway lighting influence your feeling of

mfort?

Very serious )2Serious )1Fair 0

Acceptable +1

Very acceptable +2

Do you feel safer using subways if the subway lighting is brighter?

Strongly disagree )2Disagree )1Fair 0

Agree +1

Strongly agree +2

Are you satisfied with the existing subway lighting system overall?

Very unsatisfactory )2Unsatisfactory )1Fair 0

Satisfactory +1

Very satisfactory +2

Score

Q1–Q5+Q8

Q6 Q7 Score

Q6+Q7

108 44 44 88

6 36 37 73

10 32 36 68

45 40 39 79

142 51 45 96

Page 8: Design and performance of pedestrian subway lighting systems

Fig. 3. Comparison of scores for five types of lighting system (A–E).

626 J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628

ations in horizontal and vertical illuminance. This is

reflected in the responses to question 3 regarding thelight level difference between subway section and exit/

entrance.

It is clear from question 4 that types B (transverse

recessed) and C (longitudinal surface) fared poorly when

compared to type E (sideways cornice) as far as direct

glare is concerned, with types A (sideway surface) and D

(longitudinal recess) barely satisfactory. Analysis based

on Unified Glare Rating (UGR) (CIE, 1995) illustrateswhy type E is the most satisfactory as far as glare is

concerned. UGR is scaled so that a value of 10 repre-

sents imperceptible glare and a value of 30 represents

intolerable glare. On the assumption that the back-

ground luminance and luminance of luminaires was

similar due to similar physical dimensions of the sub-

ways and luminaires used the UGR of type E was esti-

mated to be 17.8, indicating that discomfort glare isbetween perceptible and just acceptable, while the UGR

for type C is 19.7, with discomfort glare between just

acceptable and unacceptable.

Questions about the ability to distinguish peoples

faces (Q5) showed type E as being preferred. Despite the

lower vertical illuminance Type A also scored better in

this aspect. The aggregate of all the scores (Table 3) and

the comparisons given in Fig. 3 indicates that Type E(Cornice mounted) is the most favoured arrangement,

with Type A (sideways surface mounted) also

favoured. Also, given that question 5 is about the en-

trance/exit lighting, removing these scores only mar-

ginally improves the ranking for the other three

arrangements.Questions 6 and 7 are more to do with the general

feelings and perceptions of users. The aggregate scores

for each group of 50 users ranged from 36 to 45 re-

garding deterrence of crime, and 32–51 regarding safety.

Given that the highest scores were recorded at Type E,

marginally greater than those at Type A it would seem

that responses to these questions were influenced by the

conditions in the subway at which the users were inter-viewed. The responses suggest that further improve-

ments to subway lighting performance would be

appreciated.

5. Consideration of life cycle costs

The life of a subway lighting system often depends onnon-operational factors such as inability to meet im-

proved performance standards leading to obsoles-

cence. The introduction of new standards in 1996 has led

to the renovation or upgrading of existing subway

lighting systems older than five years, before the end

of their useful life (generally taken as 8–10 years).

The external factors that affect life cycle cost (Eq. (2))

include the system design and installation, environ-mental impacts, vandalism, and frequency of cleaning

and maintenance. Clearly, vandalism is dependent on

the inclinations of subway users but the design of

luminaires and method of installation helps to reduce

failures due to this cause:

Page 9: Design and performance of pedestrian subway lighting systems

J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628 627

Life cycle cost ðper luminaireÞ¼ initial costþ ðenergy costþ cleaning costs

þ repair costsþ lamp replacement costÞ� number of years: ð2Þ

The subway lighting systems studied consist of two

types of luminaire; corrosive resistance surface moun-

ted, and recessed type with prismatic diffuser, having

one or two 36 W fluorescent tubes controlled by either

conventional or electronic ballasts. The initial cost de-

pends on luminaire type, number of lamps (tubes), andtypes of ballast (controls), with electronic ballasts about

20% more energy efficient. Operation is around the clock

(i.e., 8760 h over a year) and electricity cost is 1 HK$per

kW h (1 US$¼ 7.8 HK$). As fluorescent tubes generally

have an 8000–10,000 h rated life, re-lamping every year

is regarded as being cost effective. Annual cleaning is

also regarded as sufficient. Repair costs depend on the

number of failures (faults), which are a function of en-vironmental conditions and other factors. Failures can

be grouped under five main headings; luminaires faults;

electricity supply failure; protective device tripping;

wiring short circuits; and vandalism.

Luminaire faults dominate, accounting for about

two-thirds of the total faults. It is thought that the high

levels of humidity and ambient temperature, presence of

corrosive or polluting substances speed up the deterio-ration of internal components such as ballasts, capaci-

tors, and wiring. Around one tenth of faults were caused

by wiring short circuits. The various electrical faults may

be a result of poor workmanship, both of concealed and

surface conduits at the expansion joints of subway

structures leading to the ingress of water. The cleaning

of subways using the high-pressure jet cleaners may also

be a factor as the joints of conduits and adaptable boxcovers in surface or concealed conduit systems are sus-

ceptible. Heavy rain and typhoon conditions can cause

flooding especially in low-lying land such as the north–

west of the New Territories, also leading to wiring short

circuits or tripping of protective devices.

Almost one tenth of failures are ascribed to damage

due to vandalism. Most of the subways have a height of

2.5 m so people can easily reach the luminaires fixed atceiling height. Anecdotal evidence suggests that subways

located in the vicinity of newly constructed public es-

tates record few cases of vandalism in the first few years,

but this dramatically increases over a number of years,

before decreasing again. This seems to correlate with the

age and behaviour of children who are initially too

young to damage the lighting equipment when first

moving in to the vicinity, but are capable to do so aftergrowing older. Furthermore, higher rates of vandalism

occur during long holidays, and were noticeable worst

during the period of the 2002 World Cup. In contrast,

few cases were reported in the older residential areas

where the population is more aged.

Lack of published data precludes a detailed assess-

ment of life cycle costs for each of the subway lighting

systems surveyed, although it seems clear that unless thesubway suffers a very large number of failures, annual

electricity cost dominates the ongoing cost of operating

a subway lighting system. The use of twin tube lumi-

naires may not be a good choice, given the higher energy

cost, and increasing the number of single tube lumi-

naires can be cost effective, and also provide improved

uniformity of illuminance. Cleaning and re-lamping

costs are not too significant, even if cleaning frequency isincreased to twice a year where high levels of pollution

occur, therefore increasing initial lighting levels to

compensate for dirty environments may not be cost

effective either.

6. Summary and conclusions

This study has examined the lighting conditions in 20

selected subways and found that some did not comply

with the local design recommendations in respect of

maintained illuminance. A key factor in light output

deprecation is the prevailing level of TSP in the air,

and a correlation exists between the rate of depreca-

tion and recorded TSP levels in the vicinity of the sub-

way. Designers can consider use of EPD air pollutiondata as a reference to estimate depreciation of light

output due to soiling and to make allowances in design

and/or maintenance. Nonetheless, the design MF of 0.65

used in design seems reasonable for all but the worst

cases.

Responses to the survey questionnaire suggest that

the design average maintained illuminance of 100 lx is

barely acceptable for the average subway user, buthigher levels would be welcomed, particularly by the

elderly. The raising of illuminance to 150 lx will have an

energy and cost penalty and may not by itself have

significant impact on the perceptions of users, unless

overall lighting quality is also enhanced. That said, im-

provements to stair/ramp lighting would be beneficial.

The questionnaire survey does demonstrate that users

feel safer and more comfortable using subways withgood quality lighting and pleasant interior finishes.

Glare seems to be a problem that affects visual comfort

of users so that the cornice mounted arrangement is the

preferred choice. This arrangement also improves the

ability to distinguish the appearance of oncoming

pedestrians.

The majority of damaged polycarbonate diffusers are

the consequence of vandalism. In newly developed res-idential areas, rigid protection guards are installed to

reduce damage by children, but further accumulation

of dust is inevitable. Thermally toughened glass has

Page 10: Design and performance of pedestrian subway lighting systems

628 J. Burnett, A.Y.-h. Pang / Tunnelling and Underground Space Technology 19 (2004) 619–628

outstanding resistance to adverse environments and

tolerance to damage and may be a suitable alternative.

Acknowledgements

This paper is based on work undertaken by the sec-

ond author in satisfying the academic requirements of

the BEng(Hons) degree in Building Services Engineer-

ing.

References

Bracknell Forest Borough Council, 2002. Local Transport Plan

Annual Progress Report. Available from <http://www.brac-

knell.gov.uk/Council/EnvironmentNew/pdf/LTPAnnualreport01-

02.pdf>.

British Standards Institution, 1992. BSI. Degree of protection

provided by enclosures (IP code). BS EN 60592: 1992.

British Standards Institution, 1996. BSI. Road lighting – Part 9: code

of practice for lighting for urban centres and public amenity areas.

BS 5489-9.

Chartered Institution of Building Services Engineers, 1992. CIBSE.

Lighting guide LG06: the outdoor environment.

Chartered Institution of Building Services Engineers, 1994. CIBSE.

Code for interior lighting.

Commission Internationale De L’eclairage, 1995. Discomfort Glare in

Interior Lighting. CIE 117-1995.

Commission Internationale De L’eclairage, 1977. Depreciation of

Installations and Their Maintenance. CIE 33-1977.

Commission Internationale De L’eclairage, 2000. CIE. Guide to The

Lighting of Urban Areas. CIE 136-2000.

Department for Transport, 2000. Personal security issues in pedestrian

journeys. Available from <http://www.dft.gov.uk/stellent/groups/

dftmobility/documents/page/dftmobility503822.hcsp>.

Environmental Protection Department, Hong Kong SAR Govern-

ment, 2003. EPD’s Air Quality Monitoring Network. Available

from <http://www.epd.gov.hk/epd/english/environmentinhk/air/

airquality/>.

Highways Department, Hong Kong SAR Government, 1996. Public

Lighting Design Manual. Available from <http://www.hyd.gov.hk/

publications/index.htm>.

Institution of Lighting Engineers, 1997. Technical Report No. 13.

Lighting of Pedestrian Subways, second ed.

Koskela, H., Pain, R., 2000. Revisiting fear and place: women’s fear of

attack and the built environment. Geoforum 31, 269–280.

Legislative Council, Hong Kong, 12 March 1980. Oral answers to

questions. Security in pedestrian subways, p. 591.

Painter, K., 1996. The influence of street lighting improvements on

crime, fear and pedestrian street use, after dark. Landscape and

Urban Planning 35, 193–201.