A CRITICAL STUDY ON THE FIRE SAFETY

FOR BIG HOTELS IN HONG KONG

WU HOI HUNG

Ph. D.

THE HONG KONG POLYTECHNIC UNIVERSITY

2007

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 T h e H on g K on g P olytech n ic U n iv ersity

Department o f Building Services Engineering

A C ritical Study on the F ire S afety for

B ig H otels in H ong K on g

W U HOI HUNG

A thesis submitted in partial fulfilment o f

the requirements for the degree o f Doctor o f Philosophy

June 2006

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 UMI Number: 3265613

INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignm ent can adversely affect reproduction.
In the unlikely event that the author did not send a com plete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.

UMI
UMI Microform 3265613

Copyright 2007 by ProQuest Information and Learning Company.

All rights reserved. This microform edition is protected against

unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company

300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CERTIFICATE OF ORIGINALITY

I hereby declare that this thesis is my own work and that, to the best o f my

knowledge and belief, it reproduces no material previously published or written nor

material which has been accepted for the award o f any other degree or diploma,

except where due acknowledgement has been made in the text.

(Signed)

WU Hoi Hung______ (Name of student)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 ACKNOW LEDGEMENTS

I wish to thank Professor W.K. Chow for his guidance, encouragement and support

over the past years.

Thanks are also due to Mr. S.S. Han and Mr. Angus Cheng for supporting the

experimental work; and to all those who have helped me during my study.

Special thanks to Ms. Linda C.W. Yeung and Ms. Brenda Chan for helping in the

editorial work; and Ms. Nancy Tang for her generous help in typing the manuscript.

ii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 ABSTRACT

The abstract for the thesis titled “A Critical Study on the Fire Safety for Big Hotels

in Hong Kong” submitted by WU HOI HUNG in 2006 for the degree o f Doctor of

Philosophy at The Hong Kong Polytechnic University.

Fire safety is always a major concern in hotel operations and guest services. A

critical study on the fire safety in big hotels in Hong Kong was carried out in this

thesis. A detailed review on all the fire safety codes and requirements relating to

hotel accommodation was carried out. Local codes on the fire services installations

(FSI) required issued by the Fire Services Department were reviewed and the codes

from the Buildings Department were studied. Hotel licensing requirements by the

local Home Affairs Department were introduced. Fire safety requirements in the

Mainland China code on high-rise buildings, the National Fire Protection

Association safety code and other fire safety guidelines in USA were also discussed.

The total fire safety concept was found to be very important to hotel staff and guests.

On the hardware side, the active fire protection systems commonly adopted by

firemen in fire fighting against fires in hotel were examined and analyzed critically.

These systems are also essential to the fire fighting team o f the hotel itself.

Another hardware component, the passive fire protection systems including fire

resisting co n stru ctio n s, m e a n s of e sc a p e and areas o f a c c e s s w e re stu d ied . To most

o f the hotel operations, the software component, namely a proper fire safety

management scheme, is often regarded as more important in international chain

iii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 hotels. Housekeeping and maintenance o f the passive protection installations play

a major role in the overall fire fighting strategy.

A critical hazard area in the hotel, i.e. the kitchen, was then focused on. Safety

concerns and problems usually encountered in kitchens were surveyed and analyzed.

Full-scale burning tests in part o f a sample kitchen were performed. Fire scenarios

and the action o f active fire systems such as sprinkler and water mist systems were

assessed. In addition to the heat release rate, maximum temperature and fire

extinction time were examined and analyzed. It was found that the temperature

could reach the flashover temperature o f 600 °C in some cases, and the heat release

rate could be as high as 1.4 MW. No splashing or spilling o f fuel was found for the

test arrangements in the model kitchen. The sprinkler and water mist system were

shown to be effective in extinguishing kitchen fires but less extinguishing time was

required in using the sprinkler system. These experimental results provide valuable

guidelines for the design o f fire safety provisions in hotel kitchens. This refers not

only to hardware fire safety systems, but also software fire safety management to

achieve total fire safety.

Experimental studies were carried out for studying thermal sensitivity o f fusible

links at fire dampers normally installed in hotel guestrooms. Two groups of

experiments were carried out to examine the thermal response o f fusible links with a

heated wind tunnel. The first group o f experiments involved testing 50 fusible

links following common standards used in the USA, i.e., Underwriters’ Laboratories

33 on heat responsive links at constant temperature and air speed. All the fusible

links fulfilled the operating time as recommended. The second group of

iv

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 experiments involved the testing o f fusible links at various temperatures and air

speeds. Recording the operating time o f the fusible links would give the response

time index by plotting a suitable graph.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 TABLE OF CONTENTS

Abstract iii

List o f Tables x

List o f Figures xi

Nomenclature xii

Chapter 1 Introduction 1

1.1 Objectives o f the Research Work 2

1.2 Methodology 3

1.3 Outline o f the Thesis 4

Chapter 2 Hotel Fires 8

2.1 Review on Parameters Related to Fire Safety and Problems in Hotels 9

2.2 International Chain Hotels —Fire Safety Considerations 20

2.3 Critical Fire Hazard Area in a Hotel —Kitchen 30

Chapter 3 Total Fire Safety Concept Applied to Hotel 34

3.1 Introduction 34

3.2 Total Fire Safety Concept in Hotels 35

3.3 Development o f Total Fire Safety Concept 37

3.4 Application o f Total Fire Safety Concept in Fire Safety Design 38

3.5 Review on Passive Fire Protection Systems in Hotels 41

vi

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Chapter 4 Active Fire Protection Systems 57

4.1 Introduction 57

4.2 Local and International Fire Codes for Active Fire Fighting Systems 59

4.3 Automatic Fire Detection System 62

4.4 Automatic Sprinkler System 68

4.5 Fire Hydrant and Hosereel System 71

4.6 Water Mist System 73

4.7 Non-water Fire Suppression Systems 77

Chapter 5 Fire Safety Management in a Hotel —Software Component 79

5.1 Introduction 79

5.2 Main Objectives o f Fire Safety Management in a Hotel 83

5.3 Main Objectives o f a Fire Plan 83

5.4 Integration with Fire Safety Engineering in New and Existing Hotel
Building 86

5.5 Development o f an Effective Fire Safety Management Program 89

Chapter 6 Methods o f Study 90

6.1 Fire Safety Codes Review 91

6.2 Hazard Assessment 94

6.3 Necessity o f Experiments in the Study 97

6.4 Heat Release Rate 100

6.5 Fusible Links at Fire Dampers 103

vii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Chapter 7 Full-scale Burning Tests for Kitchens and Calorimetry 108

7.1 Calorimetry 108

7.2 Oxygen Consumption Calorimetry 111

7.3 Cone Calorimeter 116

7.4 Necessity o f Full-scale Burning Tests 119

7.5 An Assembly Calorimeter 123

7.6 Kitchen Fire 124

7.7 Full-Scale Burning Tests on Kitchen Fire 126

7.8 Results on Group A Tests 128

7.9 Results on Group B Tests 129

7.10 Discussion 130

7.11 Conclusions 131

Chapter 8 Performance of Heat Responsive Link in Active and Passive Fire

Systems 132

8.1 Water Sprinklers - Heat Transfer and Empirical Equations 132

8.2 Sprinkler Test and Response Time Index 133

8.3 Fire Dampers in Fire Protection Systems 135

8.4 Heated Wind Tunnel - Experimental Setup 138

8.5 Heat Responsive Link for Fire-Protection Service —UL33 142

8.6 Local Fusible Links - Simulation Tests with Heated Wind Tunnel 148

Chapter 9 Conclusion 150

Tables T-l

viii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Figures F-l

Appendix A Major Tasks of the Fire Safety Manager AA-1

Appendix B Fire Safety Management Program in a Hotel AB-1

Appendix C Equations for Heat Transfer at Sprinkler Heads AC-1

Appendix D Calculation for Minimum Heat Release Rate for Flashover AD-1

References R-l

Publications Arising from this Thesis P-l

ix

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 LIST OF TABLES

Table 6.1 Requirements on escape routes in China codes T-l

Table 6.2 Comparison on the egress capacity T-2

Table 6.3 Travel distance limits in NFPA life safety code T-3

Table 8.1 Difference in heated wind tunnel setup T-4

Table 8.2 Heat responsive link - rating T-5

Table 8.3 Operating time for links in sensitivity - oven heat test T-6

Table 8.4 Sensitivity oven temperature in sensitivity — oven heat test T-7

Table 8.5 Fusible link test result (activation time) at air speed o f 2.54 m s'1 o f
135 °C T-8

Table 8.6 Fusible link test result (activation time) at various air speeds and
temperatures T-9

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 LIST OF FIGURES

Figure 7.1 Burning tests o f kitchen fires F-l

Figure 7.2 Hotel Chinese restaurant kitchens tested F-4

Figure 7.3 Model kitchen F-5

Figure 7.4 Heat release rates o f testing group A F-6

Figure 7.5 Heat release rates o f testing group B F-7

Figure 8.1 Heated wind tunnel test for fusible links F-8

Figure 8.2 Activation times o f fusible links F-9

Figure 8.3 Activation times o f fusible links at various air speeds and
temperatures F-10

xi

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 NOMENCLATURE

A, Surface area o f the sensing element (m2)

Aef Surface area o f the sprinkler frame (m2)

Af Surface area o f the frame (m2)

A fJ p
Surface area o f the pipework (m2)

Av Hv Wv (m2)

A t 2 (L + W ) H + 2 L W - A v (m2)

B Constant (-)
i i i_
C Orifice plate coefficient (k g 2 m 2 °C 2 )

c Conductive heat transfer between the sensing element and the framework

Ce Specific heat capacity o f the sensing element (Jkg'1°C'1)

cf Specific heat capacity o f the frame (J kg'1“C '1)

H Height o f the compartment (m)

Hv Height o f the ventilation opening (m)

he Convective heat transfer coefficient of the sensing element (W m'2 °C'1)

hef Convective heat transfer coefficient o f the sprinkler frame (W m '2 °C'1)

h f Convective heat transfer coefficient o f the frame (W m '2 °C'1)

hfj p
Convective heat transfer coefficient o f the pipework (W m'2 °C _1)

k Thermal conductivity of air (W m '1 °C"1)

kc Factor related to the conductive heat loss c (-)

L Length o f the compartment (m)

I Characteristic dimension o f the sensing element (m)

m Mass flow rate in the duct (kg s'1)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 me Mass o f the sensing element (kg)

nif Mass o f the flame (kg)

Nu Nusselt number (-)

n Exponent (-)

AP Pressure drop across the orifice plate (Pa)

Qmf Minimum heat release rate for flashover (kW)

Re Reynolds number (-)

RTI Response time index (m 2 s 2)

RTIV Virtual response time index (m 2 s 2 )

Te Gas temperature at the orifice plate (°C)

Tg Nominal gas temperature in test section (°C)

Tm Marked temperature rating of the link (°C)

Tu Nominal ambient air temperature (24 °C)

t Time (s)

t0 Operating time o f the link (s)

top Operation time o f the sprinkler (s)

u Speed o f air (m s '1)

2-1
Vef Ventilation factor (m 2)

Vg Air speed in wind tunnel section (m s '1)

W Width o f the compartment (m)

Wv Width o f the ventilation opening (m)

P constant (-)

v Kinematic viscosity o f air (m2 s'1)

xiii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Qe Effective operating temperature (°C)

9e Temperature o f the sensing element (°C)

0.(0 ) Initial temperature o f the sensing element (°C)

9non, Nominal release temperature (°C)

A0e Increase in temperature o f the sensing element (°C)

AOf Increased temperature o f the frame (°C)

A es Increase in gas temperature (°C)

re ' Time constant o f the sensing element (s)

V Isolated time constants (s)

x’ Long-term time constant (s)

x" Short-term time constants (s)

xiv

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 1 INTRODUCTION

Modem hotels nowadays have many new construction features such as double skin

facade and use a lot o f glass features and new materials [Chow 2001a]. It is also

equipped in its internal areas with lots o f state-of-the-art technology and equipment.

The fire safety concern that may be brought forward by all these installations will be

studied very carefully. In particular, all the electrical and mechanical installations

inside the hotel building itself are of great concern. The air-conditioning and

ventilation system is not just provided for the proper internal environment to the

guests and hotel employees but is also a major factor to be considered in fire safety

aspect. The running o f the air ducts [ASHRAE 1997] through fire resisting

construction and proper fire dampers and stop details to be installed are critical.

The material used in the ventilation system is a major fire safety concern both

locally and abroad and must be considered carefully. O f critical importance, the

smoke control mechanism in the ventilation system in hotels is both a practical and

academic focal point to the operators. On the other hand, the electrical building

services systems, including the lighting system, are also a major concern to fire

safety in hotels. In the author’s hotel operating experience, most o f the fires in

hotel guestrooms are caused by faulty electrical equipment or the improper usage o f

these electrical installations. Training to hotel staff is important, particularly on

strict adherence to electrical ordinances and standards upon using all these electrical

eq u ip m en t.

Furthermore, in this thesis, the fire safety and risks in particular functional areas in a

hotel, such as the kitchens and laundries, will also be emphasized. There are lots of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 electrical and mechanical equipment and installations in the kitchens and laundries.

The installations involve various kinds o f fuel and large amount o f oil, and dust will

be generated during the operational processes. In this study, the fire risk in

kitchens will be focused on as nearly all the hotels, no matter its ranking and service

standard, will have kitchens servicing the guests and employees. The risk and fire

load [Chow 1999] in kitchens are high and an extensive study on the safety aspects

in kitchens will be conducted. Experiments will also be carried out to find out the

heat release rate for a possible oil fire in the kitchen.

1.1 Objectives o f the Research Work

The objectives and key items in this thesis:

To investigate the fire safety concerns and problems in hotels, in particular the high

risk area — kitchens. Fire codes were reviewed for hotel accommodation. The

local fire codes involve those from the Fire Services Department [FSD 1998] and

Buildings Department [BD 1995, 1996a, 1996b]. Mainland fire codes [Ma 1995,

1997; MPS China 1997] and National Fire Protection Association (NFPA) fire

standards and guidelines [NFPA 1993, 1994, 1995] were also compared and

considered.

To study the total fire safety concept [Chow 2004b; Hui 2004] and its application in

the practical hotel operation. The importance o f fire safety management [Chow,

2001a; Dailey, 2000] in hotels will be focused on.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 To carry out full-scale burning tests [Babrauskas 1992, 2003; Chow 2001b, 2002,

2004a] and investigate the fire scenarios and heat release rate for kitchen fires. The

effects on operation o f sprinkler [Gupta 2001; Watts 2003] and water mist system

[Mawhinney 1993, 1994; Yao and Chow, 2001] on kitchen fires will be studied.

To carry out experiments [Chow and Ho 1992; Heskestad and Bill 1984, 1988,

1989] in rating fusible links installed in fire dampers for hotel guestrooms. This is

critical to the passive fire protection systems and the correct rating [UL33 2005] of

the fusible link is important to the damper operations.

1.2 Methodology

Local and international chain hotels in Hong Kong were surveyed to investigate the

fire safety concerns and problems in big hotels. Local fire codes and those from

the Mainland and the US were compared and reviewed. Hotel licensing

requirements [CNTA 2005] were particularly dealt with.

To understand the importance and application of total fire safety concept in Hong

Kong, particularly the practice and program o f fire safety management [Della 1999]

were reviewed.

Very few experimental data are available from systematic full-scale burning tests

[Chow 2002], especially for certain building configurations such as the hotel

kitchens. Full-scale burning tests [Chow et al 2003d] on a model kitchen in a

remote area o f China were carried out to determine the heat release rate and also the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 effects o f sprinkler and water mist system on the kitchen fire. Full-scale burning

tests are important to determine the actual fire environment and material behaviour

under fire condition.

The operation and performance o f fusible links in smoke curtains and fire dampers

were reviewed. The empirical equations [Chow and Ho 1990] and models

involved for sprinkler heads were studied. Simulation experiments were carried

out to verify the performance o f the fusible links currently used in one o f the local

hotel projects. Response time index [Chin 2002] o f the fusible link was evaluated.

The experiments are critical for ensuring that correct rating and performance of

these fusible links are achieved in actual operating conditions.

1.3 Outline o f the Thesis

Chapter 1 is an introduction to hotels nowadays. Fire safety is of critical

importance to hotel operations. The various mechanical and electrical installations

are discussed and their importance towards fire safety is outlined.

Chapter 2 is a detailed investigation on hotel fires. The fire safety concerns and

issues related to fires in hotels will be outlined. Electrical and mechanical

installations in the hotel building services system will be studied and its operation on

fire safety concern will be investigated. Particular fire risks in certain functional

areas will be studied, for example, laundries and kitchens. Local [FSD 1998] and

foreign fire codes [NFPA 1993, 1994, 1995] will be studied, particularly those codes

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 directly related to hotel and guesthouse accommodation will be reviewed. These

codes are implemented in Hong Kong in early 1990s and are under continuous

revision to achieve both fire and building safety standard. The importance of fire

safety concern in international chain hotels is specially discussed. This relates to

the hotel image and reputation if a correct fire management scheme has been

successfully implemented.

Chapter 3 is an in-depth study on the total fire safety concept and its application in

daily hotel operations and activities planning. The hardware and software

components and also the fire safety management strategy [Shipp 1998] and planning

in hotels will be studied. To the hotel operators, fire safety management may be

the most important aspect relating to the asset and life in the hotel. The various

components and attributes in the passive fire protection systems and installations

will be outlined. This passive system [Tsui and Cheung 2004] may be the first

defence and protection against fire spread. The correct selection and rating o f the

materials and installations will be critical to the success o f these passive fire

protection systems.

Chapter 4 will focus on the active fire protection systems [NFPA 1997], which

comprise all the dry and wet fire fighting systems. Various codes related to these

systems, their operation, characteristics and suitability on their usage in various

hotel functional areas will be reviewed. Just like passive fire protection systems as

described in Chapter 3, the importance and implementation o f engineering

performance-based fire codes to replace the traditional prescriptive codes will be

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 discussed. More advanced systems such as the water mist system and chemical

system will be examined.

Chapter 5 is to discuss and investigate fire safety management in hotels. This

software component which is o f paramount importance to the hotel will be studied

in depth. For the hospitality service industry, a proper strategy and training plan

should be established in order to cater for the changing service environment and

different guests’ expectations nowadays. The various components [BSI 1994] in a

fire safety management scheme will be explored further.

Chapter 6 is on the methods used for this study [Wilson 1990]. Basically, fire

safety concerns and theories for hotels will be reviewed. This forms a basis for

further research. O f critical importance, experiments were carried out to

investigate various fire scenarios o f kitchen oil fire and the operation o f sprinkler

and water mist system. A remote site in China was selected to perform the

experiments in part o f the actual kitchen setup.

Chapter 7 is a review on the site selection and its characteristics. The

environmental concerns on these experiments [Chow et al 2003d] are particularly

dealt with and the necessary expertise in setting up the experiments is outlined.

The actual experiments carried out for kitchen oil fire are outlined. All the results

will be recorded and analyzed. The most important parameters, such as heat

release rate, maximum temperature recorded and the fire extinction time will be

reported.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 In Chapter 8, empirical equations for heat transfer [Heskestad and Smith 1976,

1980] at sprinkler heads will be studied. Heated wind tunnel tests on sprinkler

heads were simulated for testing the actuation o f fusible links. Experiments were

carried out on rating the fusible links for fire dampers installed in hotel guestrooms.

The data obtained in the experiments will be reported and analyzed.

Finally, in Chapter 9, conclusions for the whole study will be given and further

studies recommended.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 2 HOTEL FIRES

Hotel buildings are very complex on its functional use and also change with time for its

major clienteles. The familiarity o f various types o f guests coming from different

countries with the hotel configurations is usually a question mark. On fire safety

concern, as only the hotel employees are familiar with the hotel layout, fire exits and

safety installations, etc., a proper training plan to staff [Malhotra 1987] is o f paramount

importance. Since usually the hotel guests cannot easily find the locations o f the fire

fighting installations and the exits or staircases, in accordance with the local or

international fire code, proper lists of fire service installations must be provided.

In this chapter, some past experience o f fire examples in hotels will be listed. The fire

safety concerns on the electrical and mechanical installations in a hotel building will be

discussed. International chain hotels will have their particular focus on fire safety and

installations required. The requirements and the special installations that these chain

hotels in Hong Kong may have on fire safety concern will be outlined. Furthermore,

particular risks in a hotel building, such as the kitchen, will be discussed. The particular

problems and issues related to fire safety in kitchens will be reported. Besides the fixed

fire fighting systems and the fire resisting construction [England et al. 2000] provided,

the software component such as the fire safety management scheme provided by the hotel

operator is equally important to guests on fire safety aspects [Bickerdike Allen Partners

1996].

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 2.1 Review on Parameters Related to Fire Safety and Problems in Hotels

Before embarking on the fire safety problems often encountered in hotels, some examples

o f past fire incidents in hotels will be described. These fire occurrences were actual

happenings although the fire size was not large and did not cause any casualties.

2.1.1 Examples o f fires in hotels

• Fire in guestrooms

A lot o f electrical and mechanical facilities are installed in a typical guestroom. Their

characteristics and effects on fire safety will be discussed in more detail later in this

section. Besides, the guestrooms are usually fitted out with a lot o f fabric and finishing.

Guests smoking in guestrooms are a primary source of fire happening in guestrooms.

Some o f the guests are heavy smokers and they do not handle the cigarettes ends well

which leads to the fire outbreak. Although these fires are usually extinguished by the

sprinkler system [Cheung and Chan 2003] inside each guestroom, the water damage and

the panic caused may be very significant.

Moreover, in modem hotels, there are also some service apartments that serve a certain

amount o f long-staying guests residing for years or months. Quite often, some pantry or

kitchenette equipment will be provided for minor cooking in the guestrooms. This will

pose a high fire risk to both the guests and hotel employees. If the cooking process is

unattended in the guestroom, it may lead to a fire accident. Furthermore, the

guestrooms are also equipped with many electrical appliances or heaters. If these

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 equipments were not correctly used and properly maintained, a fire would occur and this

actually happened in the past. For example, water boiled in kettles and equipment

malfunctioned without electrically cut off when boiling temperature was reached. This

concerns the quality and reliability o f the equipment. Similarly, some guests misused

the hairdryer as a clothes dryer and that caused a fire. In winter, clothes or fabrics in the

guestroom would be ignited by a heater if unattended or heating surface being too close

to the finishing.

• Fire during hotel renovation

Again, this concerns the quality o f project management and the attitude o f hotel

management towards the contractors. A lot o f these fire examples happened in the past,

though the fire was put out by the contractor at the very initial stage. Most o f the

construction workers are smokers and they break the rules to smoke on the site even

when they are forbidden to do so. This is a fire risk to the hotel because the lighted

cigarette ends in the site can lead to a big fire.

Careless welding is another source leading to fire during the hotel renovation. For

electrical and mechanical installations in modem buildings and certain steel or metal

framework construction for the structure and window installation, these works involve

lots o f welding work. The poor and careless attitude o f the workers during welding will

leave lots o f lighted or high temperature soot on site. If not properly cleaned or

removed, this soot will be a fire source on site. There are several examples in the past

hotel renovation projects such as the removal o f old escalators in the lobby in order to

create an atrium [Morgan and Marshall 1975], intense welding will be needed and

10

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 actually the author had seen a fire due to such welding process. Same as in another

upgrade project on installation o f a new air handling unit for the laundry room, intense

welding was needed and the sparks actually lighted the dust and lint accumulated inside

the equipment. This is an example o f poor welding technique and inefficient

maintenance and cleaning procedures implemented in the past on the equipment. Both

o f these two examples o f fire due to careless welding did not cause any casualties but the

damages to the hotel building were significant. In both cases, hotel evacuation was

required and that caused panic to hotel guests and employees.

• Fires in kitchens

Typical fire examples in kitchens will be described in later chapters.

• Fires in laundries

Laundry in a hotel is equipped with many large electrical and mechanical installations.

It is a factory and indeed a very labour intensive operation in the hotel. There are also a

lot o f equipment in the hotel laundry, including folding machines, flatwork ironers,

pressers, compressors, dry cleaning machines, washers and form finishers, etc. High

rating motors, chemicals and large exhaust volumes are typically involved in daily

laundry operation. Detailed and efficient maintenance and cleaning programs and

operation are necessary to keep the equipment in a safe condition. All the heated

surfaces or motors should be kept away from linen or clothes. All lint and dust on the

equipment interior or surface should be properly removed. Several fires had happened

in the past in the laundry, though most o f them were put out by the hotel internal fire

11

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 team and did not cause any casualties. One example involved the 50 HP motor o f the

washer while it was washing about 100 kg of laundries. Due to poor maintenance and

carelessness, the dust or lint on the motor surfaces was ignited from the motor heat. The

fire was put out by portable extinguisher but the washer was damaged as the motor and

washer body itself was seriously burnt.

• Fire due to lighting installation

Lighting provides the suitable visual environment in the hotel lobby, restaurants,

conference halls and guestrooms, etc. Lighting is understood to generate heat, in

particular for those high power rating fittings usually adopted such as halide lamps or

floodlights. Lighting systems are usually not regarded as a high fire risk and not so

many concerns are given to minimize the chance o f fire occurrence. In particular, the

electrical characteristics, circuitry and power rating need to be checked in order to

prevent the occurrence o f a fire.

The focus effect o f some halide fittings might also lead to fire. There had been several

such fire cases in hotels, such as the reading lamp installed at the bedside panel and the

focus heat actually burnt the duvet on the bed. Similarly in the restaurant, when the

guests’ clothes were accidentally placed too near to the column light fitting [Canter

1980], the focus heat from the light fitting burnt the clothes and the hotel had to

compensate for this damage to the guest. Moreover, the hotel management also needs to

install a lot o f lighting stripes on the decorations and finishing at the front o f the house

areas, especially during festival seasons. The electrical quality o f the whole installation

12

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 has to be checked by qualified electrician so as to prevent fire due to overheating or

electrical failure.

2.1.2 Problems and concerns for fire safety in hotels

From aesthetic point o f view, hotel buildings are similar to other types o f buildings when

viewed from outside. Hence, there is a thinking to provide a global solution in the

design o f hotels on the provisions for visual, thermal and safety aspects [Buchanan 1999]

so that hotels can be operated similar to offices, hospitals or government buildings. This

concept however may cause confusion and in most of the cases do not tally with the

operation needs o f the hotel management.

According to the local fire code [FSD 1998], whether the prescriptive code or the

engineering performance-based fire code is adopted, it is still required to provide the fire

protection and suppression systems for the hotel building itself. Similarly, other

electrical and mechanical systems are also provided in the hotel for the comfort and

safety o f the guests. Moreover, the fire safety system is usually incorporated into the

hotel central building management system. This system will manage the operation,

status and commands the fire safety system operation in the hotel.

On the contrary, the operational characteristics and service needed to be provided to

guests pose additional difficulties and problems on fire safety in the hotel premises.

First o f all, the needs for hotel guests differ significantly from those o f office staff or

clients to a theatre. Basically, guests will sleep in the guestrooms and dine in the

restaurants. Because o f the accommodation provided inside an unfamiliar hotel

configuration and guestrooms, added concerns should be placed on fire safety to

13

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 safeguard the life and assets inside the hotel. In some o f the hotels nowadays, it is a

usual client segment to have long-staying guests or service apartments. To better cater

the needs o f the guests, warming or even simple cooking equipments are provided in the

guestrooms. This means that the type and profile o f guests inside a hotel really is a

concern on fire safety.

Secondly, the daytime and night-time requirements o f a hotel guest will be very different

from other building users. This relates to the specific and so many services that the

hotel will offer to the guests. During daytime, the guestroom may not be occupied as

the guests will go out for site scenery if they are tourists or attend meetings or

conferences for corporate guests. On the contrary, some corporate guests may also use

the guestroom as an office for all the business function requirements and meetings, as a

work station or a lounge, etc.

Thirdly, a main outstanding difference is that a hotel will operate 24 hours daily for the

year round. This definitely will pose additional concerns on the design, installation and

maintenance [Della 1999] on the fire service installation to be provided. Seemingly,

there is no downtime for nearly all the installations provided in the hotel. Whenever a

maintenance program is needed for the fire service installation, alternative systems have

to be considered in order to provide an equivalent fire safety level to the hotel. The

hotel management also needs to face challenges by having a team o f qualified engineers

that would handle any situations without undermining the fire safety standard.

Fourthly, modem hotels also host various types o f activities and functions in its lobbies,

convention centre and conference halls. The functions range from simple pleasurable

14

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 events to major business meetings, product launching and even leisure and sport

activities. Recently, the Hong Kong Special Administrative Region government also

hosted some World Trade Organization (WTO) meetings in hotels. There may be a

sudden gathering o f large crowd o f people inside and outside the hotel. This will be a

big challenge to the hotel management and they often need assistance from the local fire

brigade on how to deal with this kind o f emergency situation. In the recent WTO event

in Hong Kong, the Fire Services Department basically tested again all the fire services

installation for all the hotels hosting any WTO delegations. All the internal hotel fire

teams were on full alert during the whole WTO event, together with the full support from

the local Fire Services Department. The various functions which might include

fireworks, pyrotechnics and cooking demonstration added the fire risks although the

function lasted for only a very short period inside the hotel. The hotel management

needed to think very carefully in each function on how to provide safe environments for

these functions.

In the last decade and in modem hotels, the hotel management needs to provide “personal

service” to the guests. The guests will pay for the hotel in order to have the equivalent

service provided for them to feel like their own home. This “personal service” or

individualization o f comfort [ASHRAE 1997] will surely pose additional concerns on fire

safety.

2.1.3 Hotel functional areas

To the hotel operator, there are three essential elements in the hotel as far as the space

allocations are concerned. They are the private zones, the front o f the house areas and

the back o f the house areas. Private zones are basically the hotel guestrooms, suites and

15

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 service apartments where fire safety in these areas is our top concern. For the front of

the house areas, they include the restaurants, hotel lobby, health clubs, conference halls,

meeting rooms, shopping malls and lounges, etc. These areas also serve the hotel guests

and usually will have large crowd o f people. In the back o f the house areas are the

laundry, kitchens, hotel in-house offices, storerooms, plant room and loading bay.

Particular focus will be placed on the fire safety in these areas as they may not be

occupied sometimes, for example, the plant room. Hence, serious considerations on fire

safety in the following areas in a hotel must be raised:

• Guestrooms, suites and service apartments

As related in previous sections, fire safety in these areas is o f top concern to the guests as

they are residing in the room particularly during the night-time. Various activities may

be conducted in the guestroom, for example, smoking, cooking and the improper use of

electricity may all increase the fire risk in the guestroom.

• Plant rooms

In the local fire code, a special license, such as the dangerous goods license, has to be

applied for certain special plant rooms. The E & M plant rooms may include main A/C

plant room, boiler room, main switch rooms and swimming pool filtration room, etc.

Special and even individual fire protection systems are provided for these plant rooms.

16

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 • Food and beverage areas - restaurants and convention halls

Cooking activities and smoking are major concerns in these food and beverage outlets.

Moreover, the gathering o f large group o f guests in these areas creates additional

concerns on fire exit and evacuation. In trendy hotel restaurant design, the open kitchen

concept is often adopted and actual cooking being done in the restaurant and this must be

a challenge to the fire system designer and the hotel management.

• Laundry plant

It is both a factory and a plant room to the hotel. As a factory, it is labour intensive with

many hotel employees working at the same time in an area, typically densely installed

with large electrical and mechanical systems. As a plant room, it houses many laundry

equipments and the handling o f a lot o f detergents and chemicals.

• Back o f the house areas

They consist o f all the supporting areas for servicing the hotel guests and employees. In

particular, fire may occur in the loading and unloading bay and car parks. Flammable

vapour is involved if the petrol is not properly handled. All the main electrical and

mechanical systems serving the hotel building are basically located in the back of the

house areas. Hence, proper and detailed program on maintenance o f the installations

must be implemented. For international chain hotels, usually the corporate office will

have periodic visits by senior management personnel to carry out audit on the quality and

standard o f all these back o f the house areas.

17

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 • Kitchens

Details will be discussed in the subsequent sections.

2.1.4 Electrical and mechanical systems in a hotel

To the architects, hotel is a prestigious building and the guestrooms usually occupy about

60 to 70 % o f the total gross floor areas approved for the hotel project. For the success

o f the hotel building, the architect needs to consider not just the layout o f the guestroom

but also the structure [Buchanan 2001], the construction and finishing materials and the

coordination o f all the electrical and mechanical systems. To the hotel operators and

owners, they have to maintain the profitability and competitiveness o f the hotel building.

This cannot and should not be in conflict with their primary objective to provide a safe,

clean and comfortable environment for the guest.

It is the prime responsibility o f the hotel management to maintain properly the following

installations:

• Electrical installation and appliances

Proper use o f electricity and strict adherence to electrical codes and ordinance must be

enforced by the hotel operator. A proper selection o f electrical devices and appliances

are essential and should be carefully considered during the purchasing stage.

18

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 • Air-conditioning and mechanical ventilation system

Such system not just provides a comfort environment to the guests and hotel employees

but it is needed to provide a clean and hygiene environment also. All the fusible links,

dampers and fire devices installed in the ventilating system must be checked and

maintained properly.

• Fire protection system

Active fire protection systems include the sprinkler system, hosereel and hydrant system,

fire alarm system and portable extinguishers, etc. Passive fire protection systems

include the building construction with proper fire rating, fire doors or walls, dampers,

seals and smoke lobbies, etc.

• Extra-low voltage systems

They are essential systems to the proper execution o f the fire evacuation procedures,

including the public announcement systems with overriding features.

• Laundry and kitchen equipment installations

Care and proper maintenance must be implemented to remove the oil deposit, or dust or

lint collected in the kitchen and laundry equipment respectively.

19

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 • Cold and hot water system

Proper maintenance and care are required for the fire protection system for the boiler and

calorifier room.

2.2 International Chain Hotels - Fire Safety Considerations

Every guest has its own expectation towards a hotel. To most o f the guests or chain

hotel management group, a hotel is an area designated for guests to have a pleasant stay

and safety environment under a quality and effective infrastructure. The International

Hotel and Restaurant Association (IHRA) defines a hotel as “an environment which

always delivers overnight accommodation” and which is licensed to market its business

as a hotel.

International hotel chains with their main objectives to fully satisfy customers’ needs,

always try to offer a wide and varied range o f guest services. These hotel chains bring

in more spacious rooms and offer a wide range o f refined food and beverage services.

This is particularly obvious in the 1980s with China adopting an open-door policy that

attracted even these international hotel experts to develop their management expertise in

the Far East. Nowadays, hotel industry becomes more and more competitive. The

international hotel chains push for more o f their expansion plans o f corporate and

business people in the Far East, Europe and the Middle East. All these hotel chains, not

just talk about luxury o f bedrooms, state-of-the-art facilities, friendly surroundings and

20

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 high quality environment, but they also must focus on the basic concept for a hotel, i.e.

the safety provided to guests.

2.2.1 General fire safety requirements for an international chain hotel

In this section and in accordance with the author’s experience, the general fire safety

requirements will be outlined in an international chain hotel. The areas included are

public areas, guestrooms and back o f the house areas. Credits are allowed for sprinkler

protection, recommendations for fire safety for existing finishing and fire alarm

procedures. These requirements lay down the minimum standards that are acceptable in

a chain hotel in order to protect the human life and asset in a hotel.

• Fire warning system

In a chain hotel, there should be a means for giving warnings [Canter 1980], in case of

fire, to people in the building.

A fire warning system consists o f detectors and manual call points strategically situated

throughout a building, which are connected to a control indicator panel. Depending on

the fire warning system installed, information will be displayed at the control panel. In

the case o f addressable system nowadays, the location o f the actuated detector or

breakglass unit will be given on an automatic liquid crystal display (LCD) at the control

panel.

21

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 In the event o f a fire, the control panel will identify which zone, detector or manual call

point has been actuated and therefore the location o f the fire. It will also actuate the fire

alarm for evacuation, etc. Some fire warning system may also be connected to fixed

extinguisher systems and have a direct communication route to the local fire brigade (like

the Chubbs link in Hong Kong).

In addition, the fire warning system may also include the connection o f relay contacts

which may be regarded as the interface between the warning system and the main

building services, such as lifts, air conditioning systems, dampers, etc. Operation o f the

warning system will prevent these services from working, or will change their mode of

operation, i.e. lifts will home down automatically to the lobby floor. Air-conditioning

system will be switched off and dampers will close down all open ducts or orifices in

order to prevent the spread o f fire or smoke throughout the hotel [Cheny 1986]. The

aim o f a fire warning system is to ensure early evacuation of the building and to reduce

the physical losses caused by fire damage.

Fire detectors detect one o f the three characteristics o f fire: heat, smoke or flames

[Morgan and Gardener 1990]. Heat detectors are suitable for most buildings and show a

bigger resistance to adverse environmental factors. They are good for detecting fires

with very rapid heat evolution but little or no smoke. Due to the lack o f response to

smoke, heat detectors are not suitable for most areas in a hotel and are only installed in

areas where smoke detector could give unwanted alarms, e.g. kitchens.

Ionization smoke detector works by detecting a change in electrical current within the

detector as the smoke particles interfere with the system and equilibrium o f charged ions.

22

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Generally, ionization smoke detector will respond quickly to smoke containing small

particles but have a less rapid response to smouldering fires.

On the other hand, manual call points [Della 1999] are the square “breakglass” appliances

strategically located around the hotel building. Generally, no point in a building should

be further than 30 m from a call point, although this distance should be reduced where

there are specific fire hazards.

Detectors and call points are arranged in zones, which may be considered as unit fire

compartments. Besides showing the location o f a fire, the control and indicator panel

should also provide information on power supply fault and wiring faults, etc.

Where intercommunication or public address equipment is used in lieu of conventional

sounders to transmit a general alarm, the signal should take priority and override other

facilities o f the equipment. The alarm signal, which may be followed by a voice

transmission o f essential information for safe evacuation should be distinct from other

signals which may be in general use on the system. In the event o f a false alarm, it is

important that the cause and extent o f the problem is determined as quickly as possible.

Installations which are the subject o f a series o f false alarms should be referred to the

designer. The provision o f a two-stage alarm system may need to be considered in some

nuisance case.

23

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 • Portable fire fighting equipment

The sitting o f portable extinguishers should always be as close as practicable to the fire

risk, adjacent to exits, on escape routes and away from extremes o f temperature.

Portable fire fighting equipment should be securely fixed to wall with the carrying

handles approximately 1 m from floor level. Fire blankets should be provided in high or

special risk areas, for example, in all kitchens and workshops. It is most important that

portable fire appliances are kept in their allocated positions and not misplaced or

misused.

Extinguishers are made in a number o f sizes. If the allocation o f water extinguishers is

to be based on the criterion floor area, then one 9 litre extinguisher should be provided

for every 200 m2 o f floor area [BSI 1988], In deciding what will constitute appropriate

means for fighting fire within the hotel, it is necessary to consider the nature o f the

materials in each area and thus the class o f fire likely to be involved.

On the contrary, hosereels are strongly recommended particularly where the floor areas

are in excess o f 800 m2 [BSI 1988]. They are the most effective extinguisher, depending

on the class o f fire. Once they are in action, a substantial volume o f water can be

provided and the water supply can also be controlled. For ease of handling, it is

normally convenient to restrict the length o f hose on one reel to 30 m o f 19 mm diameter

internal bore hose.

24

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 • Fixed fire fighting equipment

Water sprinklers, due to the effective extinguishing ability and accessibility o f water,

provide good general protection when installed in hotels. However, in situations where

the use o f water could be hazardous or when it is essential to avoid or at least minimize

water damage, o f which special risks are present, the installation o f alternative systems,

such as carbon dioxide, foam, halon or dry powder should be considered. In many cases

in hotels, the extinguishing systems will operate automatically and may also be connected

to the detection and alarm systems to provide complete protection.

Water sprinklers comprise a system of pipework, pumps, control valves and heat

sensitive valves in the sprinkler heads which release water onto the seat o f a fire. The

rate at which the water is released will depend on the severity o f the fire, given the nature

o f any combustible materials present. It is the flow o f water which actuates any

connected alarm systems. A clear space o f at least 0.3 m should always be maintained

below the level o f the sprinkler head throughout the room.

Besides water sprinkler system, foam installation can also be used in special cases.

Foam installation is a self-contained system capable o f making, conveying and releasing

foam. The foam may be low, medium or high expansion foam.

For carbon dioxide (CO 2 ) installation, the gas is usually stored in a cylinder or

refrigerated tank and is connected to the outlet nozzles by a system o f pipework. Such

system may be designed to operate automatically or manually. Carbon dioxide systems

provide good protection for hazardous plants in a hotel, e.g. transformer areas, electrical

25

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 equipment and computer facilities, etc. One disadvantage o f this system is the reduction

o f oxygen in the protected area on actuation, which necessitates built-in warning and time

delays to allow evacuation o f the area.

On the other hand, the components o f halon systems are similar to those in carbon

dioxide installation, although less storage is required for the same level o f protection, due

to differences in the extinguishing action. There are two types o f halon systems: “Halon

1301” and “Halon 1211” [NFPA 1997]. Generally, halons are used for protection in the

same sort o f area and situation as carbon dioxide system. Like carbon dioxide system,

the operation o f halon system has health and safety implications due to the toxic effect of

the halons themselves. In addition, halons are known to affect the atmospheric ozone

layer and are now only used when no suitable alternatives exist.

Similarly, the dry powder is kept in a pressurized container, or is connected to a gas

cylinder with a system o f pipework leading to the outlets. On actuation, the powder

become fluidized in the expellant gas and is conveyed to the outlets. These systems are

suitable for fires involving flammable liquids, electrical equipments or where water

damage must be kept to a minimum. They are not suitable in situations where

re-ignition may occur.

• Emergency lighting system

Adequate lighting, whether natural or artificial, is important in enabling hotel guests to

locate an escape route and move quickly along it to a place o f safety. Hotel emergency

lighting is required not only on complete failure o f the normal power supply but also on a

26

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 localized failure if such a failure would present a hazard. Any new emergency lighting

system should conform to the local codes. The normal standards require that emergency

lighting should maintain an adequate level o f illumination for two hours in accordance

with the hotel licensing requirements [CNTA 2005]. The greatest need o f emergency

lighting is those areas where there are guestrooms, corridors and escape routes.

There are several types o f emergency lighting systems. Emergency light fittings are

classified according to the type o f power supply they use. The electrical power needed

to operate them may come from one o f the three sources:

a. Individual batteries in each light fitting

b. A central battery system which powers all the luminaries

c. An electrical generator which powers all the luminaries

Luminaries in which power is supplied from a central source, i.e. the central battery

system or a central generator, are referred to as “slave” luminaries. This is in contrast to

“self-contained” luminaries where the power source is contained within the luminaries.

In addition to the relative positioning o f the power sources to the luminaries unit, it is

also possible to have “maintained” and “non-maintained” emergency lighting systems.

“Maintained” systems refer to those where the luminaries are operational at all times.

“Non-maintained” systems only become operational when the normal power supply fails.

27

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 • Means o f escape in case o f fire

The basic principle o f satisfactory means o f escape [Tsui and Cheung 2004] in a hotel is

that persons should be able to walk unaided to a place o f safety, regardless o f where a fire

might break out in the hotel building. Satisfactory means o f escape should be an

integral part o f the hotel building construction. This excludes portable ladders and

self-rescue devices.

Hotel guests or employees should be made aware o f the escape route from their particular

location. The escape route should be easy to follow and adequately signposted. The

following factors should be considered:

a. The routes should be sufficiently wide, and of sufficiently short distances to allow

speedy and safe evacuation.

b. Ideally there should be alternative routes leading in different directions to that all

persons should be able to turn their backs on a fire. However, a single exit may be

permissible when the travel distance is limited and where the fire risk is low.

c. The route should enable everyone to escape unaided.

d. The route should lead directly to the open air or via a protected route to the open

air.

e. The distance between work stations and the nearest fire exits, should be minimized.

The travel distance should be kept as short as possible. It is necessary to know the

number o f people likely to be inside the building when examining means of escape.

Except where a single exit is sufficient, one of the exits should be assumed to be

28

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 obstructed by fire. Therefore, in the case of two exits, each should be capable o f letting

all the occupants through. When more than two exits are provided, the largest should be

discounted in accessing the aggregate widths o f the others. Exits should not be

narrower than other parts o f the escape routes. Lift and escalators are not usually

acceptable as part o f an escape route.

• Fire resisting doors

Fire resisting doors are provided to prevent the spread o f heat and smoke and to protect

escape routes. The resistance o f a fire door is determined using tests approved by the

local fire or building authority. Two aspects o f fire resistance are measured, namely the

stability and integrity. All fire doors provided for the protection o f escape routes should

be fitted with smoke seals.

All fire doors, except those to cupboards and service dusts, should be fitted with

self-closing devices to ensure the positive latching o f the door. Examples o f suitable

devices are weights, floor or overhead springs and overhead door closers. Self-closing

doors should be permanently marked on both sides, at about eye level, with the words

“Fire door - Keep closed”. Fire resisting door to cupboards and stores on escape routes,

which are not provided with self-closing devices, should be permanently marked on the

outside at about eye level, “Fire door- Keep locked closed”.

29

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 • Fire exit doors

Exit doors on escape routes which are not the normal route o f travel from the hotel should

be indicated by notices bearing the words “Fire exit”. If such an exit door is not visible

from any point on the escape route, a notice in the same white and green coloring should

be provided at such a point, reading “To Fire Exit” and accompanied by directional

access if necessary. Illumination o f exit and directional signs should be provided by an

emergency service, e.g. an emergency generator or batteries, independent o f the main

supply.

2.3 Critical Fire Hazard Area in a Hotel —Kitchen

As mentioned in the above sections, kitchen is a high fire risk area. Complex and

intense application o f various building services systems and MVAC system are installed

in the kitchen. In this section, various systems in the kitchen and the concerns and

problems on fire safety will be mentioned. Also, training for kitchen staff and

maintenance management o f various systems are equally important to prevent the

occurrence o f fire.

2.3.1 Introduction

The fire loading and risk in a kitchen are high, not just because o f the flammable grease

deposited in the MVAC system but also there are various types o f fuel used in the

kitchen, such as towngas, electricity, solid fuel or even LPG. Particularly at night, when

30

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 the number o f employees in the kitchen is low with the steward staff doing the cleaning

operation, the outbreak o f a fire will be a surprise to everyone. Most o f the fires that

happen in the kitchen relate to the poor maintenance of the building services system or

the carelessness in the operation of the kitchen equipment.

First o f all, in international chain hotels, chefs o f various nationalities are employed to

run the kitchen operation. Also, during festive seasons, a lot o f casual or short-term

workers will be working in the food and beverage outlets and hence adequate training is

considered for them to handle a fire outbreak.

Secondly, in case o f peak business seasons, good kitchen management is important to

avoid carelessness in handling kitchen equipment and the procedures in the kitchen for

preventing fire. Moreover, as discussed before, the fire loading in the kitchen is high

and any reckless operation will cause a fire. Furthermore, there are very complex

electrical and mechanical systems in the kitchen and the correct maintenance program is

usually neglected because these maintenance works are usually carried out at night where

the supervision is poor.

2.3.2 Electrical and mechanical systems in a kitchen

Kitchen equipment and their proper cleaning and maintenance are critical.

Electrical Installation - Electrical system reliability is important in this damp, oily and

hot environment. It is required to focus on the correct procedure to handle electrical

equipment and the safety aspect raised.

31

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Steward Cleaning Equipment - Again, this is critical for the electrical safety and the

maintenance operation required.

Fresh Air and Exhaust System - This is required as per local hygiene code and

operational needs. As conveyed before, proper maintenance to guarantee its

performance is the top priority.

Fuels - In local license code, only towngas and electricity are allowed and extra fire

protection measures are needed for other fuel, e.g. solid fuel.

Dumbwaiters and Food Conveyor - This is critical for the passive fire protection system

as fire or smoke can spread to other parts o f the hotel if the wall or floor openings are not

sealed properly.

Fire Protection System - special fire protection systems, such as FM 200 or fast response

sprinkler heads may be adopted.

2.3.3 Fire protection systems in a kitchen

Fire protection systems in a kitchen include:

a. Water sprinkler or water mist system [Andersson and Holmetedt 1999] is

required by local code and the correct fitting and rating o f the sprinkler heads

are critical to the extinguishment o f the fire. Correct rating o f the sprinkler

32

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 heads [BSI 1990] is crucial to avoid false alarm and the improper burst of

sprinkler head that cause significant operational difficulties and financial losses.

b. Hosereels are also required as per local fire code. All the kitchen employees

must be trained on how to operate this portable equipment.

c. Fire extinguisher and fire blanket are usually the first defence again any fire

outbreak. A lot o f minor kitchen fires are put out by the CO 2 or water

extinguishers, and the damage to the kitchen equipment and related facilities

will be minimized. Proper training on their usage is important, particularly

when there are a lot o f casual and temporal workers in the kitchen.

d. Detectors and manual pull stations are required as per the local fire code and

licensing requirements. They are the first alarm and any defects must be

cleared as a priority. Correct sitting o f these detectors are important to avoid

false alarms which will be a nuisance to the operation.

e. BTM system and FM 200 [NFPA 1997] are usually used in Chinese cooking

range and teppanyaki cooker respectively. They are expensive and delicate

systems and must be handled and designed with care.

f. Emergency lighting & battery luminaries are required as per the local fire code

and licensing requirements. Usually, these back-up lighting are needed to

maintain for two to three hours.

33

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 3 TOTAL FIRE SAFETY CONCEPT APPLIED TO

HOTELS

3.1 Introduction

Total fire safety concept should be implemented in hotels and basically four major factors

will be considered [Chow 2004b]:

• Risk factor, such as the occupancy load factor (OLF) and the fire load density (FLD)

• Fire services installation (FSI)

• Passive building construction (PBC)

• Fire safety management (FSM)

In subsequent sections, each of the above factors will be discussed in more detail. To

the hotel management team, all these four factors are important and critical to fire safety

[Building Regulations 1985] in the hotel. However, with changing guests’ needs and

hotel building usage, fire safety management strategy, the software component o f the

total fire safety concept, is considered even more important and should be focused so that

all the hotel staff are familiar with all the necessary procedures, strategies and measures

[NFPA 550 1995] on fire safety. In this section, the total fire safety concept in hotels

and its development will be outlined. The passive building construction in a hotel will

be discussed in detail.

34

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 3.2 Total Fire Safety Concept in Hotels

Fire engineering is quite a new subject and being developed rapidly in recent years. It is

a multi-disciplinary subject, involving physics, chemistry, fluid mechanics, computer

application, electrical and mechanical engineering and instrumentation. With the recent

big fires in Hong Kong, Korea, USA, China and Europe and the current trend of boom

economy, all academics and practitioners are facing escalating challenges on how to

solve all these complicated fire safety problems. The related fire science issue becomes

more and more complex and hence the subject o f fire engineering always is an integral

part o f public safety science concern [Mckinnon 2003] and gradually receiving much

more attention. To the hotel management team, total fire safety concept includes and

relates to nearly most o f the contents o f the subject “fire engineering” and actually

reflects the key rationales behind the subject “fire engineering”.

Normal operational mode o f the building often has to be maintained and the fire safety

inside the hotel premise to be kept. To better protect the hotel building itself and

minimize the damages due to possible fire outbreak, not only the active fire measures

should be provided during the building design stage. The active fire measures basically

directly prevent the outbreak o f fire and also are technical designs to control and restrict

the fire growth and development to other hotel areas. These active fire measures

include the fire alarm and detection at the initial fire stage, water sprinkler or water mist

system and the smoke management system. Besides the active fire measures, passive

fire measures are also needed. The passive measures include the study on fire resistance

of building materials, compartmentation, refuge floors and fire exits and escape routes,

etc. These passive measures are to improve or increase the ability o f the structure and

35

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 building construction or finishing to withstand the effects o f fire. From the point of

view o f maintaining the fire safety in the hotel building, the combined successful

operation o f both the active and passive fire measures is necessary. With the operation

o f these measures, the protection o f the structure against the fire outbreak can be

achieved and at the same time evacuate the guests and hotel employees to emergency

escape routes.

However, from the fire prevention and protection perspectives, these active and passive

measures are complimenting each other and function as a whole and complete technical

infrastructure. Hence, establishing the total fire safety concept [Shields and Silcock

1987] helps to theoretically and philosophically unite fire safety with both these active

and passive fire measures together. Total fire safety concept is used to guide and

educate hotel employees and guests on the correct behavior during outbreak o f a fire.

This will improve the quality for all related personnel participating in the building

management and the fire safety scheme in the hotel and will strengthen the fire safety

management system in the hotel. In other words, total fire safety concept is used to

formulate the software component for the hotel building fire safety scheme so as to

control the hardware and technical components, i.e. the active and passive fire measures.

Moreover, establishing the total fire safety concept will on one hand help to raise the

design level to another prospective while fully understand the complete picture o f fire

safety. On the other hand, the concept not just improves the fire management system

but also strengthens and reinforces the quality o f all management personnel involved.

By properly implementing the total fire safety concept, the technical components and

functions o f both the active and passive fire measures are united together. Meanwhile,

36

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 the scientific management for these active and passive systems can always be kept in the

proper operational mode.

3.3 Development o f Total Fire Safety Concept

Total fire safety concept has been gradually developed and implemented over the past 30

years. Although there are significant achievements and progress in fire science and fire

safety engineering [Fire Safety Symposium 1989] with plenty o f experimental data and

model tools being set up, more research works are needed so that all the fire engineering

designs are based on fundamental fire engineering principles and in a scientific manner.

With total fire safety concept in mind, fire safety design, installation and management

can be dealt with as a whole. The concept can be applied to any building types with all

the fire engineering design and measures [Fire Safety Engineering 1989] being

interactive and harmonized together. The total fire safety concept actually considers the

influence between various fire safety measures and also formulates a proper fire safety

strategy for the hotel building environment.

There are quite a number o f countries researching or practicing the implementation of

engineering performance-based fire design. They do not just rely on traditional

prescriptive building and fire codes but also establish functional terms in their

requirements o f the building regulation. The UK has been doing this since 1985 but

there is no guidance or recommended solution on how these alternative solutions to fire

safety design can be assessed or established. How various fire safety measures interact

with one and other still cannot be known. Detailed in the NFPA 550 standard [NFPA

37

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 1995], the first systematic development on total fire safety concept and the adoption of

the fire safety concepts tree can be seen. This concept tree outlines the relationship

between fire protection, fire precaution and fire prevention measures. Documented

development on the total fire safety concept is from Australia, led by Vanghan Beck.

They focused on their total fire safety concept in the following six areas [Eaton 1991]:

• Nature o f occupancy

• Occupant avoidance

• Active fire measures

• Passive fire measures

• Fire growth and development

• Fire fighting strategy

The interaction o f these areas later evolved into the first edition o f fire engineering

guidelines [SFPE 2002; BSI 1994, 1997] which later became the subsequent BS7974

[BSI 2001].

3.4 Application o f Total Fire Safety Concept in Fire Safety Design

The fire safety design can be considered in two different stages. One is the early fire

development stage and the other is the post-flashover stage.

For the early fire development phase, smouldering fires will be managed [Ohlemiller

2002] which may not be disastrous to occupants or cause building damage. The critical

38

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 factor for this stage is to protect the occupants and the building itself against the fire

effect. The occupants are needed to be alerted by public announcement or fire alarm

system [Shields and Proulx 2000] to bring them safely to the exit or escape route.

Additional fire safety measures such as emergency lighting or smoke extraction system

may be needed in case occupant evacuation is initiated. The fire safety strategy in this

stage requires the interaction between the occupants, fire alarm system, escape route and

possibly the smoke management system [Alpert 2002; Kung et al 1991; Klote and Milke

2002].

For the post-flashover stage [Bryan 2002], the passive fire measure will possibly be

subjected or challenged by the heat from the fully developed fire. The fire developed

fully as there may be a delay in informing both the in-house fire fighting team or local

fire brigade and that the water sprinkler system may be non-existent or inoperative. The

primary purpose at this stage is to prevent fire spread to other parts o f the hotel, causing

further hotel damage and human life loss. The fire safety strategy basically concerns

three points, namely, the building characteristics and material properties, building usage

or fire load and if any internal or ventilation openings are involved.

A number o f parameters [Schifiliti et al 2002] must be considered concerning the above

two stages o f fire. Firstly, “How big is the fire?” has to be considered. Design fire is

important and specified by the relationship between the release rate and time, and with

heat release rate being viewed as the most important parameter in any fire hazard

assessment. Hence, full-scale binning tests are critical as there are basically insufficient

data on how various combustible materials behave and bum under different ventilated

39

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 environment. Certain full-scale burning tests in a sample kitchen will be outlined and

reported in later sections.

The behavior o f the hotel occupants [Proulx 2002] toward a fire alarm, the time required

for them to recognize the alarm or take action or to arrive at the fire exit [Nelson and

Mowrer 2002] are both important and crucial to human lives. This is a psychological or

sociological issue but we still lack exact and complete knowledge [Hall 2000; Zukoski

1994] on the above problems and the time required.

Fires have to be reported to the local fire brigade so that appropriate action and rescue

can be arranged without any delay. There are particular design data or model developed

so far for fire brigade operation. In Hong Kong, the operation and response time from

the local fire brigade always receive good comments. However, there are still rooms for

improvements such as the response to fire at ultra-high hotel buildings, arson fires,

terrorist attacks and chemical fires.

Numerous tests have been carried out by various international bodies on the building

materials, structure and finishing, such as BSI, ASME, ISO and NFPA. However,

whether these tests and rating reflect the actual behavior o f the building material in a fire

is still questionable. Hence, more sophisticated and refined material tests on fire should

be conducted to reflect the actual material behavior under a real fire, whether it is a t2 fire

or a quick and severe fire [CIBSE 1995] that has major disastrous effects on the fire

resistance o f a door [England et al. 2000].

40

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 3.5 Review on Passive Fire Protection Systems in Hotels

Passive fire measures form a major part o f the hotel construction and structure against the

effect o f fire. These passive measures are also a significant and major section in the

standard licensing requirements on building safety conditions issued by the Home Affairs

Department for purposely built hotels. In this section, the aims o f these passive fire

measures in a hotel; their major components; ideas from the various codes [BD 1995,

1996a, 1996b] such as the local codes on means o f escape (MoE), fire resisting

construction (FRC), and means o f access for fire brigade (MoA) will be reviewed. Life

safety code NFPA-101 and standard licensing requirements from the Home Affairs

Department for the issue o f operating license to local hotels will be discussed. How

these passive fire measures integrate with the previously discussed ‘Total Fire Safety’

concept will be outlined.

3.5.1 Aims to be achieved by passive fire protection systems

In the past decades, modem hotels have adopted new architectural designs and concepts

in response to changing guests’ needs and hotel corporate image. For hotel industry in

the 1980s, nearly all hotels were built simply with simple concrete structure, connected

with floors o f guestrooms and facilities. With the incorporation o f advanced electrical

and mechanical systems into the hotel infrastructure, more and more complex hotel

configurations have emerged. The use o f more daylighting design, ventilation and

air-conditioning system and better fire protection system all contribute to the appearance

o f new architectural features, such as the use o f large atrium spaces in the hotel lobby,

open corridor design for resort hotels with all part o f the hotel being connected

41

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 horizontally and vertically. All these modem hotel buildings are symbols o f economic

growth o f the country, like China in the past 20 years. However, it seems that the

current building or fire codes, based on the existing prescriptive mles, are not so well

organized or arranged to confront the challenges o f all these new architectural designs.

Besides those new architectural features, a lot o f new finishing materials are used and the

characteristics o f these materials under fire are sometimes unknown. All these new

architectural features, building complexity that may involve groups o f buildings, ultra-tall

hotels and advanced materials used cause problems on design for new hotels, in particular

on fire and smoke hazard concern. These new features and materials are concerns to

fire prevention and also fire fighting for the local fire brigade.

The main aim o f the passive fire measures [Tsui and Cheung 2004], o f course, is to

maintain the safety o f the occupants inside the building and also the life safety of the fire

fighting or emergency teams. Based on this main aim, the measures must lead or

educate the hotel building owner or management the necessary fire protection and

prevention in the building itself and its surrounding buildings and also to protect the

assets within the hotel. Passive fire measures must comply with the local code on

building and fire safety. Also, in many cases for international chain hotels or

multinational building owners, they must satisfy the insurance requirements laid down by

these major management and multinational companies or owners. Not just international

chain hotels but also local hotels need to maintain the hotel or their corporate image and

most importantly, to maintain the hotel business and minimize the financial loss in case

o f a fire. That is why the passive fire measures play a significant part in the hotel on fire

strategy and indeed is an environmental protection concern also. In some hotel

buildings that are converted from heritage or historic buildings in China, Singapore or

42

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Hong Kong, the passive measures must be seriously considered because it is both our

obligation and moral to conserve these buildings.

To summarize, passive fire measures aim to provide fire prevention, building protection

and human life safety for the hotel building. The local building codes or standard hotel

licensing conditions have actually laid down the minimum requirements for the building

itself in satisfying the safety and health needs anticipated by the society. Those fire

measures represent the expectation o f the community toward fire safety required in the

hotel buildings.

3.5.2 Essential parts o f the passive fire protection systems

The first essential part is the fire resistance in the passive fire protection systems. The

fire resistance o f the structure, finishing or materials will limit fire spread and avoid

structural collapse in case o f fire. The requirements are laid down in the code of

practice on Fire Resisting Construction, 1996, and a major part in the whole fire safety

strategy. For the initial fire stage or pre-flashover stage, it is important to lay down

proper fire resistance for the building materials or finishing so as to restrict flame spread

at this fire developing period. In the following stage or when the fire has developed to a

post-flashover fire, the fire resistance in the structure or finishing basically resists the fire

spread and avoids total building or structural collapse. The general building elements

involved are the fire partition walls, floor slabs, smoke and intuminent seals, fire rating

doors and glazing, staircases, emergency signage such as exit signs. During each hotel

licensing inspection by Buildings Department, the building surveyor will inspect most of

43

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 the sealing or fire stopping around the ducts or building services installation through the

walls or slabs to ensure that they are properly sealed without any gaps or openings.

Passive fire protection systems [White 2002], in particular the fire resisting construction,

are important and cannot be traded off by other active fire systems provided. For

modem hotels with so many floors, duct spaces and connected air-conditioning systems,

it is critical to seal all the gaps and holes at all fire rated partitions so as to prevent and

limit fire or smoke spread to other parts o f the hotel building. Hence, good

housekeeping practices have to be implemented by relevant hotel engineering and

maintenance department to ensure that scheduled inspection, repair and maintenance

have been carried out to maintain the fire resistance in the building structure, partitions

and finishing.

The other essential element o f the passive fire protection systems is the escape routes or

exits. The requirements are laid down in the local code o f practice on means of escape

in case o f fire, 1996. Advanced developments and designs in local hotels bring in more

fire safety considerations and problems, particularly on egress and exit aspects. As

compared with old and traditional buildings, some o f these new hotels are unique as

several fire safety issues are needed to consider at the same time. These include the

building height for ultra-tall buildings, change of functional use inside the hotel and

possibly the increase in hotel occupancy loading as many o f them adopt the service

apartment concept. High occupancy loading in the hotel implies a potentially larger

number o f causalities and damage to the building in case o f a fire occurs. Large group

o f people are anticipated to search or run to the escape route, resulting in possible longer

escape time due to queuing and chaos and pre-movement time issue. The hotel

44

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 management needs to accommodate this issue carefully by revising all their safety

measures, especially on the fire safety management implementation. Furthermore, upon

the various advanced and complicated layouts adopted in modem hotel building design,

the occupant age has to be considered when planning the whole fire evacuation strategy.

Children and kids are considered as a high risk group that needs more care and help

during fire evacuation.

The aim o f the escape route is to enable the occupants to reach a safe place via an escape

path which is free from smoke and fire [Butcher and Parnell 1979]. This principle must

be accomplished during the whole evacuation stage so as to minimize death and injury o f

occupants. For either prescriptive codes or engineering performance-based fire codes,

serious considerations on the capacity o f the escape routes are requested so that the

maximum occupant density in the hotel will not be exceeded.

Good housekeeping and maintenance practice should be implemented and supervised by

senior hotel management. Often happening in hotels is that obstructions and storages

are placed at lift lobbies, staircases or even the escape routes. These are totally

unacceptable and senior hotel management must instruct that all the exit routes are to be

kept free from any obstructions at all times and all exit doors are maintained operable

from inside without using a key. No storage of dangerous goods along the escape route

or lobby is allowed and all the fire doors should not be wedged with wooden blocks.

The third essential element of the passive fire protection systems is the access for

fire-fighters or emergency personnel. The requirements are laid down locally in the

code o f practice on the provision of means o f access on fire fighting and rescue, 1995.

45

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 This is sometimes a problem in Hong Kong and it can be seen from the media concerning

the difficulties in some fire cases for fire-fighters to access the building. Serious

considerations must be followed up on the design o f these access routes for the fire

brigade so that they can access the building which is on fire. Misunderstanding on the

various requirements between various statutory departments must be sorted out at the

design stage so that sufficient space is allowed for the fire-fighters to approval and

extinguish any fire incident.

On the fire safety management or the software component practiced by the hotel

management team, it will be more important whether there is a team o f trained hotel staff

who can lead the fire brigade or fire-fighters to the fireman’s lift or access route. The

fireman’s lift must be well maintained so that it can be used in case o f fire or other

emergency situations such as electrical power suspension. The hotel team must inform

the firemen exactly the location o f fire and if there are any guests or employees being

trapped at the fire location. Good housekeeping practice is o f paramount importance at

this stage as all the fireman’s lift lobbies or access routes must be free from any

obstructions or blockage. Time is very important for any rescue operation and the hotel

management team must coordinate fully and assist the fire brigade team in any hotel fire

incident. The correct sitting o f the smoke vents at the hotel exterior, the fire alarm

panel, fire sprinkler pump room, street hydrant and hosereel pump room should all be

clearly shown to the fire brigade so as to organize the rescue.

46

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 3.5.3 Fire resisting construction - main characteristics and requirements

Hotel [CNTA 2005] should be separated from any adjoining buildings by an external

wall having a fire resisting period of not less than two hours. Also, no openings should

be made in such external walls that are within a distance o f 900 mm o f any part o f any

building on the same site or within 450 mm o f the boundary with an adjoining site.

Openings may however be made on external walls within a distance o f 1.8 m o f any part

o f any building on the same site or within 900 mm o f the boundary with an adjoining site

provided that these openings are protected by fixed lights with fire resisting glazing.

Elements o f construction [Fitzgerald 1997] in any hotel basement and the separation

between the basement and any hotel adjoining storey should have a fire resisting period

of not less than 4 hours. Where different occupancies are happening within the same

use o f a hotel, separation should be made between them by walls and floors capable of

resisting the action o f fire for a period o f not less than that required for the elements of

construction o f the compartment in which it is situated, subject to a maximum o f 2 hours.

The local FRC code needs to concern the fire resistance o f any element o f construction

e.g. compartment wall, floor, roof, fire separation at internal and external openings and

adjoining buildings and also fire separation at any bridges or tunnels connected. Hence,

approval o f the fire resisting construction table, methodology or smoke vent disposition,

adjoining building protection and fire damper or smoke lobbies at bridges or tunnels

connected are necessary.

A hotel should be divided into compartments by walls and floors such that no

compartment exceeds 28,000 m3. Compartment walls, floors, separations and lobbies

47

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 should be constructed with all joints completely filled with non-combustible material to

prevent the passage o f smoke or flame. Any opening in a compartment wall or floor for

the passage o f electrical and mechanical service installation or holes left after the hotel

construction should be protected with fire dampers or other suitable forms o f fire stop to

maintain the required fire resisting period of the compartments. Where ducts, pipes,

wires and any insulation passing through the wall are o f combustible material, such

material should be contained within an enclosure having an FRP corresponding to that o f

the surrounding structure [Fleischmann and Buchanan 2002]. Where access openings

are provided to the enclosure, such openings should be provided with self-closing doors

having an FRP o f not less than half that o f the structure.

At any internal unprotected opening such as at escalators and circulation staircases, a

barrier o f not less than 450 mm measured vertically downwards from the underside o f the

floor should be provided to surround the opening. The barrier should be constructed of

material having an FRP o f not less than 1 hour. The barrier shall extend not less than

450 mm below any false ceiling hung in the vicinity o f the opening.

For special hazard areas like kitchens, special requirements are necessary. They should

be enclosed by non-combustible construction having an FRP o f not less than 1 hour and

openings should be provided with doors having an FRP o f not less than half an hour.

Kitchens should also be provided with protected lobbies between each door and any

escape route from the main hotel building portion.

All fire resisting doors having an FRP should be arranged to be self-closing with notices

for them to be always kept closed. All such doors should be closely fitted around their

48

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 edges to impede the passage of smoke or flame. Doors including frames should be

tested in accordance with BS 476: Part 20 and 22 : 1987 and certified as being capable of

resisting the action o f fire for the specified period.

A bridge, like the one at Wanchai Century Novotel Hotel, uniting a hotel to an adjoining

office building should be provided with a fire shutter having an FRP o f not less than 2

hours and no openings should be provided in the hotel within a distance o f 900 mm from

the junction o f the bridge and the walls o f the hotel within this distance should be o f

non-combustible construction having an FRP o f not less than 2 hours. Underground

tunnels uniting a hotel (like the one at Kowloon Hotel) to an adjoining building or facility

should be provided with a fire shutter having an FRP o f not less than 4 hours.

It is the basic aim o f FRC in any design to restrict the effect from fire, through

conduction, radiation or convection. Usually, a two-layer model is formed in a

compartment fire. A fire plume is gradually formed until it reaches the ceiling where it

will be deflected to form a horizontal ceiling jet. If the compartment wall or floor is not

o f the specified fire resisting period, the fire will spread to other parts o f the hotel.

However, depending on the fire size and compartment, in some cases it may be feasible

for the fire plume to be too large that it will not be contained in the room compartment.

Under the fire safety design concept, FRC is to limit the spread o f fire and the collapse of

the hotel building [Milke 2002] so as to reduce the damaging effects o f smoke, heat and

flames on the hotel guests and employees. The design has to cater for the fire separation

within the hotel building and between adjoining buildings and the structural integrity o f

the hotel.

49

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 3.5.4 Means o f escape - main characteristics and requirements

Every exit route shall lead directly to a street or to an open area having unobstructed

access. Such access to a street shall not be closed with doors or gates unless such doors

or gates are fitted with panic bolt. The enclosing walls o f every staircase shall be so

continued at ground floor level as to separate from the remainder o f the hotel any passage

or corridor leading from the stair to any ground level exit doorway to which the stair

gives access. Every hotel shall be so constructed that there are available from each

storey not less than two exit routes.

The maximum travel distances that will be permitted from any part o f hotel shall be 36 m

o f which not more than 24 m may be along a corridor or not more than 30 m may be

along a balcony approach. If the exit route is in one direction only, the maximum travel

distance shall not be greater than twice the length o f the exit route between the entrances

to the enclosures o f the required staircases. Every hotel should be so constructed that

there are available from each storey not less than two exit routes.

Every required stairway in a building shall be separated from the hotel by a wall having

an FRP o f not less than half an hour. Any opening in the wall separating a staircase

from the hotel shall be protected by a self-closing door having an FRP o f not less than

half an hour provided that no such door shall be required between a balcony approach

and any stairway leading from there.

The exit door o f any hotel or storey with direct access to a protected stair shall be

self-closing with an FRP o f not less than half an hour. Every internal corridor giving

50

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 access to rooms shall be enclosed by partitions having an FRP o f not less than half an

hour; the doors to rooms from the corridor shall be self-closing with an FRP o f not less

than half an hour. A lobby between the internal corridor and the staircase will not

normally be required in a hotel in which the highest storey is not more than 30 m above

ground level. The exit route from any room or storey to any part o f a staircase which

serves a storey more than 30 m above the level o f the ground shall be through a lobby.

Such lobby shall be designed as an integral part of the staircase so that it could not be

readily incorporated as part o f the accommodation, and shall be a protected lobby. The

means o f escape from any part of the hotel shall be so arranged that it is not necessary to

pass through one staircase enclosure in order to reach an alternative stair.

When a fire occurs, hotel occupants are not expected to take a very long distance to reach

the safe exit [Joyeux 2002] and the ground level is often considered as the most

appropriate place o f safety. Hence, alternative routes must be provided for immediate

access to the place o f safety in order to avoid occupants being trapped at the fire scene.

Under the MoE design, a structural means has to be provided with safe routing o f travel

from any point o f the building to the place o f safety. The design has to cater for both

the horizontal and vertical evacuation of occupants via the staircase, lobby and exits

provided.

The MoE code concerns critically the exit width, number o f exits from room or

individual stores, maximum distance travel, staircase capacity loading, refuge floor,

lighting and access to staircase and finally the exit to the place o f safety usually at ground

level.

51

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 3.5.5 Means of access for fire fighting and rescue - characteristics and requirements

The basic aim o f the code for means o f access is to provide fire-fighters the necessary

safe and unobstructed access to the fire scene to perform their rescue and fire

extinguishment operation.

At least one fireman’s lift [Buckley and Bradbom 2000] has to be provided. Where

more than one lift is installed in a hotel, the lift which is to be the fireman’s lift shall be

designated by the local fire brigade. Fireman’s lift shall be provided to enable fire

services personnel to reach any floor that may be on fire in the hotel without having to

traverse more than two floors. Separate liftwell must be provided for the fireman’s lift

and up to three lifts may share the same well provided all these lifts are designated as

fireman’s lift. A notice shall be displayed outside the liftwell indicating which one will

be the fireman’s lift. The lift car shall be of a minimum size o f 1.35 m2 net floor area

with a rated minimum loading of 680 kg.

Time is essential in any fire fighting operation [Klaene and Sanders 2003] in order to

save lives. The emergency vehicles from the fire brigade should be able to reach the

fire scene without any obstruction and in a reasonable time under a safe environment.

Hence, the code o f means o f access for fire-fighters lays down requirements that the

building should have emergency vehicular access so that fire appliances can reach the fire

site safely.

Internal means o f access would be more efficient for the fire-fighters although current

advanced equipment like aerial devices may provide external access to individual fire

52

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 incident floor o f the hotel. To protect fire-fighters from the exposure to heat, flame and

smoke from the fire, it is necessary to have protected routes inside the hotel building so

as to minimize their travel distance in the fire environment. These protected routes also

serve as the refuge places for the rescue personnel.

From the design point o f view, means of access provide the safe route o f access from the

hotel exterior to any point o f the building. It is important to concern and plan the

emergency vehicular access from the hotel exterior to the building. The vertical access,

via fireman’s lift or protected vertical shaft, must be planned for fire-fighters to carry out

rescue operations. For horizontal access at each floor, protected lobby should be

established for travel from the vertical access.

In brief, the code for means o f access for fire-fighters and rescue outlines the

requirements for emergency vehicular access, the access at ground level and the

provision and requirements for firemen’s lifts. It also lists out the requirements for the

number o f access staircases and the provisions for rescue staircases.

3.5.6 Passive fire measure - alternative fire engineering approach

Nowadays, many countries are practicing the performance-based fire safety design

[Custer and Meacham 1997]. Fire safety goals and design objectives have to be

identified in performance-based design. However, the advantages are that there is

flexibility to have a range o f technical solutions for the same design problem or issue.

By doing so, it will help to bring in innovations to the design, new materials, advanced

technology which assist in making the whole design process more efficient. This is

53

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 particularly obvious for buildings with special features like hotels or with special hazards

due to their height, proposed use, site location and construction which specially may need

particular fire safety objectives and standards to be established. It is laid down in the

building code that “the Building Authority recognizes that fire safety may be approached

in a number o f ways, the best o f which is not necessarily prescriptive”.

The code also mentions that “the Building Authority assesses the acceptability o f any

alternative or complimentary approach to fire safety in a building by reference to such

criteria on the means o f escape, the means o f access, the fire services installation, the fire

resisting construction, the size, the height, the use, location and the management o f the

building”. Hence, besides the passive fire protection systems, the code also brings in

components like fire safety management and active fire protection systems.

As an alternative to prescriptive codes for the fire safety provision, currently the fire

safety engineering approach [Tubbs 1999] is also adopted. The approach includes basic

engineering principles, models and scientific methodology in order to provide a design

solution and framework to achieve the performance requirements and objectives as

outlined in the building laws. In brief, the fire safety engineering approach formulates

the performance-based codes and requirements which will be a flexible approach to reach

a more efficient and cost-effective design.

3.5.7 Passive fire protection systems —integration with total fire safety concept

For the fire safety engineering approach [Chow 1999], it allows the designer to determine

their own methodology to achieve the specified goals and objectives which will be in

54

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 very clear and precise terms. However, there may be disadvantages in adopting the

performance-based design. First of all, the local building authority, which has the

obligation to approve any building design, may be reluctant to approve design as they

may not be familiar with the goals, objectives and methodology used. Moreover, the

engineering approach may be more academic as it needs the application o f engineering

principles and scientific methodology [Hadjisophocleous and Tamim 1998] which most

designers or government officials may not get acquainted. This concerns the quality and

education o f the designer and reviewer and also the necessary quality control measures to

assure the adequacy o f the design by engineering approach.

It is understood that once the occupancy or building use is changed, the fire protection

needs will be revised in prescriptive codes. That is why the establishment o f the key

assumptions, goals and objectives are so important in performance-based fire safety

design. It also states the importance o f the components o f fire safety management in the

whole building design and operation. The subsequent inspection, testing, repair and

maintenance o f all the fire protection systems provided (whether it is active or passive)

are important to upkeep the goals and objectives originally set in the design stage and

assumptions.

In the past, MoE, MoA, FRC, FSI and FSM will be considered separately in prescriptive

fire safety design. Now, a more comprehensive fire safety strategy can be focused on so

that the software fire safety management (FSM component) can be used to control the

building hardware provisions (MoE, MoA, FSI and FRC component). The design goals

and objectives originally set and planned will not be forgotten even after the issue of the

55

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 necessary occupation permit if a fire safety management scheme is implemented [SFPE

2001].

56

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 4 ACTIVE FIRE PROTECTION SYSTEMS

4.1 Introduction

Before an effective and cost conscious fire safety strategy is finalized, a clear picture on

the fire safety objectives is needed so as to satisfy the client’s need and project

specification [Chow and Wong 1998]. These fire safety objectives must be related to

human life safety, protection o f asset and environment and the conservation o f cultural

heritage.

Fire safety measures must be formulated, whether passive or active, communication

systems or egress systems, in the design o f a building. However, as related in the

previous chapter on total fire safety concept, the synergistic effect and integrating

together o f all these measures and systems will make a building safe. These integrated

measures and systems include, on the passive fire protection side, the proper fire rating or

resistance o f the building constructions and fire stops at the floor and walls penetrations.

On the active fire protection side, they include the fire suppression system, smoke

management system, fire detection system and the egress system.

The active fire protection systems [FSD 1998; CNTA 2005] are designed and installed to

meet the requirements in the building and fire code so as to function as designed in case

o f fire. The active fire protection systems, usually powered by electricity, standby

power or UPS, are operated under automatic devices action to control and suppress the

action and effects o f fire.

57

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 In recent years o f research and developments [Chin 2002], a wide range o f active fire

protection systems are available for use in actual fire fighting operations. The first

active system in fire safety operation is the detection devices which will sound the alarm

to inform the occupants on the possible fire outbreak. These include the smoke or heat

detectors, breakglass points, and public announcement emergency system. Then, the

fire suppression systems which include the fire hydrants, hosereel, portable extinguisher,

sprinkler or water mist systems and possibly chemical systems will operate to control the

fire and limit the spread o f the fire and smoke to other parts o f the hotel. Finally, the

emergency illuminating system is to provide the necessary luminance for evacuation of

the hotel occupants to the place o f assembly.

The operation o f active fire protection system is believed to raise the alarm or concern of

the occupancy to any possible fire outbreak or growth, limit the smoke spread, suppress

the fire and control or restrict the fire or smoke spread to within or outside the fire

compartment, make possible the fire-fighters’ operation to happen, avoid building

collapse, minimize life and asset loss and to permit most rapid egress to the safe place.

However, in real situations and due to environment or other restraints, some o f these

active fire protection systems do not perform in accordance with their design

expectations. More intensive and detailed research must be carried out to provide more

valuable data. These supportive data will avoid initial design assumptions or judgments

being drawn up blindly during the design stage when the suitable systems are being

proposed for the particular hotel building. In this chapter, local and international fire

codes for hotel concerning active fire protection systems will be discussed. The

operation and characteristics o f active systems will be outlined, such as fire detection,

sprinkler, water mist, hydrant and chemical systems.

58

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 4.2 Local and International Fire Codes for Active Fire Fighting Systems

In this section, the requirements for active fire fighting systems from various local and

international fire codes will be discussed. Locally, for the application o f hotel operation

license, the fire safety requirements as listed in the Hotel and Guesthouse Ordinance

[CNTA 2005] have to be followed. Basically, the requirements laid down in the “Code

o f practice for minimum fire service installations and equipment and inspection and

testing o f installation and equipment” have to be complied with. In China, the

requirements in the China fire code issued by the State Council and the Ministry of

Public Security and also the fire code for high-rise buildings are referred to. In the US,

there is a particular section in the NFPA life safety code-101 [NFPA 1994, 1995] for

hotels and guesthouses.

From the local code o f practice, audio or visual alarm system should be provided for any

part o f the hotel where the area occupied by the occupancy on any one floor exceeds

2000 m2 and the hotel occupants, due to their risk exposure by transient presence either as

visitors or shoppers, will require such system or advice. Such system will include the

alarm bell system, visual advisory system and public announcement emergency system,

etc.

If water supply [Cheung and Chan 2003] is undesirable for the fire suppression system in

the hotel, then automatic fixed installations other than water shall be provided. The

automatic system or devices should be supplied also with essential power and an

independently powered electrical generator shall be installed to provide the necessary

electrical capacity for the essential services in the hotel. Portable fire extinguishers, fire

59

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 blankets or sand buckets are provided whenever necessary in accordance with the code,

for various locations in the hotel. The locations are primarily the pantry, switch rooms,

plant rooms and also restaurants and kitchens as per the local restaurant licensing

requirements from the Food and Environmental Hygiene Department. Usually a 2 kg

dry powder or a 4.5 kg C 0 2 gas fire extinguisher will be provided for the pantry or switch

room location.

It is important to detect the outbreak o f fire and alert the hotel occupants in the shortest

duration possible. Hence, it is required to install an automatic smoke detection system,

in accordance with the Fire Offices’ Committee for automatic fire alarm installation 12th

edition, for the entire floor if the guest floor is used for sleeping accommodation. On

current changing use o f the guestrooms to other similar function, like service apartments

for long-staying guests, heat detector at some locations inside the guestrooms may be

accepted to avoid false alarms. The alarm o f such system shall be transmitted to the fire

services emergency centre by direct telephone line, which is also linked up with the hotel

fire alarm system. Regular testing, inspection, maintenance and repair should be carried

out for the automatic fire alarm system as this is the first line o f defence o f the building

against fire. The main automatic fire alarm system control [Tamura 1994] is usually

installed in the fire control centre of the hotel. This control centre is located at the

ground floor o f the building for easy access and checks by the local fire brigade.

In case a fire is detected by the occupants, such as the kitchen chef or in-house guests,

they can alert the whole hotel by sounding the manual fire alarm system. This system is

installed with one actuating point (e.g. breakglass unit) and one audio warning device at

each hosereel point, usually near the exits. It is in the fire code requirement that the

60

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 actuating point shall initiate a series of automatic devices action for the fire pump to start

and audio warning device actuation.

For fire suppression systems, a number o f related systems have to be provided to control

and limit fire and smoke spread. A fire hydrant and hose reel system should be installed

to ensure that every part o f the hotel is reached by a length o f not more than 30 m o f the

fire services hose or hose reel tubing. Daily service, checking and repair o f the system

are necessary to reconfirm that all these hydrants and hose reels provided are in operating

manner and working properly. With past experience on site or upon daily hotel

operations, this simple housekeeping and maintenance procedure is usually neglected or

treated in a very poor manner.

Another fire suppression system that should be provided is the water sprinkler system.

Automatic sprinkler system shall be installed for the entire hotel building including

staircases, common corridors in accordance with the Loss Prevention Council Rules for

automatic sprinkler installations. Except for areas with special risk, such as in kitchen

with cooking woks, computer rooms or switch rooms, sprinkler system is proved to be a

very effective system in limiting fire spread or fire extinguishment in a hotel. Several

past fire incidents happened in local hotels and the fires were basically extinguished by

the operation o f fire sprinkler before the action o f the in-house fire team and local fire

brigade.

It is important that all exits should be indicated by illuminated exit signs bearing the word

“EXIT” in block letters and characters o f not less than 125 mm high with 15 mm wide

strokes [NFPA 2005]. Color contrast for translucent surrounds to lettering shall be

61

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 complied. If an exit sign is not visible in any location in the hotel, say due to turns at

corridor or blockage by other architectural features, suitable directional signs conforming

to BS5499-4 [BSI 2000] shall be erected with an exit sign and shall be provided at

conspicuous locations to assist hotel occupants to identify the exit routes in case o f an

emergency. Emergency lighting should be installed in the whole hotel and all exit

routes leading to ground level. A self-contained secondary lighting system will be

accepted if the illumination level o f not less than 2 lux for duration o f 3 hours in the event

o f power failure is provided.

Staircase pressurization or static/dynamic smoke extraction system shall be installed with

the local code o f practice for minimum fire service installations and equipment.

Whenever ventilation or air-conditioning control system is installed in a hotel, it should

stop mechanically induced air movement within a fire compartment. All ventilation

systems should comply with the building (ventilation system) regulation and a letter of

compliance should be obtained from the Fire Services Department.

4.3 Automatic Fire Detection System

For fire protection and safety, time is the most important parameter. The alarm response

time, evacuation time and suppression time should be reduced. The prime objective of

any fire detection and alarm system is to reduce human life loss from fire and secondly to

protect the asset property. An advanced and early warning performance o f the detection

system is expected, particularly when designing and installing the system in accordance

62

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 with performance-based codes. In the following sections, various fire detection systems

and their applications will be outlined.

4.3.1 Introduction

Fire alarm systems are the first defence and feature among the protection elements of the

hotel building. If the systems are properly designed, installed and maintained, the

systems can surely reduce life loss and property damage due to fire. By adhering to the

requirements specified in the local code o f practice for minimum fire services

installation, NFPA 72 national fire alarm code or FOC guidelines, a qualified fire

detection and alarm system can be ensured.

Several characteristics o f the hotel building have to be finalized before selecting the

appropriate fire detection system. A holistic approach should be taken that the effects of

the combination o f all the fire protection features must be greater than individual

operational effects o f each component. Therefore, effective protection level must be a

combination, interlocking and interrelated functioning o f each fire protection system.

Hence, designers or owners for the hotel building should not just provide a fire alarm

system or sprinkler system but they also have to think o f all other passive fire protection

features or other valuable protection systems as a whole in order to achieve the safety of

the hotel building.

63

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 4.3.2 Fire alarm system - components

A normal fire alarm system may provide several signals. The alarm signal, whether it is

true or false, should be attended immediately in a hotel [NFPA 2005]. The trouble

signal, which indicates a failure or fault in the alarm circuit, should be attended

immediately. The other critical signal is the supervisory signal, which indicates the

activation o f the fire protection systems that are connected to the fire alarm system, e.g.

sprinkler system, dry chemical and gaseous system, etc. The components in a fire alarm

system can be classified as the central system control unit, the main or secondary

electrical supply, and the initiation device circuits such as the detectors, the visible or

audible devices. In many o f the local and foreign applications, there must be a direct

link connection to the local fire brigade where immediate response to any fire alarm can

be activated. The secondary power system has to be properly selected for the fire alarm

system. The secondary power system must be operating to supply energy to the fire

alarm system within 15 s and followed by 3 hours o f emergency operation. The main

and secondary power must be properly maintained and tested to ensure that the fire alarm

system will function during an emergency.

4.3.3 Automatic fire detectors

In any occurrence o f fire, certain environmental changes are involved like heat, smoke

and radiant energy. O f course, human beings are the best fire detectors but their

detection can be distracted by many other factors that make these early detections

ineffective. Several space and fire characteristics should be considered before any

detector selection. Firstly, different types of fires may have different environmental

64

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 emissions, for example, some fire may have no smoke or even flameless. Secondly, the

environmental emission being detected must reach the detectors and exceed a certain

minimum threshold for the detector activation. Thirdly, the environmental emission

from other non-fire situation that may lead to false alarm, for example, boiling water that

activates the detector in the kitchen or guestrooms. Hence, all the fire codes will

indicate the requirements leading to the proper selection and location o f these fire

detectors in order to accomplish the fire protection objectives. Fire alarm system

designers need to consider the amount o f environment emission from any potential

effects on the proposed detector options. Finally, the designer has to select the

automatic fire detector that can offer the fastest response time and minimum disruption

due to false alarm or detection.

4.3.3.1 Heat detectors

Fixed temperature heat detectors will initiate an alarm if the space air temperature near

the ceiling reaches the pre-determined temperature setting. However, there may be a

rapid increase in the space air temperature by a fire which the fixed temperature detector

cannot raise the alarm. Therefore, the rate o f rise detector is usually designed to

function when the rate o f temperature rise exceeds a preset value, say 7 to 8 °C per

minute. This rate o f temperature rise is normally expected under a potential fire

condition. However, care should be taken when locating and selecting the rate o f

temperature detectors in order to avoid false alarms. Usually, these detectors will not be

installed in the hotel kitchens, laundries, near heaters or air diffusers, etc. There are also

several types o f rate o f rise temperature heat detectors, namely line type and spot type.

65

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 4.3.3.2 Smoke detectors

Ionization smoke detectors normally contain a small amount o f radioactive material that

polarizes the air in the detector chamber leading to a small current flow though the air

between the two electrodes in the detector. The smoke particles, possibly due to the fire

emission, will decrease the conductivity o f the air in the sensing chamber and activate the

detector if the conductance falls to a pre-determined level.

Photoelectric smoke detectors work on two principles, either by obscuration o f the light

intensity over the light beam path or by scattering o f the light beam.

4.3.3.3 Selection o f detectors

A wrong selection o f detectors will definitely lower the detection efficiency or cause too

many pre-matured false alarms. In the choice o f detectors, the fire protection goals and

objectives have to be firstly identified. The fire protection objectives can be life safety,

property protection or simply to satisfy local fire codes. Then, general design objectives

should be formulated that will outline how these fire protection goals are satisfied.

Thirdly, the occupancy pattern should be considered, say the possible fire threat in a hotel

building and examine if any dangerous goods storage or other fire suppression system are

available. Then, the potential environmental emission from the probable fire will be

considered and the ambient situation in the space. Finally, detectors are selected based

on its sensitivity towards the ambient condition to avoid unexpected false alarms.

66

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 After selecting the detectors, the installation and the proper spacing o f the detectors

should be considered. In some cases, smoke or heat detectors are used to control other

active fire protection systems, for example the gaseous system or pre-action sprinkler

system. Comprehensive preventive maintenance and regular testing are required for all

the selected and installed automatic fire detectors. In hotels, bi-weekly tests on all these

fire detectors should be arranged to ensure that they are functioning and operating safely

in their areas.

4.3.3.4 Audio and visual devices

These devices are part o f the fire alarm detection system. They give warnings to the

hotel occupants and indicate to them the necessity to evacuate to a safe place. A fire

alarm system can send a single bit o f information by sounding an audible signal with

strobe light flashing or multiple bit o f information by a public announcement system.

Usually when provided with more detailed information, people tend to evacuate rapidly

and effectively to the safe place. The audibility and intelligibility o f the audible devices

are both important to provide a clear message for evacuation o f the occupants.

Visual devices are usually to supplement the function o f the audible devices. The local

hotel licensing requirements, NFPA 72 and life safety code NFPA 101 [NFPA 1994;

1995] all specify the necessity o f visual alarm devices in the hotel building. For visual

devices, the source intensity and also the illumination at a distance from the source have

to be considered in particular.

67

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 4.4 Automatic Sprinkler System

Automatic sprinkler system is designed and each sprinkler head will respond to the heat

from the fire and distribute water over the fire source. Sprinkler system may be

considered in several aspects, namely its thermal response, the distribution o f water,

ability to suppress the fire and its control capability. A number o f recent researches are

available [Chow and Wong 1998] on quantifying the sensitivity o f sprinkler system.

These research works had come up with some terms, like the time constant or response

time index [Chow and Ho 1990] as a measure o f the sensitivity o f the sprinkler system.

On the other hand, the distribution pattern o f most sprinkler systems is tested only for

overall coverage under specific geometric conditions. Little can be achieved from an

engineering design point o f view on the specific spray patterns from sprinkler operation.

Currently, the total amount o f water actually delivered to the fire source per unit floor

area still cannot be predicted since the spray patterns will change with the water

discharge pressure. It can be expected that the total heat absorption rate o f the sprinkler

spray will be proportional to the total water droplets surface area and the temperature

difference between the water droplet and the ceiling smoke layer. The total cooling

capacity produced by the sprinkler system will depend on the depth o f the ceiling jet and

the distance through which the water droplets travel through the smoke layer.

Sprinklers function in a number o f ways to control or suppress fire [Cheung 1995].

Firstly, the operation o f sprinkler system will be most efficient to distribute water in

producing the cooling effect by the water spray. Particularly, the radioactive heat

reflects back to the fire for sustaining combustion will be significantly reduced if fine

68

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 water droplets are produced that will generate further cooling effect. From the fire

triangle in a combustion process, oxygen is crucial to sustaining fire growth. When the

sprinkler system operates, the water droplets produced will evaporate to steam which will

be more than 1700 times in volume in comparing with the water ejected. The steam

generated will deprive the fire from the needed oxygen to sustain the fire growth. The

sprinkler system will work well in an unventilated enclosure because the fire is contained

and under the combined effect o f the above mentioned “spray cooling” and “oxygen

deprivation” effect. Moreover, in ventilated spaces, the strong updraft due to the fire

will blow away the small water droplets and hence critically reducing the sprinkler

system efficiency. Hence, designer will specify sprinkler system based on distributing a

variety o f water droplet sizes over the fire source ground and to maintain a low ceiling

temperature and control or suppress the fire.

During the outbreak o f a fire, traditionally it is expected that a number o f sprinklers

would be operating at or around the fire source. The operation o f the sprinkler above

the fire source may not be able to extinguish the fire but they can work together with

other newly activated sprinklers to cool the atmosphere and to prevent other sprinklers

outside the fire scene from operating. In particular, the newly operating sprinklers will

cool and wet the surrounding areas including the probable stored combustibles and

further assist in preventing the spread o f fire. At the burning fuel level, the water

droplets reaching the burning fuel will reduce the burning rate to such a stage that the fire

will not spread to nearby additional fuel, together also with the water droplet pre-wetting

effect on the nearby fuel. However, at the ceiling level, the sprinkler operation will

absorb the heat from the rising fire phase to prevent operation o f additional nearby

sprinklers and also avoid the building structural damage due to excessive heat from the

69

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 fire source. Usually, the area upon which sprinklers are operating will be greater than

the fire source area. Finally, the fuel in the fire source is burnt out and the fire is

extinguished. It is also believed that walls or compartments will assist the sprinkler

system by reducing the number o f sprinklers operating. As fewer sprinklers are

operating, the higher water pressure will increase the flow which in turn increases the

cooling effect o f the water droplets. Also, a thicker smoke layer will enable the droplets

to go through longer distance and hence improve also the cooling effect o f the sprinklers.

In normal situations, the water discharge from a sprinkler is restrained by a cap holding

against the orifice by a system o f lever pressing down on the cap and anchored to the

sprinkler. There are several types o f sprinklers, namely the fusible sprinkler, bulb

sprinkler and thermo sensitive elements, etc. Automatic sprinklers have several

temperature ratings. Usually, the sprinkler temperature rating can be distinguished by

viewing the color codes on the glass bulbs o f the sprinkler. Sprinkler o f higher

temperature rating [Yao 1988] will be installed where high heat release rates are

expected. This is to prevent and reduce the sprinkler operation outside the fire area.

To suit various ceiling or fitting out requirements, various sprinkler types are available,

namely pendant sprinkler, recessed sprinkler, flush-type sprinkler and concealed

sprinkler, etc.

For designing sprinkler system [Cheung 1997; LPC 1990], designer will need to satisfy

the local code o f practice for minimum fire service installation, FOC and UL

requirements and NFPA 13 fire code. In general, several sprinkler system types are

available such as the wet-pipe system, dry-pipe system, precaution system and deluge

system, etc. The design o f sprinkler system is based primarily on the fire hazard

70

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 classification o f the occupancy. The fire hazard classification may be light hazard,

ordinary hazard or extra hazard groups.

4.5 F ire H y d ran t and Hosereel System

Fire hydrant and hosereel system are fixed piping system that brings water from a reliable

water source to the hotel building where hoses can be used for fire fighting by the

firemen or occupants. These systems must be provided in tall and large area buildings.

The hydrant and hosereel system increase the fire fighting efficiency by eliminating the

need to lay hose from the fire engine to the fire site. Even when the buildings are fully

sprinkled, the hydrant system will serve as a back up and support the sprinkler operations.

Designers will design the fire hydrant system based on the requirements from the local

codes o f practice for minimum fire service installation, NFPA 14, NFPA 101 life safety

code and the FOC rule.

The common purpose o f the hydrant and hosereel system is to deliver water for manual

fire fighting. However, the system design may vary to achieve this purpose. Whereas

one system may have water connecting from the fire engine to the hose in the building,

while others may have fully automatic connected hoses with automatic water supply.

Again, like sprinkler system, the design process o f hydrant system starts with the

confirmation on intended use o f the system. The hydrant system may be designed for

full-scale fire fighting, first aid fire fighting or both. The three uses mentioned

correspond with the three classes o f hydrant system in NFPA 14 [NFPA 1993], that is

class I, class II and class III.

71

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 In accordance with NFPA 14, there are several limitations on the selection o f the hydrant

system. Manual hydrant systems should not be used in high-rise building design.

Surely, sprinkler and pump system [Cheung and Chan 1997] will be required in

accordance with the fire code and hence it is reasonable to install also pumps for the

hydrant system to achieve an increased safety level in the hotel building. Moreover,

manual systems are not approved to be used in class II and class III system. This is

because pre-connected hose system obviously must have an available water supply ready

for fire fighting. Furthermore, dry systems are not allowed in class II or class III

buildings because o f the potential risk o f untrained occupants that may delay the water

availability in fire fighting.

The required numbers o f hydrant and hose connections depend on the building layout and

design. Two methods are recommended. One is the “actual length” method in which

the hose connection is sufficient to reach all portions o f the area served with a 30 m hose

and with a nozzle reach o f 9 m. The second method is the “exit location method”. By

this method, 64 mm hose connections are located in exit stairs, as required in the building

code. As building exits are distributed within the building to provide enough egress and

hence hose connections will also be considered distributing adequately upon this

principle.

There are also several system components in the hydrant and hosereel system. The steel

pipe and fittings selected must withstand maximum system pressure from 175 to 300 psi.

The hose, hose racks, nozzles and hose cabinets should also be selected very carefully.

Hoses are usually kept on compatible racks and should always be positioned in a readily

accessible location with convenient reach o f the standing occupant. Moreover, hoses

72

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 must be clearly visible and located in places not easily to be obstructed. If the hoses are

kept in a cabinet in most o f the modem building design, the doors should have a glass

panel or other means for easy identification. The valves or pressure reducing devices

must be durable and able to withstand the maximum system operating pressure.

Throughout the whole hydrant and hosereel operations, a detailed preventive

maintenance program, inspection and testing must be implemented to ensure its proper

functioning.

4.6 Water Mist System

In the past decade, the use o f water mist fire suppression systems has become more

frequent in Hong Kong. First o f all, the general extinguishing principles o f water mist

system are examined [Knoxville 1994; NFPA 1997]. The extinguishing action o f water

mist system is believed to be due to the dilution o f the air supply at the fire scene when

the water droplets evaporate to vapor and surround the heated areas. The cooling effect

o f the water droplets is also a dominant factor. The water droplets in the water mist

system must be relatively fine and the amount o f water must be sufficient relatively to the

fire size.

The basic extinguishing mechanisms [Mawhinney and Kim 1994] involved are heat

removal, oxygen displacement, elimination o f radiant heat, vapor/air dilution and the

kinetic effects. It may not be difficult to understand the engineering principle in

extinguishing fires for the first three mentioned extinguishing mechanisms. However,

73

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 computational fluid dynamics may be involved to further investigate the vapor/air

dilution and kinetic effects extinguishing principle.

On heat removal or cooling effect, it is understood that when water mist system operates

and water droplets are applied to the fire, heat will be removed from the fire scene.

Heat is absorbed from the hot gases, the burning fuel and also from the surrounding

objects and surfaces. The cooling o f the surrounding surfaces and objects contributes to

reduce the speed o f fire spread. The use o f water mist systems and fine droplets will

increase the speed o f extracting heat from the burning fuel as the rate o f heat transfer will

be increased due to the increase o f surface area of the water mass. The conversion o f

water droplets to steam needs latent heat o f vaporization extracting from the fire. If

enough heat is extracted from the fire, the flame temperature may drop to a level below

that necessary for sustaining the combustion reaction. At the same time, the effect of

oxygen depletion will also mean the fire might be extinguished with only a portion o f the

theoretical minimum required for flame cooling.

Some researchers believed that the fire suppression effect by oxygen depletion will be

more dominant than flame cooling. Water droplets expand about 1900 times upon

vaporization. If evaporation takes place immediately, the water vapor will displace the

air in the vicinity o f the droplet. Rapid evaporation, expansion and air displacement

happened in the fire compartment by steam generating from the water droplets injecting

to the fire scene. If the oxygen level is reduced to below a certain critical level, the fire

will be burning inefficiently and easier for extinguishing by the cooling effect.

74

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 However, the dilution o f oxygen by water vapor during suppression process [Liu et al.

2001] will be limited by the average temperature o f the gases in the fire compartment.

This will explain why water mist system is more efficient in large fire extinguishment

than small fires in a fire compartment. When the fire compartment temperature is

higher, more water vapor dilutes the oxygen in the air. This also explains why turning

the water mist system on and off will reduce the extinguishing time in a given

compartment. More water will be evaporated because o f the higher compartment

temperature during the off stage o f the cycle.

Another extinguishing mechanism, the elimination o f radiant heat, reduces the spread of

fire to unignited fuel surfaces and the vaporization at the fuel surfaces. In other words,

radiant heat elimination protects the occupants and assets in the space from direct radiant

heat damage. The efficiency o f radiant heat elimination depends on the droplet diameter

and its mass density. Water mist system with higher concentration o f very fine drops

has been proved to be more efficient in reducing the radiant heat transfer. Actually, the

reduction in heat transfer to the fuel surface will reduce the rate o f generation of volatile

vapor.

The extinguishing mechanism by vapor/air mixture dilution should also be considered in

detail. The water vapor and air entrained in a water mist system operation will dilute

the vapor/air mixture to below the flammability limit. This extinguishing effect is

referred to as a secondary mechanism because it is hard to see that dilution alone can

extinguish the fire. In designing water mist system, it is still infeasible to quantity the

relationship among the flammability limits of different fuels, the fuel vaporization rates,

spray evaporation rate and finally the mass flow rates o f water mist and entrained air.

75

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The kinetic effects o f mist on burning fuel also have to be considered. To fire-lighters,

a general flare-up at the instant o f application of water spray on liquid fuel fires may not

be unfamiliar. The sudden intensification o f the fire has been commented by some

researchers to the effect o f droplets striking the fuel surface and causing increased

vaporization rate. Moreover, enclosure effects also will improve the water mist system

performance. The performance improvement may be due to restricted ventilation and

heat entrapment. Hence, it is easier to extinguish an under-ventilated fire in an

enclosure than a well-ventilated unenclosed fire.

It is understood that ‘water mist’ is a very fine water spray [Yao et al. 1999] that will

remain suspending in the air for a period o f time. Four characteristics o f water mist are

considered which will influence its efficiency as a fire suppression system. They are the

drop size distribution, flux density, spray momentum and additives needed.

The relationship between drop size distribution and extinguishing capacity [Downie and

Polyneropoulus 1995] o f water mist is complex. Moreover, drop size distribution also

does not determine the ability o f a spray to extinguish a given fire. There are other

factors such as enclosure effect, flux density and fuel properties involved in determining

whether a fire can be extinguished.

Whether a water mist system will extinguish a fire depends only partially on the drop size

and spray velocity. This surely needs the mass o f water spray that interacts with the fire

be enough to absorb a critical portion o f the heat from the fire. Spray flux density is

therefore an important characteristic o f the water mist fire suppression system.

76

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The success and failure of a fire suppression system depends on the variations in the

spray momentum [Andersson and Holmetedt 1999]. The three factors forming the spray

momentum are the spray velocity, its direction relative to the fire plume and the mass of

the water droplets transported onto the fuel surface. The more control on spray

momentum exercised, the greater will be the control on total water requirements, time to

extinguishment and overall system reliability. In addition, water mist systems with

additives, such as sodium chlorine or alkaline salt, are o f interest for application in

machinery spaces and engine compartments.

There are some practiced applications of water mist system in machinery spaces, turbine

enclosures, marine accommodation, hotels, computer rooms, tunnels and aircraft

passenger compartments, etc. For hotels, the prime objective is life safety although

hotel property protection is also critical. For remote heritage churches, parks building

and galleries, the objective is mainly to protect the property and heritage preservation.

Water mist [Putorti and Twilley 1995] is currently viewed as a possible alternative to

halon 1301 for use in electrical equipment rooms. Potential applications include use in

telecommunication central office buildings, control rooms and computer rooms.

4.7 Non-water Fire Suppression Systems

Halogenated systems or agents, such as halon 1211, halon 1301 and halon 2402, are used

in portable fire extinguishers and central extinguishing systems. The extinguishing

mechanism o f halogenated agents is not very well known. It seems that a chemical

reaction appears to interfere with the combustion process. The agents act by breaking

77

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 the chemical species involved in the flame chain reaction. While all the halogens are

active in this way, bromine is much more effective than chlorine or fluorine. Moreover,

the toxicity o f halon 1301, halon 1211 and halon 2402 should be considered. There are

two types o f application systems, namely the total flooding system and the local

application system. It is understood that halons are potential ozone-depleting substances

and should not be released to the atmosphere except to extinguish the fire.

Due to the ozone depletion potential o f halon systems, currently halon replacement

agents and systems are sought for. There are some practical applications, such as the

FM200, Argon and Inergen system, etc.

The use o f foam [NFPA 1997] as extinguishing agents and systems can be considered.

Foam is produced by mixing a foam concentrate with water at the appropriate

concentration, and than aerating and agitating the solution to form the bubble structure.

Low expansion foam is used to extinguish flammable or tank fires by the application to

develop a cooling, coherent blanket. Foam can also be used to diminish or halt the

generation o f flammable vapor from non-burning solids or liquids. It is also important

where aircrafts are fuelled and operated. Currently, warehouses and buildings storing

large quantities o f combustible and flammable liquids are protected by foam-water

sprinkler systems. The protection required is a function o f the type and quantity of

liquid stored, building height, and storage configuration. Foam breaks down and

vaporizes its water content under attack by heat and flame. It therefore must be applied

to a burning liquid surface in sufficient volume and rate to compensate for this loss, with

an additional amount applied to guarantee a residual foam layer over the extinguished

liquid.

78

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 5 FIRE SAFETY MANAGEMENT IN A HOTEL -

SOFTWARE COMPONENT

Fire safety management is a very important software component to the hotel operation.

It is also a basic concept in providing total safety to the building itself. In the following

sections, various views and discussions on the application o f fire safety management will

be outlined. Fire safety management should be started by using an engineering

approach at the design stage, on the design o f fire provisions for new hotel building

projects. In the fire safety manual, all the fire safety objectives should be listed out.

On the other hand, a ranking system is used for the fire provisions, in comparing with the

new fire codes, on the passive fire protection and active fire protection systems installed

in existing hotel buildings. From the comparison, the fire safety management program

is planned to supplement the inadequacy in the existing hardware components. Based

on the fire safety objectives and assumptions originally planned, a fire safety plan can be

worked out for the hotel. Control legislation by the government department is

recommended on the implementation o f these fire safety management plans. The hotel

fire safety management program and the implementation o f the hotel fire evacuation

plans will be examined in the following sections.

5.1 Introduction

Fire safety management is very critical in the concept o f providing total safety in a hotel

building. Various views from the literature towards fire safety management are

available. Malhotra [1987] had a very detailed review on fire safety management and

79

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 the fire grading report [HMSO 1952] prepared in 1952 in the UK is regarded as the first

document on fire safety management. In this report, it mentions the maintenance o f the

provisions for means o f escape, regular inspection o f all doors, passages and staircases

and also about external facilities which might become unsafe due to external exposures.

Next, the BS5588 [BSI 1983, 1985] “Fire precautions on the design and construction of

buildings” Parts 2, 1983 & Part 3, 1985 are available on advices concerning the

management and evacuation procedures in case o f fire. The appendices o f these two

parts focus on giving valuable guidance to the management on using staff for fire safety,

training and function, keeping records, preparing notices and calling the fire brigade. At

the same time, these appendices also advise on the defined actions in case of fire and

evacuation procedures. The hotel management or owner must work out a fire safety

plan by referring to the size o f the building and its occupancy. In using engineering

approach [BSI 1997, 1999] in hotel design, qualified fire engineers should be involved to

formulate the design objectives and the fire safety plan and design. In working out the

fire safety management program, the fire safety engineers must familiarize with all the

fire safety design assumptions and objectives.

Following the introduction o f BS5588 Part 11 in 1997 [BSI 1997] on “Code o f practice

for design offices, industrial storage and other similar buildings”, it focused on the

effective management combined with appropriate staff training to be important in taking

correct actions and having the occupants evacuated safely. Genuine and coordinated

advices should be sought from the local fire brigade and the hotel management should be

aware o f the statutory requirements regarding the maintenance o f means o f escape, fire

80

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 warning systems, portable fire extinguishers, escape lighting and fire safety instructions

to staff.

At least two points should be considered carefully in the recommendations from the

above document concerning the commissioning and handover o f fire safety installation

and fire safety manual. One important point is that a fire safety manager should be

appointed to take overall control o f the premises and day-to-day safety management of

the building. In a hotel, the fire safety manager is usually the safety and security

manager or the Director o f Engineering. Moreover, it is also necessary to prepare a fire

safety manual which should be kept in a safe place inside the hotel and maintained by a

competent person.

The fire safety manual should explain the fire safety planning, constructions and systems

designed; and their relationship to overall safety and evacuation management. It refers

to the documentation produced at the hotel design stage for using different types o f fire

protection system in different incidents and the responsibility o f staff. The manual

should also include drawings o f the hotel building identifying the different smoke control

zones and fire detection zones. Finally, the manual should record all the routine fire

maintenance activities and the drawings o f the basic fire precaution measures.

Recent views [HMSO 1996] on fire safety management focus on the hazard both to the

people inside the hotel and the building content itself. In BSI-DD240: Part I : 1997

“Fire Safety Engineering — Part I : Guide to the application o f fire safety emergency

principles”, fire safety management is important to the success o f the safety engineering

81

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 design. People involved in the fire safety management program are represented in the

Qualitative Design Review team.

In BS ISO/TR 13387-1 : 1999 “Fire Safety Engineering - Part I : Application o f fire

performance concepts to design objectives”, the prevention and control o f fires, the

evacuation o f occupants and the maintenance o f fire safety system are critical parts of the

fire safety management programs. Malhotra in 1987 specified that fire code to be added

in supplementing the building code and maintenance o f fire protective measure was

mentioned.

System approach to fire safety management was considered in the National Fire

Protection Association and Fire Safety Concepts Tree. How the fire safety goals and

objectives are affected will be examined together on all the fire safety features provided.

Following this approach, a Fire Safety Management Handbook was published by the

American Society o f Safety Engineers, USA.

The implementation o f a fire safety management program [Della 1999] can help to

reduce the hotel insurance premium, business interruptions and create an efficient

working environment. The program will also improve the hotel reputation to the public,

customer service and even realize quality gains. In formulating the fire safety plan,

needs and capabilities should be first assessed, and the hotel facilities and any fire hazard

should be analyzed. Moreover, it is necessary to develop and implement fire prevention

and fire protection control and evaluate the overall effectiveness. There are several

critical components in forming a fire safety management program, namely, inspections,

education and training, fire suppression, emergency service, evaluation o f fire possibility,

82

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 fire prevention, reports and records keeping and most importantly, the communication

between all relevant parties involved.

5.2 Main Objectives o f Fire Safety Management in a Hotel

In case o f a fire, the main objectives [Malhorta 1987] o f the fire safety management

program are to ensure:

• All the fire safety measures provided will be available for use.

• Hotel occupants will be able to use all the fire safety measures.

• Hotel occupants will be assisted to escape to a safe place.

Failing to implement the above management program and guidelines would lead to heavy

casualties during fire outbreaks. Hence, fire safety management in a hotel at least has to

focus on the following three aspects:

a. The hotel management has to ensure that all the fire safety measures provided are

maintained, serviced and kept in a proper functional order.

b. The hotel management has to initiate appropriate actions during fire outbreaks

which would assist occupants to reach a safe place.

83

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 c. There may be renovations or refurbishments in the hotel. Hence, the hotel

management has to review the adequacy o f current fire safety measures in case

there is a change o f building, building use and new technology on fire services

installation.

The hotel management has the obligation to ensure that all fire safety provisions are

maintained properly. The management should also train their staff on the available fire

protection systems so as to assist the local fire brigade or even deal with smaller fire

incidents.

5.3 Main Objectives of a Fire Safety Plan

A fire safety plan should be prepared in fire safety management, there may be three

essential components [Malhorta 1987]: A maintenance plan for proper service and

maintenance o f the hotel passive and active fire protection systems; a proper and

organized staff training plan which will include major training schemes for all hotel

employees; and a proper and organized fire action plan that will list out all required and

detailed actions to be implemented by each staff in case o f a fire.

In the hotel fire maintenance plan, basically the following should be included to maintain

all the passive fire protection systems (such as escape routes and fire doors), to maintain

all the active fire protection systems (such as fire detection system, sprinklers, portable

extinguisher, hydrants and hosereels and special systems), to verify the system

performance and the system integrity interfaces at regular intervals, to include

84

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 information and plans on layout, escape routes and egress signs for hotel occupants and

to have good housekeeping such as proper disposal o f rubbish and the use o f heat sources

like gas cookers. This is particularly important when the hotel guestrooms nowadays

are very often used as service apartments where usual residential cooking or life styles

will be happening inside.

A proper hotel staff training plan should include detailed descriptions o f the duties of

each staff, nomination o f the fire warden, training on the use o f hotel equipment,

particularly safety in handling electrical and gas equipment in kitchens and laundries,

proper procedures and guidelines to guide guests and staff to the safe place.

The hotel fire action plan should include proper guidelines and procedures on reporting to

the local fire brigade, proper guidelines to assemble hotel guests and staff and bring them

safely to the assembly places, guidelines, procedure and techniques to attack fire in the

case o f a small fire incident, procedures and proper routing to assist the local fire brigade

and to conduct a proper roll call at the assembly place for all hotel guests and staff

reported to be inside the building.

A fire safety manager should be appointed and responsible to the hotel top management.

It is recommended to have a permanent staff for a hotel with more than 100 staff; or the

expected hotel occupancy loading is more than 500. Other than that, some members o f

the hotel staff can be appointed as fire safety officers. Hotel staff recruited as

‘fire-fighters’ in large hotels should be trained, for example, in a local fireman training

school in Pat Heung in the New Territories.

85

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 In addition to the above three plans mentioned, a fire prevention plan is also required.

This plan will identify the use and maintenance of hotel items which could be ignition

sources, or combustibles which can lead to rapid fire spreading upon ignition.

Examples are taking care o f all electrical appliances in the guestrooms, kitchens and

laundries; and waste materials and rubbish handling. In other words, good

housekeeping in a hotel should be done properly.

5.4 Integration with Fire Safety Engineering in New and Existing Hotel

Buildings

Recent advance on the role o f fire safety management in fire safety engineering was

discussed below [BD 1998; Chow 1999]. It was reported that safety systems might not

be properly maintained and proper management o f the building should be implemented.

The importance o f integrating fire safety management into fire safety engineering during

the concept and design stages was pointed out. Advanced techniques such as fire

models [Forney and Moss 1994; Yang 1999] should be used at the design stage with an

awareness o f the management impact. It is essential that a fire safety engineer

understands how the effectiveness o f the safety systems design will be influenced by fire

safety management. The duties o f a fire safety engineer are proposed.

In passive fire protection system, they include structural fire protection measures

including compartmentation, fire doors, fire stops and protected means o f escape. In

active fire protection system, they include hosereel / fire hydrants system, emergency

86

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 lighting and exit-way guidance, smoke management system, fire detection system, fire

suppression system and fire alarm system.

The work o f the fire safety manager is important and the major tasks supposed are in

Appendix A.

In addition, a fire safety engineer can direct the work o f the fire safety manager in three

aspects: to assist and ensure that the active fire protection systems are properly

maintained and tested, to ensure that passive fire safety systems are not made ineffective.

For example, the removal of lift doors while replacing lifts in a big old highrise building

that had led to a big fire in Hong Kong [Chow 1998] and to manage the hotel building in

conforming to the assumptions made on fire safety design, such as controlling the fire

load and occupancy loading.

However, in Hong Kong, hotels are constructed at different times and it is difficult for

them to satisfy all the new fire codes [FSD 1998] and regulations. As mentioned

previously, in any building design, the means o f access code, means o f escape code, fire

resisting construction code and the fire services installation code should all be satisfied.

Before 1972 [Chow and Wong 1998], the fire codes were not detailed and even the types

o f building occupancies were not clearly defined. Commercial buildings could not be

distinguished from residential buildings during this period and they were usually

regarded as old buildings. However, there were major improvements by 1987 and a lot

o f the fire safety provisions, for example, minimum corridor width, were incorporated in

87

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 the code. Since 1987, sprinkler system has been required nearly for all non-residential

highrise buildings and the fire safety provisions have significantly been enhanced.

For those existing buildings, a fire safety ranking system is suggested to be adopted and

based on current code in order to assess the fire safety aspects on the passive building

design and active fire protection system. If there are any deviations between the

existing fire safety provision and those in the new codes, then a fire safety management

scheme needs to be worked out to compensate for the shortfall. It is difficult for the

existing buildings to satisfy the new codes. However, an alternative solution may be

adopted in applying engineering performance-based fire codes upon fire safety provisions

design. Nevertheless, it will take certain time to develop the necessary codes and the

cost might be high.

To give total safety to hotel buildings, fire safety management must be properly

implemented. Fire safety management might not be controlled by building regulations.

It is noted that the fire codes for dealing with fire safety provision for new hotels are not

yet in use. In Hong Kong [Effective building management, 1998], the Buildings

Department, Fire Services Department, and the licensing section of the Food,

Environment and Hygiene Bureau and even the Health and Safety Officer o f the Labour

Department should take care of the fire safety management. In other words, an overall

fire safety system, or a total safety system, including design, construction and

management o f the fire safety system should be worked out. However, there are still

not yet any codes related to this topic. The fire safety management program in a hotel is

described in Appendix B.

88

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 5.5 Development of an Effective Fire Safety Management Program

Understanding the fire safety organizations and having basic knowledge o f the available

resources are important in fire safety management. In Hong Kong, the government

departments, such as Buildings Department, Fire Services Department and relevant

academic and professional institutions can provide most o f the information for this fire

safety management program. Knowing where to go, what facilities and equipment they

process and their ability to respond would be o f great assistance in organizing a plan of

actions. First-hand knowledge o f fire fighting resources at one’s command is one o f the

keys in determining whether a fire of a certain size can be controlled with a minimum of

damage.

Then, the fire chemistry and its relationship with the particular hotel involved must be

understood. Whether it is the fire triangle or fire tetrahedron, it is important for the

hotel management team to prevent the combination o f fuel, oxygen or heat that can

initiate a fire.

In previous sections, essential elements have been mentioned that make up the fire safety

management and also the fire prevention plan and control o f hazardous materials in a

hotel. Care and maintenance o f active and passive fire protection systems in a hotel are

considered. In particular, various legal aspects, legislation and organizations involved

in forming an effective hotel fire management program are discussed.

89

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 6 METHODS OF STUDY

Throughout the whole study, a practical research methodology has been applied in order

to investigate various fire safety aspects [CIBSE 2003] in big hotels. First o f all, it is

understood from previous chapters that hotels are complex building constructions. Fire

safety codes in Hong Kong, USA and China have been reviewed. This is essential so

that we can know the differences in the statutory fire codes in various countries and their

applications in big hotel projects.

For special risks in the hotels, the fire safety aspects in kitchens will be studied. Hazard

assessment in hotel kitchens will be carried out by full-scale burning tests [Chow 2001b]

in a model kitchen in a remote town in Northern China. This will give the heat release

rate [Babrauskas and Peacock 1992] and air temperature distribution when a model

kitchen is burnt. Experiments will also be conducted for burning kitchen, under the

operation o f a sprinkler system and water mist system respectively to study the fire

extinguishing time.

To maintain the integrity o f the passive building construction, the operation o f heat

response links at fire damper is important during the breakout o f a fire. Heated wind

tunnel tests at fusible link samples will be conducted, in accordance with the setup and

requirements specified in UL33 [UL 2005] concerning heat response links for fire

p ro tectio n ser v ic e. T h e te sts are im portant to understand the p erform an ce and integrity

for locally made fusible links.

90

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 6.1 Fire Safety Codes Review

In previous chapters, the concept o f total fire safety in hotels has been studied. Building

codes have been studied that concern the passive fire protection systems and the fire

service installation codes that relate to the active fire protection system in Hong Kong.

Furthermore, the main licensing requirements on fire and building safety are listed out as

specified in the local Hotel and Guesthouse Ordinance. These local codes and

ordinances must be fully satisfied for any hotel buildings in Hong Kong. In comparison,

the NFPA 101 [NFPA 1994] life safety code will be listed and reviewed.

It is understood that hotels or guesthouses are mainly used for stays o f a relatively short

duration. The accommodation o f sleeping hotel guests is a main characteristic in the

provisions o f the life safety code because occupants who are asleep are unaware of a

developing fire. The guests when awakened to be alerted to the fire outbreak might also

be confused. The possibility o f being asleep and familiarity with the surroundings

during a fire are main factors that endanger the hotel guests’ safety. Hotels lead to

additional safety issue, because escaping guests need to transverse smoky and hot interior

corridors in typical hotel configurations before searching the exits.

The content o f a hotel is classified as light hazard and designed in accordance with NFPA

13 standard for the installation o f sprinkler systems. The difference in classification in

the code is based on the human life threat when comparing with the extinguishing effects

o f the automatic sprinkler system. The occupant load calculations for areas o f hotels

used for non-residential purposes should be based on the occupants load factors

applicable to the use o f the area.

91

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 In hotel buildings protected by automatic sprinkler system, exit enclosures shall have a

fire resistance rating o f not less than one hour, and the protection rating o f doors shall be

not less than one hour.

Hotels are not allowed to have any door locked against escape while the hotel building is

occupied. This requirement [NFPA 1997] allows the door to be equipped with a locking

service that permits the door to be opened from within the building. The safety code

requires a minimum o f two separate exits from each floor. A third exit is required when

the hotel guests loading o f a floor exceeds 500 and a forth exit when it exceeds 1000.

This requirement will probably have little effect on modem hotel design practices

because hotel guest floor large enough to accommodate more than 500 guests would

probably be provided with more than two exits based on travel distance limit

consideration.

If the hotel building is protected through by an automatic sprinkler system, the common

path o f travel is permitted to be 15 m, which is the same distance permitted for the length

o f existing dead-end corridors, regardless o f sprinkler system considerations. Travel

distance within a guestroom or suite to a corridor door shall not exceed 23 m. However,

travel distance from the corridor door o f any guest room to the nearest exit shall not

exceed 30 m. The distance o f travel from the termination o f the exit enclosure to an

exterior door leading to a public way shall not exceed 30 m.

Emergency lighting must be provided in all hotel buildings with more than 25 rooms.

Any vertical opening must be enclosed or protected. All hazardous rooms in a hotel,

subject to possible explosion, must be effectively cut off from other parts o f the building.

92

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Particular attention must be paid to new draperies, curtains and other similar loosely hung

furniture and decorations in a hotel regarding their fire resistance characteristics and

compliance testing.

Detection, alarm and communications systems must be installed in hotels. Guest rooms

and suites specifically required and equipped to accommodate hearing impaired

individuals shall be provided with a visible notification appliance. The code requires

that, in additional to the normal distribution of manual fire alarm stations, the front desk,

telephone operator’s location or similar location must also be equipped with a manual

pull station. The intent is that a pull station is to be available at the location where guest

phone is an emergency. The smoke alarms installed in sleeping rooms are usually

single-station alarms that are provided for the sole purpose o f notifying the occupants o f a

smoky condition within that room. Thus, the alarms are not part o f a required automatic

detection system and are not required to initiate the building alarm system.

Hotel buildings must be protected throughout by an approved automatic sprinkler system.

In some cases, hotels are required to use quick-response sprinkler throughout guestrooms.

The technology associated with quick response sprinklers help to maintain tenability

within the room o f fire origin.

Tables 6.1, 6.2 and 6.3 also summarize some o f the differences in China code [MPS

China, 16-87; MPS China, 45-82] and NFPA requirements [NFPA 1994, 1995],

particularly on escape routes and travel distance limits.

93

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 6.2 Hazard Assessment

There are certain risky areas in a big hotel, particularly the kitchen, which has been

mentioned in previous chapters. Hazard assessment has to be carried out for the kitchen

in a big hotel. Material burning must be understood, whether it is solid, gas or liquid.

Burning involves oxidation which is also accompanied by the release o f energy as heat or

light. The first parameter o f hazard assessment [Babrauskas 2003] o f a material might

be the amount o f heat it produces when it bums, i.e. the heat o f combustion. The

chemical composition o f the material will almost determine the amount of heat produced,

which depends on the physical form o f the material.

Methane and propane, the simplest organic compounds, are common gaseous fuels. The

richest variety o f materials is organic solid, like wood, paper, textiles and plastics,

commonly encountered in hotel fires. The molecules that form most organic materials

comprise o f many thousands o f atoms linked together to form chains and networks

necessary for its useful mechanical properties. Polymerization is the process and the

origin of almost all plastics and synthetic polymers. Nearly all organic materials bum

readily with the release o f water vapour and carbon dioxide upon complete combustion.

It is found that the heat produced per unit weight of oxygen consumed is the same within

about 10 percent. This is another important parameter and allows us to use oxygen

consumption as a reasonable measure o f the heat produced by a burning organic material.

Organic materials fall into two classes, namely hydrocarbon or cellulose. From energy

point of point, oil or natural gas is a better fuel than wood, which consists mainly of

cellulose.

94

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Hazard assessment concerns the production o f heat, smoke or toxic gases produced

during burning and the rate at which these fire products are produced. Some factors

contributing to the burning rate will be examined. Burning is in general a vapour phase

phenomenon when the fuel and oxygen are brought into contact at a molecular level.

The rate o f burning depends on how fast the chemical reaction o f oxidation occurs. At

the preliminary stage or in premixed flames, the burning rate depends on the inherent rate

at which the substances combine. The rate is quite fast and hence it will be very

dangerous to contact the air and combustible vapour and the burning process is difficult

to be interrupted.

The burning behaviour o f materials may differ greatly. They may vary from the fast end

o f the spectrum as explosion to the slow end side like auto-oxidation and smouldering.

Auto-oxidation is the combination o f a material with oxygen at a rate far too slow to

produce the heat and light normally associated with fire. Self-heating can occur if the

heat produced by the auto-oxidation is not removed. On the other hand, smouldering is

a burning process quite different from flaming. Only a small fraction o f combustible

materials, those which can produce porous char during the course o f combustion, will

smoulder. Most smoulders are organic which can yield both volatile fuel species and a

rigid, porous char structure.

Different burning behaviour and characteristics o f gases, liquids and solids pose different

types o f hazards. The methods for measuring and controlling them are also different.

It may be better to consider fire hazard control relating firstly to the likelihood of

ignition, then the control o f fire spread and lastly the management o f the fire if ignition

and spread cannot be prevented.

95

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Flammable gases usually bum quickly once ignited. The main action to avoid ignition

is firstly the determination o f the flammability limit, and secondly to operate so that the

concentration is kept beyond those limits. Moreover, the possibility o f ignition can also

be reduced by the addition o f chemical inhibitor to raise the lower flammability limits.

In practice, ignition control is done by stringent storage and handling safeguards. For

example, dangerous goods licence is needed to be obtained for the hotel for storage o f

matches or fireworks materials. Alternatively, it may be possible to reduce the oxygen

available in closed environment and therefore raise the effective flammability limit

without a major breakthrough in the prevention and suppression o f large flammable gas

fires. The most effective means o f minimizing the fire impact is advanced planning,

preparedness and response o f trained emergency personnel.

Oil in kitchen is a major fire hazard. Burning actually occurs in the vapour phase and

the most hazardous combustible liquids are those with a high vapour pressure. Serious

measures exist to minimize the escape of flammable vapours in the handling o f volatile

liquids. If a fire o f such kind is initiated, means are also available to prevent the supply

o f additional combustible fuel to the fire. These include designs for the venting of

storage tanks and flame arresters, etc.

For fires involving solid materials, there is a variety o f tests to determine their

susceptibility to ignition. The fire hazards posed by inorganic structural materials are

most likely to be passive. For example, structural steel may lose its strength, concrete

can crack and spall, and glass can break or melt. Such materials are therefore rated on

the ability to withstand such high temperature effects. Improvements in ignition control

96

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 of materials have come about as a result of flame resistant developed for both natural and

man-made materials.

6.3 Necessity of Experiments in the Study

Very few experimental studies on heat release rate measurement for combustibles had

been carried out [Chow 2001b]. Hence, carrying out experiments for this heat release

rate study is important so that fire safety problems can be addressed properly.

Moreover, experiments [Chow et al 2003c] had not been carried out before to verify the

performance and operation o f fusible links used in fire services systems. In designing a

fire services installation for a big hotel, an engineer requires to choose the most

economical design which satisfies all the specifications, including the aesthetic

requirements and owner’s needs. In designing an experiment [Chow 2002] to measure

the heat release rate when burning a model kitchen, questions on cost are too frequently

focused. The main reason may be because o f the difficulty in deciding a good estimator

on the time necessary for the experiment. Also, it is partly a traditional consideration

that science is above monetary considerations. Economic factors must be taken into

account in view o f the increasing cost o f experiments in research. Needless to say,

experiments must be carried out in a cheaper way when it can be done equally effectively

with less expenditure. It will be difficult to decide whether a given release rate

measuring experiment should be carried out, considering its probable cost involved.

Budget cost estimates [Bright 1990] must include direct expenditures for materials,

salaries and overheads. When the experiment is conducted, sometimes we will be

shocked by the magnitude o f overhead costs. In general, the overhead costs include

97

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 laboratory rent, electricity, heat and administrative expenses o f the laboratory. Usually,

these costs amount to nearly 40 percent of the total direct salaries and wages.

A usual source o f waste is the misuse o f the time o f salaried personnel with scientific and

proper training in handling all the heat release rate experimental apparatuses and

procedures. There are many jobs which less highly trained assistants could carry out

equally well. One main reason for this misuse o f talent is the low salary scale so often

paid for scientists in China.

Costs will also be a serious consideration on whether to buy a piece o f apparatus for the

experiment or to build it, like the Chinese calorimetry in Lanxi [Chow et al 2003d] in

Northern China. If it is true that this is available commercially just what is needed, it is

usually cheaper to buy it. In considering the cost of building an apparatus, it makes a

difference whether a salaried machinist has time not needed for other work or whether he

is occupied for his full capacities. The former situation is practically a pure hypothetical

one in most laboratories. It is necessary to add the cost o f the supervising and designing

scientists’ time and also the cost involved in making changes and adjustments in an

untested design.

In planning actual experiments for measuring heat release rate in kitchen, it is necessary

to have a good basic understanding o f the nature o f the problem and any relevant theory

associated with it. Moreover, an experiment usually will be designed on the basis o f

one or more preliminary hypotheses. These can be constructed more effectively if good

knowledge o f the theory and background o f the situation is available. It is sensible if

the problem is analysed and is thereby enabled to cast it into the simplest form. It will

98

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 also be possible to divide the problems into parts when they can be much more easily

answered separately than together. Moreover, it will be desirable or even necessary to

approach the answers in stages, starting with the most idealized and simplified version

possible. In certain cases, it is possible to construct a single experiment whose outcome

largely determines the fate o f a given hypothesis in theory. Nevertheless, it is important

to design the experiments that are as far as possible crucial with respect to the hypothesis

under consideration. This may not be easy as it sounds because there are alternative

interpretations for the results o f most experiments. It will be bad to carry out an

experiment without a clear-cut idea in advance o f just what is being tested. Usually,

however, a certain objective is the reason for undertaking a research, and yet when the

experiments are over, it becomes apparent that the questions asked were not the ones

whose answers were really needed. It will be safe to go right back to the origin o f the

inquiry and ask at every stage on why doing this experiment, why testing the fire

scenarios under the operation o f water mist and sprinkler system.

Equipment or apparatus is needed in the heat release rate measurement tests. The

apparatus may serve to hold certain variables constant and to change other variables in a

prescribed way. It often but not always involves measurement. Apparatus may also

provide appropriate conditions for changing objects or materials as a preliminary to

observation. In designing the heat release rate tests in a kitchen, specifications must be

made up for every part, no matter small, based on the duty that part is to perform The

design questions can be qualitative or quantitative. If the qualitative possibilities are not

known, it is unlikely to make decisions correctly. Here, experience is o f great help, but

the experience o f others is used by the wise. However, quantitative decisions are

usually harder and may require calculations, sometimes based on difficult theories. It is

99

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 needed to minimize the cost o f the design by close calculations, or to calculate a lower or

upper limit for some specification which may affect for the design. Misunderstood or

incomplete instructions are a serious source o f waste and delay. The more the technician

is told about the uses o f the part, the more likely he is to produce what is needed. It is

important to be able to tell at all times that the different parts o f an apparatus are in

proper adjustment and working conditions since instruments are expensive in the heat

release rate measurements. They should be used where they will do the most good.

Expense can also be reduced by using the same instrument for several tests, provided that

this can be conveniently arranged, or by switching. To facilitate test operation, all

adjustments should carry at least a rough scale. It will be harder to tune up a

complicated system if it is not possible to go back to a previously attained setting. The

more precise is the given adjustment needs to be made, the finer the scale should be.

The basic purpose o f our experiment [Chow and Meng 2004d] is to test the effect o f a

certain selected variable, the heat release rate, when a model kitchen is burnt. To do this

successfully, other variables which may influence the result need to be kept constant.

As a result, one o f the principal problems is apparatus design on how to keep these

variables from disturbing the situation.

6.4 Heat Release Rate

In this study, heat release rate measurements were carried in a model kitchen with the

operational effects o f water mist and sprinkler systems. Heat release rate (HRR) study

[Babrauskas and Grayson 1992] was developed in 1980s and is an important aspect of

100

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 fire protection engineering. In practical fire engineering and safety design, the question

“How big is the fire?” should be answered. With the heat release rate measurement, the

necessary quantifying answer to the above question and the very cornerstone o f fire

protection engineering are provided. The knowledge o f heat release rate becomes

necessary for all trades, including fire protection design engineers, building officials,

chemists and all other technical specialists. When going into the 1990s, the focus on

heat release rate becomes obvious in most of the fire research and problems.

Heat release rate is a major fire parameter in any fire safety engineering design because it

defines the size o f the fire. With this quantification, it also indicates a lot other fire

parameters, such as smoke and toxic gas production. It is important in fire engineering

for having the accurate heat release rate on items like building contents furniture, wall

linings and industrial storages, etc.

To successfully measure heat release rate from burning materials, early research works

[Krasuy and Babrauskas 2001] were usually interrupted by difficulties in constructing an

efficient small-scale adiabatic systems used for materials testing. For testing building

products or furniture under large-scale apparatus, this will involve great costs and

difficulties which hamper these tests from practical applications. The use o f oxygen

depletion calorimetry as the methodology to measure heat release rate has overcome the

above difficulty. Currently, it will be comparatively cheap and easy to carry out both

bench and full-scale heat release rate tests on materials in fire environment.

It is believed that heat release rate measurements were initially carried out on systems

that monitored temperature rises in flue o f furnaces and reproduced the same temperature

101

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 rise from metered control burners. The first FM construction materials calorimeter was

built in 1959 but it is a very bulky apparatus. Actual bench-scale work on materials was

done on insulated box systems which used calibrated thermocouples in the flues of

smaller apparatus. However, poor insulation and thermal feedback between the sample

and the radiator are some o f the problems with the instrument. The International

Organization for Standardization both examined this apparatus and also started to

develop their own insulated box test. Both were eventually rejected as oxygen

consumption calorimetry became available.

Oxygen consumption calorimetry is based on the theory that for a large number o f liquids

and gases, a constant net amount o f heat is released per unit mass o f oxygen consumed in

combustion. The heat release can be readily measured if all combustion products from a

burning test are collected in an extraction system and the flow and oxygen concentration

in that duct is accurately measured. Such a technique has considerable advantages for

large-scale reaction to fire tests as it does not limit the burning tests by carrying them out

unnecessarily in compartments. This technique has gained considerable support and

there is now a list o f the laboratories working with large-scale oxygen depletion

calorimeters.

On the other hand, bench-scale oxygen depletion calorimeter [Babrauskas 2003] has been

designed with a number o f geometries, such as the cone calorimeter. This has become

the standard bench-scale test method for measurement of heat release rate. The cone

calorimeter has also made heat release measurement viable to the researchers with

standard laboratory facilities.

102

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Room fire test in full scale for surface products, ISO9705 [ISO 1993], was published in

1992. This standard provides for measuring heat release from wall lining materials

when attached to the surfaces o f a standard room rig. It is similar to a test method also

being developed by ASTM.

6.5 Fusible Links at Fire Dampers

Fusible links are installed in fire dampers and smoke curtains, and will be activated to

operate the damper in order to maintain the integrity o f the fire rating for air-ducts

passing through fire walls. Simulation tests, using the heated wind tunnel, will be

conducted in accordance with UL33 [UL 2005] to measure the performance o f fusible

links commonly used in Hong Kong. The results will be reported in later sections.

Hotels are usually equipped with very advanced and modem air-conditioning and

ventilation systems for the comfort o f guests. If ducts need to penetrate fire rated walls,

floors or partitions, fire dampers with equivalent fire-resistance ratings must be provided

to maintain the integrity o f the fire areas. Damper will also be used to prevent back

draft in duct systems to keep unpleasant smell from being sucked upstream and spread in

the guestrooms.

Fire damper is installed in an exhaust hood [ASHRAE 1997] at the duct connection to the

exhaust system hood in a kitchen. Its main objective is to prevent fire from getting into

the ductwork. Fire dampers are useful in the hotel kitchen exhaust system because they

will help to keep fire out o f the existing ducts that may not be able to contain the fire.

103

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 They also assist to prevent the spread o f fire to other parts o f the hotel building and

provide safety in case the duct fire suppression system does not function.

However, fire dampers are considered not necessary as the installed duct system if

properly designed can exhaust the fire and smoke out o f the hotel building in a safe

manner. Moreover, closing the fire damper will turn the fire and smoke back into the

kitchen and spread more quickly. The smoke will make it more difficult for the fire

service personnel to even see to fight the fire. Furthermore, a fire kept out o f the duct

system and in the hotel kitchen space can spread quickly to combustible surrounding

greasy surfaces, increasing the hazard to other guests and other hotel functional areas.

The exhaust system may be so dirty and greasy that the damper cannot close properly and

so the damper only provides a false signal o f safety and makes it more difficult to access

the fire in the duct. It is clear that much o f the concern with dampers has to do with the

proper condition o f the ductwork beyond the damper, including its integrity and clearance

to combustibles.

Fire dampers in kitchen exhaust systems should only be installed in listed hoods at the

hood to duct connection, or they may be a component o f a listed grease extraction system

that is installed downstream o f the hood. Exhaust hoods with integral makeup air

supplies usually have fire dampers installed in the supply duct to hood connection, or at

one or more o f the makeup air register openings. Their fusible links have the same

inspection and replacement schedule.

Fire dampers and its associated fusible links also need to be checked for proper and full

operations. The links should be able to break rapidly in accordance with its temperature

104

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 setting. If it does not break fully, its surface must be properly cleaned to remove grease

and impurities. If the fire damper cannot close quickly and completely, it cannot be

effective as designed.

For local hotel building projects, many fire dampers will be used in guestrooms along the

primary air or exhaust duct systems. The local Fire Services Department currently is

using the sampling test procedures on fusible links for use in fire dampers o f ventilating

systems. The purpose o f the sample test is a site acceptance test where applicable for

locally made fusible links which are approved by local Fire Services Department. A

batch of samples will be drawn randomly from the hotel building ventilation system.

Ten percent o f the total quantity of fusible links o f the same product or a maximum up to

10 will be drawn by the fire services officer on site with the representative o f the

registered ventilation contractor for the test.

The operating temperature (bath) test will be carried out as follows:

a. The operation temperature (bath) test is a check on the absolute operating

temperature o f the links. It shall operate within 74 °C.

b. The vessel is to be prepared with a source for heating the water bath with means for

heating the liquid at a prescribed rate. An agitator is to be used as an aid in

obtaining uniformity in temperature o f the bath liquid and measuring the

temperature o f the liquid.

c. Water should be used in the bath test o f the fusible links

105

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 d. Mercury thermometers or other types of thermometers with valid calibration

certificate shall be used to determine the temperature o f the liquids employed in the

bath test. The bulbs o f thermometers should be held level with the fusible link by

a support member.

e. Place the fusible links in an upright position, completely immersed in the water bath

and independently apply a load o f not more than 0.4 kg to the end link, using a

suitable weight or adjustable spring balance.

f. Raise the temperature o f the bath liquid at a convenient rate until the liquid is at 60

°C.

g. Control the rate o f temperature rise at a rate not exceeding 1 °C in 2 minutes until

operation o f the fusible link, or until a temperature is reached at which the link fails

the test.

h. Record the temperature of the liquid and the time o f operation, as each heat

responsive fusible link operates.

All samples shall pass the test in principle. Otherwise, all links installed on site must be

replaced. Acceptance o f site test may vary on individual project basis, due

consideration will be given to the total quantity and the lots o f production being installed

In a later section, tests on fusible link samples will be conducted in accordance with the

conditions specified in UL33.

106

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Empirical equations, as described in Appendix C, concerning heat transfer at sprinkler are

also reviewed to indicate the sprinkler performance between air speeds and temperatures.

Large samples o f local fusible links were tested under various air temperatures and

speeds using the heated wind tunnel with the actuation time recorded. Graphs are also

plotted to verify the fusible link performance, such as response time index and time

constant, etc.

107

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 7 FULL-SCALE BURNING TESTS FOR KITCHENS

AND CALORIMETRY

In this chapter, the principles o f calorimetry and its development will be discussed. The

oxygen consumption calorimetry and cone calorimetry [Babrauskas and Grayson 1992]

will also be investigated and outlined. The Chinese assembly calorimeter [Chow et al

2003d] will then be described and the experimental site mentioned and explained. The

necessity o f carrying out full-scale burning tests and related issues will be discussed.

Full-scale burning tests in a sample kitchen will be reported and results analysed.

7.1 Calorimetry

The fire performance o f hotel buildings itself and other occupied spaces can be

considered as having two main characteristics: flammability and its stability to fire.

Flammability refers to the heat release rate, flame spread and also its ignitability. On

the other hand, stability refers to the performance o f the building structure itself against

fire and whether the partitions or compartments can withstand the effect o f fire.

Actually, standard testing methods for flammability [Babrauskas 2003] does not have a

long historical basis. Tests for the flammability o f textiles were not standardized until

the first version o f the current NFPA 701 Bunsen-bumer test proposed by the National

Fire Protection Association in 1938. With most other articles o f greater weight and

thickness, the heat release rate, flame spread, and ignitability may all need to be

108

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 considered. In many applications, it is found that these three variables may be highly

correlated.

On measuring heat release rate tests at a small scale, the earliest test method [Babrauskas

and Grayson 1992] was the FM Construction Materials Calorimeter, developed by

Thompson and co-workers in 1959. The apparatus, with a specimen size o f 1.22 m by

1.22 m, was tested in the horizontal and face-down orientation. The principle of

apparatus might best be described as a ‘substitution test’. A specimen was placed into

the apparatus and subjected to a prescribed exposure from an oil burner fire. The exhaust

temperature was recorded as a function o f time. A second test was then made, with a

non-combustible blank substituted for the specimen. The principle o f measuring was by

metering the amount o f propane flowing into the evaluating burner and the combustion

energy represented by the metered propane was then taken to correspond to the heat

release rate o f the specimen.

Improved design principle was available in 1972 which in present-day terminology could

be called an ‘isothermal’ design. In this design, a control section is established in the

stack, where thermocouples are used to monitor the temperature o f the exhaust gases.

The heat release rate is determined by including a mass flow measuring device in the

propane supply line. The specimen’s heat release rate is equal to the decrease in the

propane flow expressed in heat units. This instrument significantly simplified testing,

since it no longer required extensive calibrating runs for each test. However, there are

some limitations on its use. These included the very high apparatus complexity, the

sensitivity to exhaust pressure fluctuations and the need for a long equilibrium time prior

109

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 to the test. This NBS-I calorimeter was finally replaced by the NBS-II instrument and

was removed from service.

While the NBS-I calorimeter was not of much practical use, the SRI calorimeter

accommodated vertical specimens 457 x 610 mm in size and with a usable flux range of

15 to 70 W m'2. This apparatus was also not much used.

For the NBS-II calorimeter, the origins o f this apparatus were similar to those o f the SRI

calorimeter. The apparatus was intended to be more rugged, and to be less susceptible

to certain sources o f error identified on the NBS-I unit. A maximum heat release rate

measuring capability o f 1000 kWm'2 was provided for vertically oriented samples, with

twice that for horizontal ones. A major feature o f the new apparatus was that provisions

for a load cell were made from the very start, so that there would be a continuous record

o f the specimen mass loss. The apparatus, as constructed, worked generally to

specification. Problem areas included the inability to test single-sided vertical

specimens and the sensitivity o f the response to barometric pressure variations

propagating down the exhaust system. A more serious obstacle to adoption by other

laboratories was the fact that the apparatus required a high pressure natural gas supply, a

very large air compressor, and several other difficulties of construction and installation.

Thus, its users were restricted to exploratory research studies.

The Cone Calorimeter [Babrauskas 2003] was designed at NBS after experience had been

gained with the desirable and undesirable features of a number o f previous devices. The

calorimeters routinely operated at NBS included the NBS-I, NBS-II, and OSU apparatus.

In pursuit o f the best implementation o f the oxygen consumption principle, a number o f

110

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 other purely-developmental calorimeter was also constructed. Different heating

arrangements and various concepts for oxygen consumption gas trains were explored in

these units. The FMRC small-scale flammability apparatus and the SRI calorimeter

were also studied in detail in this design process, although they were never installed or

operated at the NBS laboratories. The instrument that resulted has been called the Cone

Calorimeter, due to the truncated cone shape o f the heater.

7.2 Oxygen Consumption Calorimetry

Thornton showed that for a large number o f organic liquids and gases, a more or less

constant net amount o f heat is released per unit mass o f oxygen consumed for complete

combustion [Huggett 1980]. Huggett found this to be true for all organic matters also

and obtained an average value for this constant to be 13.1 MJ kg"1 o f oxygen consumed.

This value may be used for practical applications and is accurate to within ± 5%.

Thornton’s rule implies that it is sufficient to measure the oxygen consumed in a

combustion system in order to determine the net heat released. This is particularly

useful for full-scale fire test applications as what had been done in the Chinese

calorimetry for kitchen heat release rate tests. Thus, for example, for compartment fires,

the oxygen consumption technique is much more accurate and easier to implement than

methods based on measuring all the terms in a heat balance o f the compartment. The

technique is now used extensively in many laboratories all over the world, both in

bench-scale and full-scale fire test applications.

ill

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The basic measuring principle is by collecting all the combustion products removed

through the exhaust duct. At a distance downstream with sufficient and adequate

mixing, both flow rate and composition o f the gases are measured. For open systems

like the room fire test, the furniture calorimeter and the standard cone calorimeter, a hood

is used to collect the combustion products and it is not possible to measure the air flow

rate into the system directly. The volume flow rate is measured downstream after the

gases have undergone expansion due to chemical reactions. For closed systems, the

mass flow rate o f the air into the system can be measured directly. The following

describe the main simplifying assumptions made:

a. The amount o f energy released by complete combustion per unit mass o f oxygen

consumed is taken as a constant, i.e. 13.1 MJ kg"1 o f oxygen.

b. All gases are considered to behave as ideal gases, i.e. one mole o f any gas is

assumed to occupy a constant volume at the same pressure and temperature.

c. Incoming air consists o f oxygen, carbon dioxide, water and nitrogen. All “inert”

gases, which do not take part in the combustion processes, are lumped into the

nitrogen.

d. Oxygen, carbon dioxide and carbon monoxide are measured on a dry basis, i.e.

water vapour is removed from the sample before analysis measurements are made.

Usually, there are two techniques to measure mass flow rate in the exhaust duct in the

kitchen fire test. The first technique measures mass flow rate m (in kg s '1) in term o f

112

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 the pressure drop AP (in Pa) across and temperature at an orifice plate Te (in °C)

1 1 1
through orifice plate coefficient C (in k g 2 m 2 °C 2 ):

(7.1)

If the test is operated within a narrow range o f conditions, C is approximately constant.

Its value can be determined via a gas burner calibration. However, if flow rates are

varied during a test or if temperature changes considerably, the effect on C o f the

Reynolds number and pressure at the downstream side o f the orifice plate must be taken

into account. The other option is to measure velocity at one point in the duct, usually

along the centerline.

Oxygen consumption calorimetry [Peacock and Babrauskas 1991] is primarily addressed

to measuring the heat release rate and the heat o f combustion [Babrauskas and Peacock

1992]. However, with some additional instrumentation, it can also be used to measure

smoke and toxic gas yields and the overall chemical composition o f the volatile pyrolysis

products from the fuel using the above measuring principle. For open systems, such as

the room fire test, the furniture calorimeter and the standard cone calorimeter, a hood is

used to collect all o f the combustion products. In a closed system, the burning takes

place in an enclosure. A closed system allows for the burning o f the specimen in

controlled atmospheres. It allows for a more precise measurement o f the mass flow rate

o f the in c o m in g o x y g e n .

It is essential that the hood and the exhaust system be designed to collect all of the

combustion products. Thus, the standard flow rate 0.024 m3 s '1 in the cone calorimeter

113

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 has been established to allow about a safety factor of 2 against spillage o f combustion

products. In applications such as the furniture calorimeter, the hood must not become so

hot to influence the burning rate o f the specimen due to thermal radiation feedback.

This may require the use o f water-cooled shields. A constriction at the entrance to the

duct serves to promote the mixing o f the combustion gases and the entrained air before

the measurement section is reached; without such forced mixing, highly inaccurate

measurements could result due to stratification. The measuring points should be

downstream o f the entrance to the horizontal section in order to establish a reasonably

developed flow profile. There must be another length o f straight duct beyond the

measuring point.

The exhaust blower must be able to withstand high temperatures o f the order o f 350 °C,

or more, depending on the fire size, the details o f duct layout, and the consequent heat

losses achieved prior to the blower location. It is especially important to note that most

simple blowers obey a ‘fan law’, which states that, as the temperature is raised, the

volume flow rate remains constant. This means that the mass flow rate drops inversely

to absolute temperature. This diminution does not occur if the blower is located in a

constant-temperature section or if a feedback loop is arranged to monitor the mass flow

rate and control the blower speed accordingly. The latter type of

constant-mass-flow-rate operation is provided as an option on the cone calorimeter. For

closed systems, the air is normally supplied under pressure, so that the use o f the fan in

the elevated temperature gas stream is avoided.

One o f the advances made during the last decade of heat release rate studies was the

incorporation o f load cells into the apparatuses for measuring heat release rate. This is

114

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 now done routinely in bench-scale and furniture calorimeter studies. Room fire tests

have also been equipped with load cells, although this may be much more difficult to do

successfully. One o f the major reasons why load cell measurements are essential is that

they allow scale-independent data to be derived. Load cells used in heat release rate

instruments have to be resistant to drift induced by temperature changes. Burning

objects often tend to shift or fall as they bum. Thus, the load cell must also be

mechanically robust and stable against jamming due to off-center load or torques.

A heat flux meter is used to set the intensity o f the thermal radiation field [Babrauskas

and Grayson 1992]. The meter may be either the Gardon or the Schmidt-Boelter type.

The Gardon type is a thin foil with the hot junction comprising a single wire brazed onto

the center o f the foil. The Schmidt-Boelter type consists o f a multi-layered thermopile

assembly. For fire test purposes, both are normally used in a configuration where the

sensing element is mounted on a water-cooled body. The meters may be configured as

radiometers or as total heat flux meters, depending upon whether they have windows.

Heat flux meters are normally used to monitor heat fluxes which are predominantly

convective. While incident radiation can be defined and monitored properly, there is no

such thing as ‘incident convection’. Convection heating can only be defined once the

temperature conditions o f both bodies have been defined. In a real fire, the heating is

usually to certain material which, as a minimum, will rise in temperature over time and at

worst may ignite. A small, water-cooled body does not represent this type o f convective

interchange at all well. If accurate measurements under such conditions are needed, the

normal procedure is to instrument with thermocouples the actual specimen exposed.

115

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The total absorbed heat flux can then be computed by inverting the solution o f the

Fourier heat flow equation.

7.3 Cone Calorimeter

For much o f the bench-scale work carried out over the past decades [Babrauskas 2003],

the cone calorimeter has been found by numerous laboratories to be the best tool for heat

release rate measurement. It has been designed to include a number o f functions,

including measuring effective heat o f combustion, ignitability and smoke and soot, etc.

It is emphasized that the cone calorimeter has been designed to use only oxygen

consumption calorimetry as its measurement principle. Other calorimeters which on

occasion use oxygen consumption principles, for example, the FMRC Flammability

Apparatus, may adopt other measuring technique. The convective fraction is dependent

on the details o f the apparatus design, and also the scale o f the specimen.

A kitchen fire burning near it maximum rate can give gas temperature over 1000 °C,

producing corresponding irradiances to walls and contents o f 150 kW m'2. Testing under

such extreme conditions may not be required; nonetheless, if post-flashover fires are to be

simulated, irradiances o f over 75 kWm'2 should be available, and preferably closer to 100

kWm'2. A significant convective component would negate the purpose of having a

radiant ignition test. Rather low convective fluxes can be achieved for specimens

oriented horizontally, face up, and with the prevailing air flow being upwards.

116

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The cone calorimeter derives its name from the conical shape o f the heater. Once the

decision had been made to use an electric resistance heater, running at a realistic

maximum temperature o f about 950 °C, its material and shape still had to be determined.

The material was simply decided, based on poor experiences with exposed-wire

resistance heaters and with silicon carbide rod-type heaters. That left the tube heater,

which consists o f a resistive wire element inside a protective tube, swaged ever a packing

of inorganic insulation. The tube is made o f Incoloy and can be bent to a desired shape.

To determine the best shape, the conical heater used in ISO 5657 ignitability apparatus

was examined. This was run to be promising from the point o f view o f the shape. The

proper shape had to have a hole in the middle, since otherwise a hot spot would occur at

the sample center, where the radiation view factor is the highest. The same heater had

to serve in both horizontal and vertical orientations. In the horizontal orientation, it was

essential that all the products of combustion flow out the hole in the middle, and not

splash on the header coil, or escape from the underside.

When measured with respect to a water-cooled heat flux meter, the results showed that in

the horizontal specimen orientation, the convective contribution was immeasurably small.

In the vertical orientation, the fraction was typically 8 to 12%. Thus, it can be said that

the objective o f having a test method where the heating is primarily radiant was

successfully met. For modelling o f test results, however, one may be more interested in

the possibility o f convective heat transfer to a specimen which is heated up, or even

burning, not to a calibration meter constrained by its water cooling jacket to be a

near-room temperature. The direction o f the heat flow was such as to represent a heat

loss from the specimen in all cases. The experimental data, however, shows a

117

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 surprisingly high variation o f the coefficient with irradiance, and thus some further

studies should be warranted.

The air flow rate through the system which is feasible must be bounded by certain limits.

The air velocity must not be so fast that ignition results are improperly affected. It must

also be not so slow that products o f combustion spill out o f the hood. If it is a closed

system, it should also be concerned about not being slow enough that the air/fuel ratio

drops into the fuel-rich regime; the standard cone calorimeter, however, has been

designed for ambient air testing, and so this consideration does not apply. The exhaust

system uses a high temperature cast iron blower to exhaust the gases and an orifice plate

flow meter. The orifice plate flow meter is instrumented with a differential pressure

transducer and a thermocouple. For specialized studies, where the entire combustion

system is glass-enclosed, it is possible to go to flow rates below 12 Is'1. Within such

enclosed systems, accurate measurements can be made down to about 9 Is'1 using the

standard orifice place; for lower flow rates, down to about 5 Is'1, the standard orifice plate

is replaced by one having a smaller opening.

In most cases, however, an external ignition source is desirable. This ignition source

should, in general, not impose an additional localized heating flux on the specimen. The

ignitor should reliably ignite a combustible gas mixture in its vicinity. The ignitor has

to be designed so as not to be extinguished by fire-retardant compounds from the

specimen or by air flows within the test apparatus.

The specimen thickness should be, as much as possible, the thickness o f the finished

product. There are limitations at both ends o f the scale, however. The instrument is

118

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 restricted to testing specimens not thicker than 50 mm. As a practical guide for testing

unknown commercial samples, it is desirable to specify that any specimens less than 6

mm thick should always be considered as needing to be tested over their in-use substrate.

Both horizontal and vertical specimen orientation are provided; however, it is considered

that the horizontal orientation is standard, while the vertical orientation is reserved for

special-purpose testing only. With proper precautions, the horizontal orientation can be

used for testing liquids and melting materials. For the cone calorimeter, a

commercial-design load cell was found which permits only up-and-down axial motion,

while being insensitive to torques or forces from other directions. The load cell has to

accommodate two different orientation specimen holders, and may need to hold

additional fixtures. Some categories o f specimens, however, present special problems.

These are specimens which either have a propensity to ignite first along the outside edge,

or when ignited, bum disproportionately vigorously near the edges.

7.4 Necessity o f Full-scale B urning Tests

Consequent to so many disastrous fires in Hong Kong, there are big concerns on building

fire safety. O f all the fire risk parameters considered, heat release rate [Peacock and

Bukowski 1994] is the most important one affecting the development o f a fire.

Nevertheless, data on heat release rate for combustible products are not yet available

[Chow 2002]. It is o f paramount importance to study the probable heat release rate by

burning those combustibles, such as the cooking oil burning in the hotel kitchen. Fire

models [Cox 1995; BSI 2002] should be considered.

119

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Full-scale burning tests should be performed to measure the heat release rate. The

results can be applied to investigate how the fire safety provisions can be improved.

These include providing better passive building design such as the structural elements;

and the appropriate active fire protection systems. The information will be important

for implementing the new building fire safety codes.

7.4.1 Developments in Full-scale B urning Tests

Before 1970, there was little need to conduct experimental studies o f the details o f room

fires. Room fire experiments were typically conducted as an adjunct to studying fire

endurance [Babrauskas and Grayson 1992]. For such purposes, it was necessary to

track the average room temperature, since this temperature was viewed as the boundary

condition determining what the wall, floor, column, etc., was exposed to. Neither the

heat release rate nor other aspects o f the room fire such as gas production rates were of

major interest. In conducting full-scale house bums, it was tried before 1950 to study

gas production rates as a means o f determining how early untenable environments might

exist.

During the 1970s, however, empirical room fire tests were regularly being conducted at

many fire research and testing facilities throughout the world. Instrumentation typically

comprised a multiplicity o f thermocouples; several probes where gas samples were

extracted; smoke meters; typically located at different heights along an open bum room

doorway; heat flux meters located on the walls of the bum room; and possibly, a load

platform. The load platform might register the weight o f a single burning item, but was

o f little use when fully-furnished rooms were tested. Despite the fundamental role o f

120

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 heat release rate in the room fire, there was no technique available to measure that.

Since neither the burning item’s mass loss rate nor the air and gas flow rates could, in

most instances, be determined, the measurements o f gas and smoke concentrations at

isolated measuring stations were not o f much use in tracking extraction rates.

Two developments needed to be become available before further progress could be made:

a robust instrument for measuring flow rates o f combustion products; and especially, a

practical technique for measuring heat release rate. Conventional velocity measurement

devices are normally precluded from use in fire applications due to several problems.

These include clogging o f small orifices and the inability to properly calibrate for high

temperate use. The new bi-directional velocity probe solved these problems of

measuring air flow rates in room, corridors, and in smoke extraction systems.

7.4.2 Room Calorimetry

The first attempt to develop some technique for measuring the rate o f heat release in full

scale was in 1978 by Warren Fitzgerald [Babrauskas and Grayson 1992]. A small room

o f size 2.7 m by 2.7 m by 2.7 m was instrumented with a large number o f thermocouples,

located in the gas space, the walls, and the exhaust duct. The room had a forced air

supply o f 0.16 mV1, from a small 0.15 m by 0.15 m supply duct, with another duct used
to let out the combustion products. The room was also equipped with a load cell and a

port for extracting gas samples. The facility was designed for measuring the burning

rate o f relatively small, free standing combustible items [Smith and Shaw 1999].

121

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The most significant efforts in recent years in the area o f full-scale test developments

have gone into developing open-air calorimeters. The reasons have to do with the

nature o f a room fire during its early stages. It is intuitively obvious that a very small

fire in a room will behave in the same way as if it were in the open air. Indeed,

post-flashover fires can reach a state known as ‘ventilation limited’. This means that so

much combustible vapours are being liberated from the fuel items that, on the average, all

o f the available inflowing oxygen is still insufficient to meet the oxygen demands o f the

fire. This does not quite mean that the combustion gases, if measured, will show exactly

zero oxygen. It is there because mixing is not perfect, and therefore, the oxygen is not

available to the fire in those locations where it is actually needed.

The ability to make heat release rate measurement under open-air burning conditions is

also rather recent. The current full-scale heat release rate methods date to two

developments in the early 1980s, one at NBS and the other at FMRC. At NBS, an

open-air full-scale calorimeter was built using the concepts o f oxygen consumption.

The device was termed the “furniture calorimeter”, since its earliest applications were for

the testing o f upholstered furniture. The term has since become somewhat misleading,

since numerous commodities have been tested in it. A calorimeter with a 2000 kW

capacity could be used to test fast-burning chairs, but would typically not have enough

capacity for a sofa. For testing items producing higher heat release rates, a ‘large-hood’

version o f the calorimeter was implemented. The instrumentation and the

measurements made are quite similar to the original furniture calorimeter.

122

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 7.5 An Assembly Calorimeter

In Hong Kong, it is difficult to select a site for full-scale burning tests as the land costs

are far too high. More importantly, there are tight environmental protection regulations

and real fire tests cannot be done. A site far from the urban area should be used for

carrying out such studies. In this way, environmental impact o f the burning tests can be

minimized. Further, there should be water, electricity and heating supply in the remote

areas which are cold.

An Assembly Calorimeter [Chow et al 2003d], has now been developed in a small town

Lanxi in a remote area o f Northern China, 200 km away from Harbin. There, a

full-scale burning hall is designated (Figure 7.1 (c), (f), (g), (h) to (p)). This is a joint

project supported by the Harbin Engineering University (HEU). The ‘duct section’ and

the associated instruments including the oxygen analyzer, carbon monoxide analyzer and

carbon dioxide analyzer at the Research Centre for Fire Engineering o f The Hong Kong

Polytechnic University were moved successfully to that site in June 2001 [Leung and

Chow 2001]. Both field [Chow and Zhu 2003a] and zone models [Chow and Meng

2004d] can be used to study the kitchen fire with heat release rate measured.

The burning halls are shown in the photos (Figure 7.1 (b), (i), (j)). A bigger hall was

used before in the first two years. A smaller one was used later on due to budget cut.

The model kitchen arrangement was placed in a small room o f size similar to the ISO

9705 room calorimeter. The room is o f length 3.6 m, width 2.4 m and height 2.4 m with

a door of height 2 m and width 0.8 m as in the photo. An exhaust hood was constructed

outside to measure the heat release rate by oxygen consumption method.

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 7.6 Kitchen Fire

There are concerns in fire safety in public areas in the Far East. Hotel fire safety is also

watched carefully as many tourists visiting big cities in there. There had been hotel fires

but most o f them were put out by the hotel staff. Some were controlled by discharging

water from the sprinkler system. These fires occurred due to poor maintenance or low

quality electrical installation and appliances in the guestrooms, kitchens and laundry

areas.

Kitchens in hotels are rated as areas with higher fire risk because o f cooking. Gas

stoves with high thermal power used to be installed in Chinese restaurants. Some of

them are even illegal installations but necessary to heat up the ‘woks’ (Chinese pans).

Such high power cooking ‘woks’ might give kitchen fires when the cooking oil inside is

heated up above the auto-ignition temperature.

As reported in the literature [Edwards 1998], about 50% o f all accidental fires in hotels,

restaurants and fast food outlets started in the kitchen involving cooking oil or fat fires.

The fuel temperature can reach up to 400 °C for a cooking oil or fat fire and the

auto-ignition temperature varies from 285 to 385 °C. The fuel temperature of

conventional test fires for Class B such as a heptane fire is only 50 °C. Therefore, the

cooking oil or fat fires are difficult to extinguish and easy to re-ignite, different from

conventional liquid fuel fires. A new fire classification, Class F, is defined by some

standard institutions such as the National Fire Protection Association, the British

Standards Institution and the International Standards Organization [Edwards 1998;

Voelkert 1998].

124

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Fire risk is usually specified by the fire load density (FLD), the total amount of

combustibles stored in a kitchen. An upper limit o f FLD is 1135 MJm" in local codes

[FSD 1998]. As the kitchen is small with high FLD, flashover would occur rapidly.

Large amount o f heat and smoke would be produced and spread out to other areas. The

space adjacent to the kitchen such as the dinning hall would be affected as investigated

before by zone models [Chow 1995].

In fact, FLD itself is not sufficient to quantify fire risk. The fire scenarios in a kitchen

have to be identified for hazard assessment. Heat release rate is the first parameter to

assess consequences o f the identified fire scenarios. There are not many such studies on

heat release rate for kitchen fires in the literature [Liu et al 2004]. It is necessary to

understand the probable heat release rates [Babrauskas and Grayson 1992] in an

accidental fire o f Chinese kitchen.

A series o f full-scale burning test [Chow et al 2003d] were carried out to investigate the

heat release rate on scenarios identified. Typical kitchen with ‘woks’ and cooking oil

commonly arranged in hotels are studied. The experiments were conducted in the

Assembly Calorimeter described as above. A model kitchen was constructed with an

exhaust hood and fan-duct system. Heat release rate were measured by the oxygen

consumption method. Two scenarios were identified in a typical Chinese kitchen setup

for a staff canteen with two woks; and in a typical Chinese kitchen setup for a large

banquet with six woks burning.

In following local fire regulations [FSD 1998] and subsequent hotel license requirements

[CNTA 2005] for restaurants and kitchens, active fire engineering systems (known as fire

125

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 services installation in Hong Kong) are required. Sprinkler systems with fast response

or higher temperature response heads are required. In addition, water mist [Chow et al

2003b] and gas protection system with clean agent such as heptafloroupropane are

allowed. Effect o f operating water system and the heat release rates will also be

evaluated.

A sprinkler system [Grant and Drysdale 2000] and a water mist fire suppression system

[Yao and Chow 2001] were installed. The cooking oil fires under the operation of

sprinkler and water mist system were studied.

7.7 Full-Scale Burning Tests on Kitchen Fire

To assess how sprinkler and water mist would control a fire, full-scale burning tests [Wu

et al 2005] were carried out. Kitchen fire scenarios with and without the operation of

sprinkler and water mist system were considered. A model Chinese kitchen as in Figure

7.2 was constructed at the Assembly Calorimeter [Chow 2004c]. Flashover kitchen

fires were studied in the autumn of 2005. Soyabean oil was used with chemical

composition including iodine, moisture, peroxide and free fatty acid. The fatty acid

composes o f Palmatic acid, Stearic acid, Oleric acid, Linoleic acid, Linolenic acid,

Arachidic acid, Gadoleic acid, Behinic acid, Erucic acid and Lignoceric acid.

126

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Two groups o f six kitchen fire tests were carried out as follows:

Group A arrangement on a hotel Chinese kitchen for staff canteen with two woks and the

arrangement is shown in Figure 7.2 (a). Each wok was filled up with 1000 ml soyabean

oil. Three scenarios were considered:

• Scenario A l :Fire only.

• Scenario A2: Fire with the operation o f water sprinkler system.

• Scenario A3:Fire with the operation o f water mist system.

Group B arrangement on a typical large banquet kitchen with six woks and each wok is

filled up with 1000 ml o f soyabean oil as in Figure 7.2 (b). Three scenarios were

identified:

• Test B 1: All six woks on fire.

• Test B2: All six woks on fire with the operation o f water sprinkler system.

• Test B3: All six woks on fire with the operation o f water mist system.

The model kitchen is o f length 3.6 m, width 2.4 m and depth 2.4 m as shown in Figure

7.3 (a). Flue gas liberated would be extracted by a fan-duct system through the exhaust

hood at a door opening o f width 0.8 m and length 2 m. A stainless steel table long

e n o u g h to se t up s ix w o k s and tw o sin k s w a s con stru cted as in F igu re 7.3 (b). The woks

filled with oil were ignited to give large amount of smoke. Tests were then repeated by

operating the sprinkler system and the water mist fire suppression system. The

operating pressures were 1.12 MPa and 1.26 MPa; and flow rates 5 Is'1 and 0.2 Is'1

127

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 respectively for the two systems. Note that flow rate for the water mist system was only

4% o f that for sprinkler.

7.8 Results on Group A Tests

Two woks filled with soyabean oil as in Figure 7.3 (c) were tested.

Test A l

Smoke was generated when the oil was heated. After about 7 min, the oil was

ignited. A sharp increase in heat release rate is shown in Figure 7.4.

Test A2

After the gas stoves were lit, it took 4 min 55 s and 7 min 54 s respectively for the

cooking oil to be lit at pan A and B. The sprinkler system was turned on after 9

min 50 s. No splashing nor spillage o f oil was observed. The fire was

extinguished after 10 min 1 s. The total burning time was 601 s. The heat

release rate is also shown in Figure 7.4.

Test A3

After the gas stoves were lit, it took about 6 min 13 s for the oil to be ignited. The

gas valve was then shut. After 2.5 min further,the water mist system was

128

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 activated. It took 16 min 7 s for the water mist system to extinguish the fire. The

total burning time was 25 min 20 s. The heat release rate curve is shown in Figure

7.4.

7.9 Results on Group B Tests

All six woks as in Figure 7.3 (d) were burnt.

Test B1

It took 6 min 20 s for fire to be ignited at wok A l, 6 min 35 s at wok A2, 4 min 45 s

at wok A3, 5 min 34 s at wok B l, 5 min 10 s at wok B2 and 6 min 10 s at wok B3.

The fire was extinguished after 9 min 14 s at wok A l, 15 min 59 s at wok A2, 13

min 30 s at wok A3, 8 min 50 s at wok B l, 8 min at wok B2 and 16 min 10 s at wok

B3.

The heat release rate is shown in Figure 7.5.

Test B2

It took 5 min 32 s for fire to be ignited at wok A l, 6 min 20 s at wok A2, 7 min 16 s

at wok A3, 6 min 27 s at wok B l, 6 min 11 s at wok B2 and 6 min 37 s at wok B3.

129

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The water sprinkler system was activated 8 min 7 s after the gas stove was lit. No

splashing nor spillage o f oil was observed as test A2. Water stopped discharging

at 8 min 20 s. The oil was re-ignited and water was turned on again at 8 min 57 s.

The cooking oil fire was completely extinguished at 10 min 50 s. The heat release

rate is shown in Figure 7.5.

Test B3

It took 6 min 14 s for the fire to be ignited at wok A l, 7 min 5 s at wok A2, 7 min 9

s at wok A3, 6 min 27 s at wok B l, 6 min 35 s at wok B2 and 6 min 40 s at wok B3.

The water mist system was activated at 9 min 33 s after the gas stove was lighted.

It took 19 min 3 s to extinguish the oil fire. The heat release rate is shown in

Figure 7.5.

7.10 Discussion

Six tests were conducted in a sample kitchen with two or six woks burning.

On group A arrangement with two woks burning, the environment was quite smoky in

test A l. The maximum heat release rate went up to 0.4 MW. In test A2, the sprinkler

system needed less time, within 20 min, to extinguish the fire for the same setup. In test

A3, the water mist system was effective in extinguishing the fire in about 25 min. It

used only 4% o f time as required by sprinkler.

130

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 On group B arrangement with six woks burning, the maximum heat release rate was up to

1.4 MW. In test B2, the sprinkler system took less time to extinguish the fire in

comparing with test B3 with water mist system. It appears that water mist under the

design condition is not so effective in suppressing bigger fires.

The operation effect o f sprinkler system and water mist system were studied. Both

water systems were effective in extinguishing the fire rapidly. From the above

preliminary tests, it appears that the sprinkler system is efficient in extinguishing the fire.

No splashing nor spillage o f fuel was observed for the woks arrangement in the model

kitchen. However, more water is required.

Calculation on the minimum heat release rate for flashover on our kitchen experimental

set up was done as in Appendix D. In some o f the cases in our experiment, the heat

release rate reached 1.4 MW or 1,400 kW which was higher than the above calculated

minimum heat release rate. Flame was also seen coming out from the test compartment.

Hence, flashover fire was reached in the experimental fire in the model kitchen.

7.11 Conclusions

Six full-scale burning tests on two different groups o f wok arrangements in a model

kitchen were carried out in a facility in a remote area o f Northeast China. The heat

release rate was measured by the oxygen consumption calorimetry. The fire might

reach the flashover stage in some cases. The sprinkler and water mist system are shown

to be effective in extinguishing the fire.

131

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 8 PERFORMANCE OF HEAT RESPONSIVE LINK IN

ACTIVE AND PASSIVE FIRE SYSTEMS

Fusible links are temperature sensitive fire protection devices designed to be part o f a fire

protection system, such as sprinklers, smoke curtains and fire dampers, etc. The system

is activated when the ambient temperature increases to the point that causes the fusible

link to break apart. At the point o f breakage, it releases the pre-load fire protection

device, thus restricting the spread o f fire. In this chapter, the application o f fusible links

in various fire protection systems and the theory involved will be discussed. Various

test methods on the reliability and performance o f fusible links will be outlined.

Simulation tests will also be carried out on fusible links at various temperatures and air

speeds and the results will be reported and analysed. The setup and equipment used in

the heated tunnel tests [Chow and Ho 1990, 1992] were outlined and described.

8.1 Water Sprinklers - Heat Transfer and Empirical Equations

Fusible links [NFPA 1997] have been applied all over building services installations in

such products as pressure tanks, damper doors, fire doors, parts washers, gas pumps,

sprinkler systems, safes and chimneys. In this section, the application o f fusible links in

water sprinkler system and the thermal theory involved will be discussed [Theobald

1987; Theobald and Westley 1988].

132

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The thermal response o f a sprinkler head [Heskestad and Smith 1976, 1980] is usually

described by the nominal release temperature 6„om, the effective operating temperature Qe,

the sprinkler fire constant t and the response time index RTI.

The nominal release temperature 6„om is commonly used in Hong Kong to rate a sprinkler

head [Hartford 1998; Theobald 1987] and is also specified by the manufacturers for Fire

Officers’ Committee approval. This will be the maximum ambient temperature for the

sprinkler head to withstand without opening and is measured by immersing the sprinkler

head into a liquid bath which is heated at a rate o f 0.5 °Cmin'1.

The effective operating temperature Qe is the minimum air-stream temperature required to

activate the sprinkler. The value takes into account the heat loss from the element to the

surroundings. Qe can also be derived graphically from the results measured in the ramp

test. Its value depends on the air velocity since the convective heat transfer coefficient

between the sensing element and air is governed by the air speed [Melinek 1988].

The heat transfer empirical equations are listed in Appendix C.

8.2 Sprinkler Test and Response Time Index

In the above section, the thermal theory and the related empirical equations on sprinkler

heads were reviewed. The principles o f two common sprinkler tests, namely the plunge

test and ramp test, will be introduced in this section. The sprinkler parameter, i.e. the

response time index will be analysed and discussed.

133

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The plunge test is to measure the thermal sensitivity o f sprinkler heads. In this test, the

test sprinkler is suddenly immersed into a flow o f constant temperature and velocity.

The results o f these tests would be related to the expected performance in the field, and

the test environment would be accurately reproduced by any interested party, which

contrasts with the sprinkler sensitivity tests previously used by Underwriters’

Laboratories and Factory Mutual.

In a real fire situation, the sensing element of the sprinkler will be heated by convection

from the hot gases. The effects o f radiation and free convection are insignificant. It

will be acceptable to assume that the temperature is uniform throughout the body o f the

sensing element. It is based on (a) conductive heat exchange between the sensing

element and supportive portions will be negligible and (b) heat required for sprinkler

activation at the activation temperature will be small in comparing to the heat stored in

the sensing element.

In Appendix C, the heat balance equation (C l) to (C7) for the heat sensing element and

framework are reviewed. The plunge test describes the air temperature by a step

function o f time, i.e. n = 0 in equation (C17). From equation (C14) and (C15) [Chow

and Ho 1990; 1992], the product ze' u 2, which is a constant, is related to the thermal

inertia o f the sensing element which is independent on the gas velocity and temperature.

This constant will be called the “response time index” o f the sprinkler head. This

resp o n se tim e in d e x h a s in c lu d ed c o n v e c tiv e h eat transfer and s o w ill b e a property o f th e

sprinkler head itself. The values o f the response time index varies from about 50 m 2 s 2
i_ _[
for Early Suppression Fast Response (ESFR) sprinkler heads to about 300 m 2s 2for

134

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 ordinary ones. However, the plunge test is considered not so suitable in studying

thermal responses o f sprinkler heads since heated air at constant speed and temperature

will be used [Chow and Ho 1990].

Another sprinkler test, the ramp test, describes a constantly increasing air temperature,

i.e. n = 1 in equation ( C l7). Solving equation (C4) to (C7) and the equation for the

ramp test will be shown in equation (C23). Developing the equations and from equation

(C25), by plotting the graph o f de against u \ a straight line will be shown.

A number o f points are needed to be considered in the above two tests. Firstly, the heat

transfer to the sprinkler is mostly by convection instead o f radiation during the early

stages o f fire. Secondly, conduction heat loss is needed to be considered as the

pipework is taken as a large heat sink. Lastly, higher values o f air speed are not

recommended and air velocities between 1 to 4 m s'1 have been found in the ceiling jet

during the early stages o f fire.

8.3 Fire Dampers in Fire Protection Systems

Fusible links are used in fire protection devices [Cote 1997] locally in Hong Kong,

including fire dampers, fire doors and sprinklers. Fire dampers and ceiling dampers will

b e in sta lled to p rotect o p e n in g s for d ucts in w a lls and flo o r s w ith appropriate fire

resistance ratings. Air transfer openings should also be protected. All these dampers

should be classified and labelled in accordance with UL standard 555.

135

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 A smoke damper can be used for either traditional smoke management or smoke control.

For smoke management in a typical ventilation system in hotels, a smoke damper inhibits

the passage o f smoke under the forces o f buoyancy, stack effect, and wind. However,

smoke dampers are only one o f many elements intended to inhibit smoke flow. In

smoke management applications, the leakage characteristics o f smoke dampers should be

selected so that they may be functionally appropriate with the leakage o f the other system

elements.

In a typical smoke control system, a fire and smoke damper inhibits the passage o f air

that may or may not contain smoke. A damper does not need low leakage

characteristics when outside air is on the high pressure side o f the damper, as is the case

for dampers that shut off supply air from a smoke zone or that shut off exhaust air from a

non-smoke zone. In these cases, moderate leakage of smoke-free air through the

damper does not adversely affect the control o f smoke movement. It is the best to

design smoke control system so that only smoke-free air is on the high-pressure side o f a

closed smoke damper.

Smoke dampers [ASHRAE 1997] should be classified and listed in accordance with UL

standard 555S for temperature, leakage, and operating velocity. Many FSD approved

type fire and smoke dampers are used in local hotels and site tested by the local the Fire

Services Department. The velocity rating o f a smoke damper is the velocity at which

the actuator will open and close the damper.

At locations requiring both smoke and fire dampers, combination o f dampers meeting the

requirements o f both UL standard 555 and UL standard 555S can be used. The

136

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 combination o f fire/smoke dampers must close when they reach their US standard 555S

temperature rating to maintain the integrity o f the firewall.

Fire, ceiling, and smoke dampers should be installed in accordance with the

manufacturers’ instructions. NFPA standard 90A gives general guidelines regarding the

locations requiring these dampers.

The supply and return/smoke dampers should be a minimum o f class II leakage at 120 °C.

The return air damper should be a minimum o f class I leakage at 120 °C to prevent

recirculation o f smoke exhaust. The operating velocity o f the dampers should be

evaluated when the dampers are in smoke control mode.

The exhaust ductwork and fan must be designed to handle the temperature o f the exhaust

smoke. The temperature o f the exhaust smoke can be lowered by making the smoke

control zones large or by pressurizing only the zones adjacent to the fire zone and leaving

all the other zones operating normally.

The odour in the hotel is one o f the first impressions the traveller has on arrival. The

odour in the hotel entry hall, at the front desk while the check-in formalities are taking

place, that o f the guestroom and finally o f the restaurant will strongly influence the

guest’s perception o f well-being.

With well designed ventilation system, the odour occurrence can be minimized but

maintaining fire system integrity is also important. Odour is at once associated with the

cleanliness o f one’s surroundings and the corresponding feeling o f well-being. The

137

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 odours which float in the air may have numerous origins: they include cigarette smoke,

body odours, cooking odours, paint odours, the odours o f certain cleaning products, etc.

These odours generally disappear when natural or artificial ventilation is used. In some

cases, natural ventilation is not sufficient, e.g. in the kitchens and restaurants, and it is

then necessary to resort to mechanical ventilation.

As soon as the volume o f air necessary for the ventilation o f the hotel is below 4 to 6 air

change, the air is confined, odour settled, and the guest feels uncomfortable. The

hotelier, conscious o f the real danger o f losing the guests, then takes steps to ventilate the

premises. The danger o f bad odour can be avoided by ensuring that the correct quantity

o f fresh air is distributed judiciously throughout the hotel. The hotelier may also

consider that different parts o f the hotel are not always occupied by a constant number of

people. By definition, occupation is variable in hotels. What happens when there are

fewer persons than anticipated? What if, for example, the “Grand ballroom” is occupied

by 200 persons instead o f 1200 which it can hold? It will then be necessary to adapt the

volume o f air to the number o f people actually present, each o f whom will require

minimum quantity o f air.

8.4 Heated Wind Tunnel - Experimental Setup

In order to carry out tests [Chow and Ho 1990, 1992] for studying thermal responses of

fusible links under different heating conditions, a hot air tunnel, known as the heated

wind tunnel (Figure 8.1) has been constructed and located at the ground floor laboratory

o f the Building Services Department o f The Hong Kong Polytechnic University. With

138

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 it, different values o f hot air speed and temperature can be adjusted to simulate a

fire-induced ceiling jet. The heated wind tunnel is constructed as shown in Figure 8.1.

The appearance o f the tunnel is similar to the one at the Fire Research Station [Chow and

Ho 1992] but with a different heating and control design.

In the heated wind tunnel, air is blown by a fan positioned at the inlet. The inlet air

stream passes through a heating section which is composed o f finned air heater. The

heated air then passes through a contraction and finally to the working section at the

outlet duct. The tunnel is basically a constant-mass-flow open-ended device. The inlet

air temperature remains in a limited range during the tests. For a particular fan speed,

the mass flow remains practically constant in the heated section. The contraction before

the working section would smoothen the air flow by reducing turbulence. An

acceptable uniform condition o f temperature and velocity can then be provided over the

working cross-section.

The dimensions o f the tunnel are 3 m long, 1.6 m high and 0.7 m wide. It is made from

1.2 mm mild steel sheet and its low thermal mass enables the tunnel air temperature to

respond rapidly to any programmed changes. The working section is wooden made and

without insulation. The relatively high thermal mass o f wood allows the surface

temperature o f the working section to be cold with respect to the fusible link temperature.

It also ensures that heat transfer to the fusible link is predominant by convection rather

than radiation from the working section.

The equipment includes a belt-drive centrifugal fan (Figure 8.1) o f which the speed is

controlled by a frequency invertor. The air mass flow is determined by the speed o f the

139

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 fan and can be controlled manually by the invertor. Remote control is allowed if the

invertor is connected to a source signal. The heating section consists o f finned air

heaters with a rating o f 15 kW. The heat output is controlled by a PID controller which

senses the air temperature in the working section. Manual override is also possible by a

remote source signal.

The fusible link to be tested is fitted to a specially made metallic fitting as shown in the

photo. Linear rates o f rise o f temperature varying between 10 to 60 “Cmin'1. may be

maintained up to a maximum o f 150 °C in the working section o f the tunnel. Smaller

rates of rises are also possible but linearity might not be maintained. At lower velocities

(below 1ms'1), the maximum temperature can be increased to 200 °C but sustained

operation at this temperature will activate a high temperature cut-out control. The time

taken for getting a certain tunnel air temperature is about 10 s for medium air velocities.

A longer time, about 30 s is required for lower velocities.

The air velocity in the working section may be varied from 1 to 15 m s'1 at 25 °C. Lower

values can be achieved by reversing the polarity o f the fan motor to obtain 0.2 to 1.5 m s'1.

The characteristic relationship between air velocity and the frequency output o f the

invertor can then be found. The tunnel air velocity can then be set to any required

operating conditions.

There is a problem in controlling the air temperature at low air speed. A time delay is

found and so giving a temperature fluctuation o f 20 to 30 °C. The problem is solved by

manual operation and the air temperature can then be set within ± 5 °C of the desired

value. Calibration is necessary for every test in order to obtain the desired temperature.

140

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Also, the rate o f temperature has to be determined graphically by measuring the time

required to get maximum heat output.

The heated wind tunnel is similar to the one constructed at the Fire Research Station

(FRS) in the sense o f getting controlled speed and temperature o f hot air. But the

equipments and their performances are different as shown in Table 8.1.

The air flow rate o f the FRS wind tunnel is controlled by changing the iris diameter.

The mass flow rate is determined using the pressure difference between the iris. Air

velocity profile at the working section is measured using pitot static tube and checked by

a vane-type anenometer. This design is applicable for test velocities 1 ms"1 or above.

However, when the air velocity lies below 1 ms"1, buoyancy effect o f the hot gas may

affect the uniform distribution o f temperature and velocity in the working section. This

would create a higher temperature at the upper part o f the section. In this case,

vane-type anenometer may not be a suitable device for measuring low air velocity.

For the heated wind tunnel, a frequency inverter is used to drive the fan. This variable

speed device is possible to change the speed of the fan by 0.2% increments (0.1 Hz), thus

providing an accurate air flow rate. The air velocity at the working section is measured

using pitot static tube and a pressure transducer. This is similar to the one in the FRS

tunnel. However, it is not applicable to choose the mean velocity o f the working section

as a reference value. This is because the air velocity at the top o f the working section,

where the sprinkler is located, may differ greatly from the mean value. The situation is

especially obvious when the air temperature is high and the air velocity is low.

141

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Therefore, a point velocity at the fusible link location, rather than a mean velocity, should

be chosen.

As the frequency inverter can accept remote control signals, it is possible to use

microcomputer control technique to create variable velocity profile during tests. This

would provide greater flexibility, easier modifications, and higher accuracies.

A programmable controller was used to control air temperature in the FRS wind tunnel.

In contrast to it, the one we used adopts a PID controller with step and current value

output. Due to the heater and sensor time delays, difficulties are encountered in

temperature control. But the problem can be overcome by switching to manual mode.

Future modifications will increase the sensitivity and relate the relationships between air

temperature, velocity and heater output by programmable controllers. The heated wind

tunnel is a reliable instrument in testing fusible links under various temperatures and air

speeds and is also adopted by the local Fire Services Department. First o f all, the code

UL33 on heat response link used in the fire protection device, was reviewed.

8.5 H eat Responsive L in k for Fire-Protection Service - UL33

UL33 [UL 2005] cover the requirements for heat responsive link used for fire-protection

service. These links consist o f devices intended for installation under load conditions

such as for use with automatic suppression systems; or automatic closure devices for

doors, windows, dampers or smoke vents. Heat responsive links are categorized by

142

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 temperature rating, type o f coating and plating, minimum and maximum design load, and

other factors which may have a bearing on their intended use.

Various types o f heat responsive links:

• Bi-metallic type - Two different metals mechanically fastened together. Each

metal has different expansion characteristics which act to release a load when

exposed to sufficient heat. This type o f device may be reused.

• Frangible bulb type - Liquid-filled glass bulb that features a load when exposed to

sufficient heat. For example, water sprinkler system.

• Fusible type - Two or more metallic parts having a solder element that fuses to

release a load when exposed to sufficient head. This is similar to the fusible links

used at the fire dampers o f ventilation system.

A heat responsive link shall be constructed for service where the maximum normal

ambient air temperature at the part o f installation does not exceed that in Table 8.2.

The operation characteristics o f heat responsive link shall not be impaired by the

application o f any applied coating or plating when tested in accordance with these

requirements. A corrosion resistant coating or plating shall be uniformly applied. A

wax coating shall not be brittle when new nor become brittle with age. A wax coating

shall not crack when tested at the maximum temperature for which the link assembly may

be installed.

143

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 A link assembly shall support a load equal to five times the maximum design load for at

least 150 hours. At least ten sample link assemblies o f the lowest temperature rating in

the Ordinary Temperature classification (or if ordinary temperature classification links

are not produced, then those produced having the lowest temperature rating) are to be

loaded to five times the manufacture’s maximum design load for 150 hours at an ambient

temperature o f 70 ± 5 °F (21 ± 3 °C).

As an alternative, a heat responsive link assembly shall support a load for one minute

when tested as below. At least ten sample links assemblies o f the lowest temperature

rating in the ordinary temperature classification are to be loaded to five times the

manufacturer’s maximum design load for one minute at an ambient temperature of 70 ± 5

°F (21 ± 3 °C). One attachment end o f each sample is to be connected to the test

supporting apparatus. The test load is to be applied to the other attachment end in the

intended direction.

8.5.1 Operating temperature (bath) test

The operating temperature o f heat responsive links [UL 2005], when bath tested, shall

operate within the range having a maximum temperature not in excess o f 5 °C or 107

percent o f the minimum centigrade temperature o f the range, whichever is greater. For

the purpose o f this determination, the marked temperature rating is to be included as one

o f the ranged values, making a total o f 11 values in the range. The previous temperature

ratings in Table 8.2 indicate all the link types.

144

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Link operation for this test includes the intended functioning o f the eutectic elements of

any rupture o f the frangible bulb heat responsive element. If partial fracture o f the

frangible bulb in the liquid environment occurs which does not result in link operation,

the temperature at which the bulb fracture occurred shall be considered the operating

temperature. Not less than ten heat responsive links o f either coated or uncoated types

o f each temperature rating are to be subjected to this test. All ten samples shall comply

with the requirements as specified above.

The fusible links are to be placed in an upright position under a load equal to the

manufacturer’s minimum design load, but not less than 4 N, and completely immersed in

the water or oil bath. The vessel is to be provided with a source for heating the liquid at

the prescribed rate and with means to agitate the liquid and measure the temperature of

the liquid bath. An agitator is to be used as an aid in obtaining uniformity in

temperature o f the liquid bath.

A laboratory mercury thermometer, calibrated in accordance with the specifications for,

ASTM E l, is to be used to determine the temperatures of the liquids used in bath tests.

The bulb o f a thermometer is to be held level with the fusible link by a support member.

The temperature o f the bath liquid is to be raised until the liquid is within 11 °C o f the

temperature rating o f the device for a device having a temperature rating o f 149 °C or

less, and within 17 °C for a device having a temperature rating o f 163 °C and higher.

The temperature rise then is to be controlled at a rate not exceeding 0.5 °C per minute

until operation or until a temperature o f 11 °C above the rated temperature is reached.

The temperature o f the liquid and the time o f operation, as each heat responsive link

operates, are to be recorded.

145

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 In previous chapter, the site test method by local Fire Services Department were

described and outlined. Recently, there was one hotel project in Hong Kong where

actually the FSD had carried out the site bath test on 20 sample links separately selected

from two hotels. The test was satisfactory and FSD approved the use o f the links in the

two hotels.

8.5.2 Sensitivity-oven heat test

A fusible link shall have the following operating time characteristics [UL 2005] when

tested in the sensitivity test oven in accordance with UL33. For a standard response

link, each sample shall have a maximum operating time as specified in Table 8.3. If the

link temperature rating is not shown in Table 8.3, the maximum operating time for each

sample is to be determined by using the formula, based on a RTI value o f 350 (m 2 s 2)

and the marked temperature rating o f the link.

The mean operating time shall be equal to or less than a 1.3 times o f the mean operating

time o f the link. Ten samples of fusible links o f end type are to be placed in the

sensitivity test over with the heat responsive element located at least 1 inch (25.4 mm)

away from the inside surfaces o f the oven and with the broadest part o f the link, or with

the heat responsive element o f the link, facing toward the air flow. The samples are to

be at the minimum load specified by the manufacturer.

The samples are to be conditioned at 24 ± 1 °C for at least 2 hours. The samples are

then to be quickly plunged into the sensitivity test oven in the vertical position. Each

146

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 link is to be observed to determine if operation occurs within the time as specified in the

above table.

The sensitivity test oven is to consist o f an 8 inches square stainless steel chamber. A

constant air velocity o f 2.54 ± 0.01 ms"1 and an air temperature as specified in Table 8.4

for each temperature rating are to be established.

The required link operating time values as specified in the above tables shall be

calculated by using the following equation relating the response time index RTI (in

m 2 s 2 ), the gas temperature at the orifice plate Te (in °C), the nominal gas temperature in

test section Tg (in °C), the marked temperature rating o f the link Tm (in °C), Nominal

ambient air temperature Tu (24 °C) and air speed in wind tunnel section Fg(in m s'1) with

operating time o f the link t0 (s):

t,„ _ R i—
T I iIn J Tm - T u (8.1)
V, T8 - T * J,

where Vg is 2.54 ms"1

Fusible links shall be marked with the temperature rating, a distinctive type or model

designation, and the manufacturer’s or private labeller’s name or identifying symbol.

The year o f manufacture shall be stamped or cast on a visible area o f the heat responsive

link. Links produced in the last three months o f a calendar day may be marked with the

following year as the date o f manufacture, and those produced in the first three months of

a calendar year may be marked with the preceding year as the date o f manufacture. If a

147

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 manufacturer produces links at more than one factory, each unit shall have a distinctive

marking to identify it as the product o f a particular factory.

In the following section, several tests on local made fusible links were conducted to

review their functional performance.

8.6 Local Fusible Links —Simulation Tests with Heated Wind Tunnel

Tests [Chow and Ho 1990; 1992] were also carried out on local fusible links currently

used in local hotel projects. By setting up the heated wind tunnel in The Hong Kong

Polytechnic University (Figure 8.1 (a) to (g)), the first batch o f 50 fusible links (Figure

8.1 (h) to (k)) were tested in accordance with UL33 at gas temperature 2.54 m s'1 and 135

°C.

The fusible link actuation times were recorded as in Table 8.5 and plotted in Figure 8.2

by a bar chart.

For the 45 fusible link samples successful broken in the wind tunnel, they all satisfy

Table 10.1 o f UL33, for breakage time to be less than 86.1 s. The five fusible links,

which did not break, might be defective or oxidized. This further confirmed the

previous on site testing by FSD using the water bath method. The results show that the

fusible links perform and operate in accordance with the requirements in UL33.

148

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Moreover, tests on 180 fusible links were carried out using the Building Services

Engineering heated wind tunnel, at air speed from 0.6 to 3 m s'1 and air temperature from

80 to 150 °C. The results are shown in Table 8.6. A graph, fusible link actuation time
1 /
a g a in s t is plotted as shown in Figure 8.3. The slope, which is the response
v s

time index (RTI) for the fusible link, is found to be 343 m 2 s 2 as in Figure 8.3. This

result matches with the general RTI as measured for sprinkler heads.

Furthermore, it can be seen that water bath tests adopted in UL33 or local fire services

department, may not be appropriate for testing fusible links. The fusible links will be

exposed to hot smoke or heated air, instead o f heated water as used in the water bath test.

Currently, water bath test on local fusible links is adopted by local fire services

department and may not be suitable as water will have a much higher specific heat

capacity than actual hot smoke. Hence, alternative method by using heated wind tunnel

tests is proposed in this study for testing local fusible links. Simulation experiments

carried out for rating sprinkler heads, involved in the testing o f fusible links at various

temperatures and air speeds were conducted in this study. The results in Figure 8.3 are

concluded in the accurate and important evaluation on the RTI. The results o f which can

be easily used and widely applied in the local fire industry.

149

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 CHAPTER 9 CONCLUSION

Fire safety in hotel premises was focused on in this thesis. Critical areas, such as

kitchens, will give safety problems. The importance of carrying out full-scale burning

tests for kitchens was pointed out such that better fire protection strategies and systems

might be worked out through evaluating the results. The performance o f fusible links in

active and passive fire systems was particularly pinpointed and tests to verify its

reliability and operation were conducted.

Full-scale burning tests in a model kitchen were then carried out to investigate the heat

release rate when a wok was burnt in the model kitchen. The operation o f water sprinkler

and water mist system were also investigated. The application of fusible links in local

ventilating system o f hotels has led to many safety concerns on its reliability and

performance. Tests were also conducted in a heated wind funnel to find out the

operation o f these locally made fusible links under various air temperatures and speeds.

The data are useful to local authorities and professionals in assessing the operation,

performance and application o f fusible links at fire dampers, sprinklers and smoke

shutters in buildings. A conclusion summary on the thesis is as follows:

The profile o f the study was described in Chapter 1 with the focus on fire safety in hotels.

Hotel fires were studied and investigated in Chapter 2. The problems and all the

param eters related to fire sa fe ty in a h o tel w ere id en tified . In th e m e n tio n ed e x a m p le s

o f fires in hotels, large crowd o f guests gathering, guests’ being unfamiliar with the hotel

configurations, and incorrect operation and misuse o f facilities were the problems. The

planning and proper maintenance o f all the sophisticated building services installations in

150

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 a hotel are critical as they nearly account for 50% o f the total hotel construction project

cost. Moreover, the importance o f fire safety in international chain hotels in Hong Kong

was outlined. These chain hotels have very detailed and planned fire safety strategies

and management programs, both on prevention and implementation. They had also

summarized in their fire safety plans all their experience and actual operations towards

fire problems and correct management procedures during an actual fire. Furthermore,

the kitchen was singled out as a critical area in the hotel. Fire safety issues and

problems in kitchens were discussed by focusing on the type o f utilities used and the

complex applications o f electrical and mechanical installations in the kitchen. By

following recent local fire codes, proper fire safety and protection system could not be

concluded and hence, this chapter actually leads us to the urgency o f carrying out more

intense study on kitchen fires by full-scale burning tests to understand the heat release

rate when a model kitchen is burnt. Furthermore, with so many ventilation systems

involved in the kitchen, the performance of fusible links are crucial to maintain the

integrity o f the passive fire provision or even the activation o f the active fire systems.

The total fire safety concept was studied in depth in Chapter 3. The goals for fire safety

are human life safety, protection o f premises and fire prevention in a hotel. The MoE,

FRC, MoA, FSI and FSM are needed to support the above three goals. It is concluded

that the adoption o f the total fire safety concept is important in the evaluation o f the fire

safety level o f the hotel building. MoE, FRC, MoA, FSI and FSM components must not

be assessed separately. All these key parts are related. A total fire safety strategy

should be focused on so as to use fire safety management as the software component to

control the hardware components, such as MoE, FRC, MoA and FSI, etc. Throughout

this chapter and from the hotel operation experience o f the author in recent years, fire

151

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 safety management is concluded as a main contributing factor towards fire safety in a

hotel.

Also in Chapter 3, the essential components of passive fire protection systems in Hong

Kong were pointed out. These components, namely the MoE, FRC and MoA, are

essential parts for building fire safety. Nowadays, prescriptive codes are used for

implementing the design o f these passive fire protection systems. However, as mentioned

in the study that with so many new architectural building features, a more scientific

method might be required and more research works are needed, particularly on the fire

behaviour o f materials and evacuation modelling. It is concluded that the adoption of

performance-based fire codes and engineering approach towards fire protection design in

buildings are highly advocated. The housekeeping and planned maintenance o f all these

passive protection systems in hotels are particularly pinpointed. The best passive fire

protection system will not function to its design intent without proper care, repair and

service follow-up.

The active fire protection systems for hotels were discussed in Chapter 4. Different

active fire protection systems were reviewed, including sprinklers, water mist, hydrants

and hosereels, detection systems and emergency lighting, etc. These active systems are

essential to the building and its occupants by limiting the fire spread, suppressing the fire

and preventing structural collapse of the building. Again, like passive fire protection

systems, prescriptive fire codes cannot only be relied on in providing better design and

safer buildings. It is concluded not to decide blindly on building design in choosing the

suitable system, adequate supportive data and considerations must be provided to see

whether the proposed active systems are feasible for the hotel building or not. More

152

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 research works are required on studying the active fire system performance and

characteristics towards controlling and limiting fire spread. Local fire codes and

licensing requirements were reviewed and compared with those codes specified in

foreign countries.

The concept o f fire safety management and its application in hotels were studied in depth

in Chapter 5. The fire safety plan, maintenance plan and the various normal and

emergency mode o f operation were reviewed. The importance o f fire safety

management was concluded as a very important software component in the daily

operation or long-term fire safety planning in a hotel. On the other hand, the importance

o f engineering approach in the application o f fire safety management was also pointed

out. Various policies and procedures were listed for an emergency fire plan

implemented in a local five-star hotel in Hong Kong. Hotel fire safety environment is

comprised o f and effected by a lot o f parameters and simply the provision of all the

hardware components may not effectively ensure the fire safety to guests without the

simultaneous implementation o f the fire safety management plan, the software

component. From the various fire incidents mentioned in previous chapters, the

importance o f fire safety management in hotel operations and the critical need to provide

full training on fire safety to all hotel employees are concluded.

Different developments in calorimetry and the establishment o f the Chinese calorimetry

for the actual full-scale burning tests for the model kitchen in this study were reviewed in

Chapter 7. On cost and environmental concerns, it is difficult to perform the full-scale

burning tests in Hong Kong and so a site was established in Northern China.

153

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 At the same time, the necessity o f carrying out full-scale burning tests was concluded in

fire safety engineering. Actually, heat release rate is a very important parameter in fire

hazard assessment. By understanding it, information can be obtained on the possibility

o f flashover during fire and whether the items adjacent to the burning object will be

ignited. Moreover, the smoke layer temperature, radiative heat flux and whether there is

an upward spread o f frame over walls can be found out. Furthermore, the heat release

rate results will be useful in recommending the use o f finishes combustibles and

eventually the fire safety provisions in kitchens.

For the six burning tests carried out in the model kitchen, in some cases, the fire could

reach a flashover temperature o f 600 °C and the heat release rate o f 1.4 MW. The

operation o f the sprinkler and water mist systems were effective in extinguishing the oil

fire. It was found that the water sprinkler system extinguished the fire more rapidly in

the burning tests. To provide better fire service protection and to arrive at a better

suppression system, further full-scale burning tests under different kitchen setups and

configurations are recommended. The results will surely tell more on what will happen

in an actual kitchen fire incident.

The use o f fusible links in both active and passive fire protection systems was studied in

depth in Chapter 8. The theory behind the heat transfer for the operation o f sprinkler

heads or fusible links were reviewed. Comparing the use o f water bath tests accepted by

the local Fire Services Department on locally made fusible links, the testing requirements

for fusible links in UL33 were also listed. Moreover, the heated wind tunnel was

constructed in The Hong Kong Polytechnic University and used for carrying out the

fusible link actuation simulation tests for a large sample (over 200) o f fusible links at

154

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 various air temperatures and speeds. For the initial test on 50 fusible links in

accordance with UL33, it further concluded its compliance with the results of the local

site tests on the 20 samples o f fusible link selected randomly on site. In the second test

on about 170 fusible links at various air temperatures and speeds, the response time index

was calculated, and the graph for the fusible link performance was plotted. Further tests

on fusible links are recommended for different ratings in order to suit its actual

application in various uses in fire services installations. This area o f tests is not

receiving much attention in actual engineering field application.

To conclude for improving the fire safety for hotels, it is recommended to adopt the total

fire safety concept on planning the fire safety provisions for hotels. By doing so,

performance based fire code can be used which will be more appropriate and economical

for providing the active and passive fire protection systems. In particular, fire models

and full-scale burning tests can also be used for simulating the behaviour of hotel fire

under various scenarios and the determination o f the design fire size. The use o f fire

safety management concept is highly recommended as a useful tool on approaching the

fire safety problem. It is concluded that the hotel management must have a proper fire

safety management program, a software component to fully relate and combine the

operational effects o f both the active and passive fire protection system in any special fire

problem in big hotels.

155

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 TABLES

Table 6.1: Requirements on escape routes in china codes

Maximum distance from room door to exit or stairs (m)

Rooms on the side/end o f the

Rooms between two exits/stairs
Type o f occupancy corridor

Fireproof class Fireproof class

I, II III IV I, II III IV

Child-care,
25 20 - 20 15 -
nurseries

Hospitals,
35 30 20 15 - -
sanatoriums

Schools 35 30 - 22 20 -

Others 40 35 25 22 20 15

T-l

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Table 6.2: Comparison on the egress capacity

China code NFPA life safety code

Capacity Uses Stairways Level

requirements (cm/person) components
(m/100 person) and ramps
(cm/person) (cm/person)

Fireproof class I, II III IV Board and care 1.0 0.5

Stories 1,2 0.65 0.75 1.00 Health care (sprinkler) 0.8 0.5

3 0.75 1.00 - High hazard contents 1.8 1.0

4 1.00 1.25 - All others 0.8 0.5

T-2

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Table 6.3: Travel distance limits in NFPA life safety code

Travel distance limit

Type o f occupancy
Unsprinklered (m) Sprinklered (m)

Assembly 45 60
Educational 45 60

Health care 45 60

Detention and correctional 45 60

Residential 53 99

Mercantile 30 60

Business 60 91

Industrial 60 75
Storage Low hazard Not required Not required

Ordinary hazard 60 122

High hazard 23 23

T-3

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Table 8.1: Difference in heated wind tunnel setup

No. Items FRS wind tunnel Heated wind tunnel in BSE

o f PolyU

1 Tunnel body 1.6 mm aluminium sheet 1.2 mm mild steel with

with insulation. insulation.

2 Fan:
• Type Axial Centrifugal
• Speed Constant Variable
• Flow control Iris-type variable shutter Frequency inverter
mechanism

3 Heater 18 kW convection heater 15 kW finned air heaters

4 Temperature Programmable digital PID controller

controller controller

T-4

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Table 8.2: Heat responsive link - rating

Temperature Temperature rating Maximum ambient temperature

classification °F
°F °C °C

Low 1 2 5 -1 3 0 5 1 -5 4 90 32

Ordinary 1 3 5 -1 7 0 5 7 -7 7 100 38

Intermediate 1 7 0 -2 2 5 7 9 -1 0 7 150 66

High 250 - 300 121 - 149 225 107

Extra high 325 - 375 1 6 3 -1 9 1 300 149

Very extra high 400 - 475 204 - 246 375 191

Ultra high 500 - 575 260 - 302 475 246

T-5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Table 8.3: Operating time for links in sensitivity-oven heat test

Operating time for links in sensitivity —oven heat test

Temperature rating (°C) Oven temperature (°C) Standard response type (s)

57.2 135 78.0

60.0 135 86.1

68.3 135 86.1

71.1 135 121.3

73.9 135 131.1

79.4 197 84.8

93.3 197 112.4

100.0 197 127.1

104.4 197 137.3

T-6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Table 8.4: Sensitivity oven temperature in sensitivity-oven heat test

Sensitivity oven temperature

Temperature rating Oven temperature

°F °C °F± 2 °C± 1

135-170 5 7 -7 7 275 135

1 7 5 -2 2 5 7 9 -1 0 7 386 197

250 - 300 121 - 149 555 290

325 - 375 1 6 3 -191 765 407

4 0 0 -4 7 5 204 - 246 765 407

5 0 0 -5 7 5 260 - 302 765 407

T-7

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Table 8.5: Fusible link test result (activation time) at air speed of 2.54ms*1 of 135 °C

Sample No. Time for fusible link Sample No. Time for fusible link
breakage (s) breakage (s)
1 45.9 26 29.1
2 41.5 27 28.7
3 42.2 28 34.1
4 28.8 29 29.6
5 35 30 34.5
6 33.2 31 34.3
7 39.6 32 28
8 37.8 33 30.3
9 31.5 34 31
10 34.8 35 29.9
11 33.4 36 36.3
12 35 37 32.4
13 35.8 38 33.6
14 33 39 32.4
15 33.4 40 360, did not break
16 30.8 51 25.5
17 31.9 42 26.1
18 31.4 43 30.8
19 30.6 44 28
20 33.8 45 30.4
21 29.2 46 30.3
22 31.6 47 360, did not break
23 31.3 48 360, did not break
24 32.1 49 360, did not break
25 35.6 50 360, did not break

T -8

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Table 8.6: Fusible link test result (activation time) at various air speeds and

temperatures

Velocity Temperature (°C)

( m s 1) 80 85 90 95 100 110 120 130 140 150

NA NA NA 593.84 661.60 296.47 211.66 176.66 115.25 100.13

0.6 NA NA NA 513.44 406.47 391.19 170.78 206.50 130.97 127.97
NA NA NA 569.29 510.34 307.28 243.22 195.88 159.40 117.47
1223.66 831.00 707.00 322.82 277.31 204.44 141.72 106.19 94.06 86.63
1.0 1206.69 578.47 677.00 307.84 385.88 142.90 136.69 74.69 132.03 76.53
1001.16 623.31 563.00 356.28 290.35 183.97 133.75 117.22 90.78 92.85
682.72 548.13 408.81 233.94 253.28 120.78 105.03 60.72 66.72 71.44
1.5 609.91 405.85 407.60 211.03 224.38 146.09 105.31 55.03 95.41 67.22
924.37 431.82 334.84 291.37 193.63 154.31 127.04 50.78 101.19 66.41
419.18 241.87 200.41 184.63 163.72 110.82 118.32 66.00 54.35 53.43
2.0 391.93 328.19 223.07 190.41 154.37 116.87 82.03 57.44 52.31 51.96
265.56 256.13 207.93 170.59 144.44 110.69 132.63 67.47 54.37 46.43
335.47 221.72 188.41 114.36 136.25 88.50 98.62 49.97 44.42 37.31
2.5 292.06 245.31 150.01 184.01 106.84 85.28 88.57 39.93 44.33 39.81
285.41 211.00 146.32 120.31 106.75 82.12 75.03 40.41 53.37 36.91
123.25 92.94 70.57 64.91 54.88 44.53 40.75 33.25 29.62 31.09
3.0 126.72 86.72 76.22 68.72 51.84 48.84 35.09 31.81 28.60 29.07
138.78 84.59 64.34 59.25 47.06 42.65 34.15 29.75 25.72 19.00

T-9

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 FIGURES

Figure 7.1: Burning tests of kitchen fires

F-l

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 (j) (k) (1)

(P) (q) (r)

Figure 7.1: Burning tests of kitchen fires

F-2

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 (V ) (w) (X )

(y) (z)

Figure 7.1: Burning tests of kitchen fires

F-3

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 (a) Staff canteen kitchen (b) Banquet kitchen with 6 woks

Figure 7.2: Hotel Chinese restaurant kitchens tested

F-4

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Exhaust gases

2.4 m
Pin

Path

D oor

Top view

(a) The room (c) Scenario A

A3 B3

WZi
Tan A2
B2

n
500 500

800 Pan A1

Front view Door

Top view

(b) Front view (d) Scenario B

Figure 7.3: Model kitchen

F-5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 400
A1
Heat release rate / kW

300
A3

200

A2

100

0 200 400 600 800 1000 1200

Time / s

Figure 7.4: Heat release rates of testing group A on a staff canteen kitchen arrangement

F-6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 1400

B1
1200
Heat release rate / kW

1000

800

B3
600

400

200
B2

0
0 200 400 600 800 1000 1200
Time / s

Figure 7.5: Heat release rates of testing group B on a banquet kitchen arrangement

F-7

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 (a) (b) (c)

(d) (e) (f>

(g) (h) (i)

G) (k)

Figure 8.1: Heated wind tunnel test for fusible links

F-8

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Sample Number 16

Air Temperature 135 °C

Time of Operation / s

Figure 8.2: Activation times of fusible links

F-9

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 1400
Testing data
1200
Fitting curve

1000

800
Actuation Time

Slope: 343m 2s
600

400

200
• •

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

In X i z!
— — / m 2s 2

Figure 8.3: Activation times of fusible links at various air speeds and temperatures

F-10

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 APPENDIX A MAJOR TASKS OF THE FIRE SAFETY

MANAGER

For daily task or normal mode o f operation includes:

• To manage the maintenance and monitor the equipment performance;

• To cany out routine maintenance and testing o f all emergency equipment;

• To ensure that all existing or new sets o f codes or regulation are being satisfied;

• To control the building contents and keep dangerous good within licence limits;

• To maintain escape route and practice good housekeeping;

• To record all near miss fire incidents and happenings;

• To maintain and keep the emergency plan, in particular the evacuation plan;

• To communicate, liaise and have a working relationship with local and external fire

brigade;

• To have plans on fire drills and train staff at all disciplines;

• To have continuous safety systems reviews and risk assessment, especially in case o f

renovations or refurbishment planned;

• To organize and plan independent safety audits and inspections;

• To maintain the access for the disabled and their escape during emergency;

• To have strict fire safety practices to be followed by contractor working in the hotel;

• To monitor and control the refurbishment in the hotel and its effect on fire safety;

• To record any building changes that may affect fire safety and evacuation; and

• To assess potential environmental impact o f fire.

AA-1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 However, in case o f fire and in emergency mode [Chow and Lui, 2001], the fire safety

manager should:

• Initiate first aid fire fighting in case site conditions allow;

• Communicate and manage and assist local fire brigade in fire fighting;

• Manage all the control room operations, including fire, security and engineering, etc.;

• Initiate and ensure that all the active fire protection systems are working;

• Switch off non-essential equipment;

• Control the crowd, mastering occupants and assist in organizing evacuation;

• Take care o f the disabled occupants in case o f evacuation; and

• Record the whole emergency process and rectify any actions that might disrupt the

efficiency during the emergency mode.

AA-2

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 A P P E N D IX B F IR E S A F E T Y M A N A G E M E N T P R O G R A M IN A

HOTEL

In this section, the elements that will form an effective fire safety management program

in a hotel will be outlined. Moreover, it will be stated in an organized and orderly way

on how to develop an effective hotel fire safety management program so as to ensure

human life and asset property safety. First o f all, how a leading international chain hotel

in Hong Kong develops and implements their fire emergency and fire safety management

program will be outlined below.

The general fire instruction for all hotel staff is to keep calm and do not panic. Panic

will certainly lead to disaster and remaining calm will certainly save lives. Any staff

member discovering a fire or seeing any smoke or smelling burning should:

a. Raise the alarm immediately by breaking the glass o f the nearest alarm.

Telephone the hotel security control on the pre-assigned number and tell them the

staff identification, and the exact location o f the fire. I f possible, the staff should

also mention the nature o f the fire, e.g. curtain on fire in a guestroom,

cooker-hood on fire in the kitchen, computer short-circuit set light to paper on

desk, etc.

b. Assist in fighting the fire if condition allows. Without putting oneself in danger,

one quick attempt should be made to put out the fire by using the nearest available

fire extinguisher or hosereel. In the kitchen area, at the same time all gas supply

AB-l

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 valves for ovens and stoves should be turned off, and electrical appliances should

be switched off.

c. Close all doors and windows if the fire grows beyond control and leave the fire scene

immediately; stand by at a safe distance and wait for the arrival o f senior staff or

fire fighting brigade and remain calm.

d. Keep all journalists and photographers away from the fire location, victims, and

evacuees. There is always intense media interest in a fire at a five-star

international chain hotel, whether or not there are any casualties.

It is the responsibility o f the hotel management to avoid the occurrence o f fire and the

following prevention plan or guidelines for the hotel staff are proposed:

a. Never smoke in the hotel, particularly in non-smoking areas.

b. In areas where smoking is allowed, make sure that matches are extinguished

before discarding them. Never throw cigarette ends or matches directly into

waste paper baskets. Make sure that smouldering cigarettes and tobacco are not

left unattended.

c. Keep heater to be at least two feet away from any furniture and curtains; and

housekeeping or staff members to switch off heaters when leaving the room.

AB-2

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 d. As previously mentioned, electrical safety is o f prime importance. Do not

overload any electrical circuits. If a multi-plug has to be used, consult the hotel

technical department for technical advice.

e. When leaving the workplace, make sure that all electrical appliances are switched

off.

f. In accordance with the Dangerous Goods licence requirement, excess material of

a flammable nature should not be stored in the workplace, e.g. paraffin, solvents

and gas cylinders.

g. Ensure that all electrical installations and wiring are properly connected and

installed by a qualified electrician.

h. Conduct training to all hotel staff so that they know the locations o f the nearest

fire extinguisher, hosereel and breakglass alarm and also how to operate them.

i. Know the fire exits and escape routes to outside the building together with the

assembly point in case o f evacuation.

j. Ensure that all the fire exits in the hotel are not obstructed or jammed and the fire

doors can all be opened.

AB-3

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 In the following, recommended actions are outlined for each hotel department when a fire

is confirmed:

a. At the telephone department and in the PABX room, the duty telephone

supervisor will take charge o f the role as the control communication centre. If

the PABX room is evacuated, the front office will take over the job. If

confirmed to be a real fire, hotel staff present will call at the operator to raise the

fire alarm and then proceed to attempt to put out the fire with the nearest fire

extinguisher. The operator in charge should call the general manager or resident

manager and all main department heads to advise locations and seriousness of the

fire. Upon receiving the confirmed fire alarm, the operator should group page

“999” meaning Fire. The operator should also inform housekeeping to activate

their own paging system (group calling “999”). This paging notifies all page

carriers that there is a fire somewhere in the building and all page carriers are to

report to their department heads.

b. For the hotel engineering department or the technical services division, the aim

always is to locate and put out the fire as soon as possible, with minimal

disturbance to the hotel guests or clients and to prevent panic and chaos.

The telephone operator in charge or security controller will immediately contact the duty

engineer when a fire alarm is activated or when advised by an individual that there is a

fire. The controller should give as much details as he has to hand about the fire to the

Duty Engineer, who will immediately go to the fire scene carrying a walkie-talkie. As

soon as he arrives at the fire scene, he will confirm the exact location and whether it is a

AB-4

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 false alarm, the fire is under control, or whether the fire appears to be out o f control. If

it is a false alarm or the fire is put out, the duty engineer will reset the alarm, advising the

security controller or telephone operator when this has been done.

When the fire has been extinguished, he will complete a full written report on building

and plant damage to be passed to the general manager.

The input from the security controller is also very critical. His responsibility is to locate

and put out the fire as soon as possible, with minimal disturbance to the hotel guests or

clients and to prevent panic and chaos. If a direct call by the hotel employee

discovering the fire is made to security control, the controller will inform the exact

location o f the fire to the duty engineer and PABX supervisor.

The security controller will immediately dispatch the nearest patrolling guard to report on

the state o f the fire and if necessary, attempt to put out the fire while waiting for the

arrival o f the duty engineer. If the fire is confirmed and is still out o f control, the

security controller will call out the fire fighting team, advising them on the location and

size o f the fire. He should alert all guards that a fire emergency is in progress and the

location o f the fire. He will send a guard to the entrance o f the hotel drive to show the

emergency services the best approach and the location o f the fire. Moreover, he should

ensure that approach roads are kept clear and any access gates are opened and guarded

where necessary. However, if the fire is in the guest floor areas, the security controller

should call security guards to confirm floor attendants. Hotel guests on one floor below

and two floors above the fire floor may be required to evacuate. The guards will then

stay in the area o f the guestroom that has been evacuated to ensure that no unauthorized

AB-5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 person enters. They will also advise control if the fire breaks out in the guestroom and

into the lift lobby prior to the arrival of the fire fighting team or fire department.

If the fire breaks out in the guestroom and into a lift lobby or smoke affects the lift shafts

or public areas, they should advise the controller immediately by walkie-talkie, any

movement in the building should then be by stairs only.

On the other hand, if the fire is in the hotel restaurants, if available, as many guards as

possible should be sent to the restaurant concerned without detracting from other security

requirements. Guards should maintain the security o f the restaurant, ensuring no

unauthorized entry o f persons. Once the evacuation has been completed, no person is

allowed into the restaurant until cleared by the fire fighting team, the fire department, the

security manager or the general manager.

If the fire is in the hotel shopping arcades, shops on either side o f the fire must be

immediately advised on the possible danger. If these shops are closed, the key holders

must be advised immediately. Two guards must be sent to the area o f the fire to ensure

no looting takes place following the sounding o f the alarm. When the fire has been

extinguished, a guard should remain in the area of the shop until relieved by the shop

manager. At the same time, all guards are advised that the emergency is over. Where

necessary, post guards over empty or damaged property if it cannot be secured from

possible entry o f unauthorized persons (including staff members, unless authorized by the

security manager or the hotel manager).

AB-6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The in-house hotel fire fighting team aims to locate and put out the fire as soon as

possible, with minimal disturbance to the hotel guests and clients and to prevent panic

and chaos. The fire fighting team will consist o f the duty engineer (team leader) and

three trained members o f each shift duty. Three persons will be nominated as the fire

fighting team at the beginning o f each shift by their shift supervisor, one from the

engineering department and two from the security department.

When a fire is discovered in the restaurant, any hotel staff members should raise the

alarm immediately by breaking the glass of the nearest alarm. He should also telephone

the PABX supervisor and tell the supervisor his name, the event location o f the fire and

the nature o f the fire; curtains and furniture on fire or cooker hood on fire, etc.

Together with the nearest staff, he should make an attempt to put the fire out with the

extinguishers or fire blankets on hand. However, do not endanger themselves is the

basic principle in handling any fire outbreak.

If the fire is in the kitchen, all supply valves for ovens and stoves should be turned off

and electrical equipment switched off. If one is not be able to extinguish the fire,

evacuate the kitchen in an organized manner and close all the doors behind. The senior

chef should ensure all staff are accounted for and report this to the manager.

The restaurant supervisor should then proceed immediately to the assembly point o f food

and beverage department and report to the F&B manager that evacuation is complete.

He should then pass a list o f hotel residents who had evacuated from the dinning rooms to

the F&B manager for transmission to the front office manager.

AB-7

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 The restaurant cashier with an escort o f two restaurant staff should take all cash boxes

until relieved by the manager. The last person to leave the restaurant should be the

assistant manager and the head waiter who should check that all side rooms and toilets,

etc. have been cleared o f guests and staff. He should then report the completed

evacuation o f the restaurant to the manager.

When the fire is discovered in the hotel shopping arcade, the shop staff member

discovering the fire should raise the alarm by shouting “Fire”. He should also break the

glass o f the nearest alarm. He should also telephone the PABX room and tell the

telephone supervisor his name, the exact location o f the fire and the nature o f the fire.

He should make one quick attempt to put out the fire with extinguishers on hand. If

unsuccessful, close the windows and doors where possible. Turn o ff any gas supply and

switch off electricity at the main, if applicable. He should remain calm and ensure all

clients and staff have left the shop before finally closing the front door. He should also

advise the neighbouring shops on either side that there is a fire and wait at a safe distance

from the shop for the fire fighting team’s arrival. He should not re-enter the shop until

he is advised that it is safe to do so by the fire fighting team or fire department or security

officer.

If the fire is discovered in the hotel carpark, the hotel staff member should raise the alarm

immediately by shouting “Fire”. He should break the alarm o f the nearest alarm. He

should also contact the PABX room and tell the supervisor his name, the exact location of

the fire, and the nature o f the fire, car on fire or wood and paper set on fire, etc.

AB-8

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 If the fire involves petrol, use the foam fire extinguisher and try to put the fire out. If

unsuccessful, move away from the fire. He should stop any carpark client moving

closer than 50 m from the fire, explaining the possibility o f the car exploding. “No entry”

signs are to be placed at the carpark entrance, and a guard should be put on duty at the

entrance to enforce this and also to allow the fire department vehicle access to the

carpark. Vehicle owners should be persuaded that it may not be safe for them to attempt

to get their own cars out o f the carpark.

For kitchen staff, he should immediately turn off the gas supply valves for ovens, stoves

and switch off all electrical equipment. Moreover, all air-conditioning and exhaust fans

should be switched off. On leaving, all doors should be closed. He should report to the

food and beverage manager to assist in evacuating diners.

For floor housekeepers, they should inform the room attendants o f evacuation order in a

calm manner; carry with the floor master key and sufficient chalk to mark crosses on

guest room doors when the rooms are confirmed as empty. They should assign room

attendants to call on guest room sections by sections. Further assistance will be

provided by the room services staff, reception staff, reservation staff and stewards.

They should also reassure guests that everything is under control. Furthermore, they

have to ensure that guests do not attempt to take with the any bulky luggage. Also, they

need to ensure that guests do not attempt to use the lift but proceed in an orderly maimer

to the designated fire exits. If they might have difficulty going downstairs, provide

assistance where practicable. After all the guests have been evacuated, ensure the floor

attendants check the end room again. When the check has been completed, chalk a

large cross on the door to indicate the room is empty. Finally, make sure all the doors

AB-9

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 are closed. They have to ensure that they are the last one to leave the guest floor.

They should assist and lead the guests to the assembly point via the fire exits; and report

to the executive housekeeper that the evacuation is complete.

For room attendants, they should remain calm on receiving evacuation message from the

floor housekeeper; carry with the section master key and proceed to knock loudly on

guestrooms within their own section. If guests do not respond, they should open the

door with the master key and tell the guests that: “There is an emergency. Please follow

me to the exit”. Moreover, they have to persuade guests not to carry with them bulky

luggage and that it is most likely they will be returning to their room in a short while.

They should remember to call each guestroom and assist the floor housekeeper to ensure

that all guests have been evacuated. They need to close the door after floor housekeeper

has marked the door with a large chalk cross and assist in leading the guests to the

assembly point.

For the front office cashier supervisor or outlet cashier, they should lock up records and

documents; evacuate cash boxes to the location pre-assigned by the financial controller

and report to the management accountant. Two persons maintain guard on the safe

deposit room unless ordered to evacuate the lobby.

For stewarding staff, they should immediately report to the executive housekeeper to

assist with the evacuation o f hotel resident guests.

For front office supervisors, they should assign clerks to stand by at the fire exits to direct

guests coming down from floors to the hotel lobby; and direct the guests to proceed to the

AB-10

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 predetermined assembly area. They have to also furnish the front office manager with an

updated guests list by name and room. They have to lead the rest o f the clerks to calm the

evacuees and provide necessary assistance. They need to stand by for further

instructions from the front office manager.

For baggage staff, they should guard the main door to prevent any unauthorized entry,

then proceed to the stairwells on the first floor to direct guests to the ground floor level.

For concierge clerks, they should assist the front office manager with evacuation o f shops

within the body o f the hotel lobby.

For the contractor and their workmen, they should know exactly what actions to take in

the event o f a fire. The chief engineer should ensure that the contractors have been

advised by their supervisor o f the location o f escape routes, fire points, fire extinguishing

equipment, alarm call points. If appropriate, the instructor must know the telephones

from which the fire brigade can be called. The contractor should have received

instructions on how and when to inform the telephone operator. The contractor on site

should understand the means o f raising the alarm and the use o f fire extinguishing

equipment. Any damage to the fire protection equipment by contractors should be

reported to the chief engineer immediately.

For cutting and welding job, the area must be checked by the duty engineer for fire

immediately after the work has been finished and also some 30 minutes later. Portable

extinguishers and hosereels should be ready.

AB-11

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 APPENDIX C EQUATIONS FOR HEAT TRANSFER AT

SPRINKLER HEADS

Since it takes time for the thermal sensing element o f the sprinkler head to heat up and

actuate, a time constant can be defined from the heat balance equation [Chow and Ho,

1990; 1992]:

If there is no heat lost or gained from the element to the frame and the associated

pipework, the heat balance equation can be simplified as:

A fundamental time constant Te' o f the sensing element (isolated time constant) can be

defined by:

de, ( A 0 ,- A 6 .)
dt r'

where r ' = 171e°e

e h.A.

AC-1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 For an isolated thermal sensing element, the time constant depends only on the

convective heat transfer coefficient he and should be independent o f the rate of

temperature increase.

If heat loss from the element to the frame is taken into account, Equation (C l) becomes:

M. (A » ,-A O .)
dt r; R,f <

KA
V <C
5a>
= ef ef
(C 5b)

Taking the pipework to be a large heat sink at a constant temperature and ignoring the

heat gained from the element, a similar heat balance equation can be obtained:

m f cf = h f A f (A 0 g ~ A 9 f )- h fP A fP A G f (C5>

Rearranging

de, ( A g ,- A 9 ,) R , A 0,
dt r’t R, f

, . rnfCf 1 , 1
where r ' = 1 1 R , = ------- and R , =
J U A J U A h
^ fAf h fAf " h fpA f P

AC-2

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Two other time constants can be derived by taking the heat transfer to the frame and

pipework into account. They are the collective time constant t ' and the apparent time

constant r " . The collective time constant x' is a long-term time constant (r » x') while

the apparent time constant r " describes a short-term phenomenon. Both values are

useful in determining the characteristics o f the sprinkler itself.

The relationship between the isolated time constant xe' with the heat transfer coefficient

he is given by Equation (C3). In forced convective air flow, the Nusselt number Nu is

related to the Reynolds number by:

Nu = / (Re) (C7)

where f (x) is a function o f any variable x.

Experimental data on the fire environment gives:

Nu = l?VRe (C8)

where B is a constant related to the geometry o f the sensing element.

The Reynolds number is given by:

(C9)
v

AC-3

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Also, the heat transfer coefficient he is related to the Nusselt number Nu by:

Nu = — (CIO)
k

Therefore, from Equations (C9) and (CIO), the heat transfer coefficient is:

K=BkA± (Cii)

k
The quantity—t= is a property o f air and is approximately independent o f temperature.
Vv

It follows from Equation (Cl 1) that for a given sensing element (B and L taken to be

constant):

he °c 4 u (C l2)

since xe' is inversely proportional to he as shown in Equation (C3), a relationship between

xe' and yfu can be derived:

reoc —)= or xe' yfu = constant (C l3)

V«

The product xe'4 u is related to the thermal inertia o f the heat sensing element which is

independent o f the gas velocity and temperature. The term ‘Response Time Index’ is

defined for this equation:

AC-4

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 R T I = x e 'yfu (Cl 4)

Since the effect o f air velocity is included in RTI, this quantity is better than the isolated

time constant xe\ However, conductive heat transfer between the element, the frame

and the associated pipework has not been included. Therefore, a quantity known as the

virtual RTI is derived to include the conductive heat transfer:

D T /
RTIy = —— ( Cl 5)
i + ~*Ju
n

The quantity RTIV is related to the apparent time constant x". This is a parameter

measured in the plunge test including conductive heat loss. It is a true measure o f the

real response time index o f the sensing element. Taking the measured response time

index to be RTIV, the value o f the real response time index R TI can be measured by

plotting —J— against

RT1V v«

Generally, the hot gas temperature increased A6g is related to the time o f heating t by:

A9g = p f (Cl 6)

Two specific cases are o f direct interest. The first is the sprinkler head being immersed

in a constant air heat bath (n = 0). The second one is the case with a constant rate o f air

temperature rise (n = 1). These are the physical principles o f two well-known tests: the

AC-5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 plunge test and the ramp test. From the two tests, the parameters Oe and RTI can then

be derived experimentally.

The plunge test [Heskestad and Smith, 1976, 1980] describes a step function of the air

temperature (n = 0). Using Laplace transforms and neglecting the heat loss o f the frame

from the element, Equations (C4) to (C6) give:

t
T (Cl 7)

where A0e = Q„om - 0e (0).

This can be calculated from A6e and t which are measured in the plunge test, but it is not

a true value o f the isolated time constant since A0e and t are measured with heat loss.

However, if the thermal losses from the element are considered, an apparent time

constant can be derived:

- t ,op
T / \ (Cl 8)
In

AC-6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 where kc is related to the conductive heat loss c in view of Equation (C l 5):

- (Cl 9)
^+ ~Jur

Substituting into Equation (C l8):

- t
T =■ (C20)

The virtual response time index RTI can be determined by this apparent time constant r " :

RTIv = x ” ^Ti (C21)

Experimentally, initial temperature sprinklers are usually kept at the ambient temperature

o f the test laboratory. Thetime taken for thesprinkler to operate when it is plunged into

a hot air-stream is taken as top. Combining with 0emeasured in the ramp test, x " can

then be calculated from Equation (C20).

The ramp or rate-of-rise test describes a constantly increasing air temperature (i.e. n = 1).

Solving Equations (C4) and (C6), the equation for the ramp test becomes:

toP = ^ E - ^ y - ^ r ' (C22)

AC-7

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Rearranging the equation:

Og —px' + 0E (C23)

The mathematical expressions for &E and x' are very complicated but they can be

determined by plotting Equations (C22) and (C23). Experimentally, a sprinkler is

immersed in an air-stream with the temperature increasing at a constant rate o f P “C m in1,

the temperature o f the hot gas at the time o f operation is 6g.

The quantity dE appeared due to the conductive heat loss from the sensing element to the

frame. In view o f Equations (C l 8), (Cl 9) and (C20), dE is related to the conductive heat

loss constant c by:

, , c y ^ - e E(0) =eE-e,m
• lu ) Os - 0 E((i) 0 ,- 0 ,(0 )

Hence

(C24)

Therefore, plotting the graph o f dE against —t= should give a straight line.
VM

AC-8

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 APPENDIX D CALCULATION FOR MINIMUM HEAT RELEASE

RATE FOR FLASHOVER

Minimum heat release rate Qmf (kW) for flashover with Vef was proposed by Thomas

[Chow et al, 2003c]:

Qmf = 378 Vef + 7.8 A t (D l)

where Ve/ is the ventilation factor and A T is the effective area o f the test chamber

Hence,

Qmf = 378 (0.8) (2) ( 2 ) + [2 (6) (2.4) + 2 (3.6) (2.4) - (2) (0.8)]

= 1202.2 kW

The heat release rate o f test kitchen exceeds the minimum heat release rate calculated.

AD-l

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 REFERENCES

Alpert R.L. (2002)

Ceiling Jet Flows
SFPE Handbook o f Fire Protection Engineering, 3rd Edition, Section 2, Chapter 2,
pages 2-18 - 2-30
National Fire Protection Association

Andersson P. and Holmetedt G. (1999)

Limitations o f Water Mist as a Total Flooding Agent
Journal o f Fire Protection Engineering, Volume 9, Number.4, pages 3 1 - 5 0

ASHRAE (1997)
ASHRAE Fundamental Handbook (SI)
American Society o f Heating, Refrigerating and Air-conditioning Engineers

Australian Standard (1988)

AS2419:1 — 1988: Fire Hydrant Installations, Part 1 - System Design, Installation,
and Commissioning

Babrauskas V. (2003)
Ignition Handbook, pages 32 - 56
Fire Science Publishers

Babrauskas V. and Grayson, S.J. (1992)

Heat Release in Fires, pages 151 - 240
Elsevier Applied Science

Babrauskas V. and Peacock, R.D. (1992)

Heat Release Rate: The Single Most Important Variable in Fire Hazard
Fire Safety Journal, Volume 18, pages 255 - 272

Bickerdike Allen Partners (1996)

Design principles o f Fire Safety
Her Majesty Stationery Office

R-l

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 British Standard Institution (1983)
BS 5588: Part 3: 1983 Fire Precautions in the Design and Construction of
Buildings - Part 3: Code o f Practice for Office Buildings

British Standard Institution (1985)

BS 5588: Part 2: 1985 Fire Precautions in the Design and Construction of
Buildings —Part 2: Code o f Practice for Shops

British Standard Institution (1988)

BS5306: Part 1: 1988 Fire Extinguishing Installations and Equipment on Premises —
Part 1: Hydrant System, Hose Reels and Foam InletsN

British Standard Institution (1990)

BS5306-2: 1990 Incorporating Amendment No. 1 and Implementing Corrigendum
No. 1 Fire Extinguishing Installation and Equipment on Premises - Part 2:
Specification for Sprinkler System

British Standard Institution (1994)

Draft Code o f Practice for the Application o f Fire Safety Engineering Principles to
Fire Safety in Buildings

British Standard Institution (1997)

DD240 Fire Safety Engineering in Buildings - Part 1: Guide to the Application of
Fire Safety Engineering Principles

British Standard Institution (1997)

BS 5588-11: 1997 Fire Precautions in the Design, Construction and use of
Buildings — Part 11: Code o f Practice for Shops, Offices, Industrial, Storage and
Other Similar Buildings

British Standard Institution (1999)

ISO/TR 13387-1:1999 Fire Safety Engineering — Part 1: Application o f Fire
Performance Concepts to Design Objectives

British Standard Institution (2000)

BS 5499-4: 2000 Safety signs, including fire safety signs. Code o f practice for
escape route signing

R-2

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 British Standard Institution (2001)
BS 7974 Application o f Fire Safety Engineering Principles to the Design of
Buildings —Code o f Practice

British Standards Institute (2002)

PD 7974(2002) Application o f Fire Safety Engineering Principles to the Design of
Buildings — Part 0 Guide to Design Framework and Fire Safety Engineering
Procedures

Bryan J.L. (2002)

Behavioural Response to Fire and Smoke
SFPE Handbook o f Fire Protection Engineering, 3rd Edition, Section 3, Chapter 12,
pages 3-315 - 3-340
National Fire Protection Association

Buchanan A.H. (1999)

Implementation o f Performance-based Fire Codes
Fire Safety Journal, Volume 32, Number 4, pages 377 —383

Buchanan A.H. (2001)

Structural Design for Fire Safety, pages 2 5 - 4 2
John Wiley & Sons

Buckley G., Bradbom W., Edwards J., Marchant R., Terry P. and Wise S. (2000)
The Fire Brigade Intervention Model
Fire Safety Science - Proceedings of the Sixth International Symposium, pages
1 8 3 -1 9 4

Buildings Department (1995)

Code o f Practice for Mean o f Assess for Firefighting and Rescue
Hong Kong Government

Buildings Department (1996a)

Code o f Practice for the Provision of Means of Escape In Case o f Fire
Hong Kong Government

R-3

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Buildings Department (1996b)
Code o f Practice for Fire Resisting Construction
Hong Kong Government

Buildings Department (1998)

Practice note for Authorized Persons and Registered Structural Engineers 204,
Guide to Fire Engineering Approach
Hong Kong Special Administrative Region

Building Regulations (1985)

Approved Document B, B2/3/4 Fire Spread
Her Majesty’s Stationery Office

Butcher E.G. and Parnell A.C. (1979)

Smoke Control in Fire Safety Design, pages 1 2 - 1 6
E. & F.N. Spon

Canter D.J. (1980)

Fire and Human Behaviour, pages 102 - 114
John Wiley & Sons

Cherry T. Don. (1986)

Total Facility Control, pages 113 - 228
Butterworths

Cheung K.P. (1995)

Design for Effective and Reliable Boosting o f Water from Fire Engines to
Water-based Fire Fighting Systems in Tall Buildings
Proceedings o f 5th World Congress, Council on Tall Buildings and Urban Habitat,
1995, pages 775 - 789

Cheung K.P. (1997)

Enhancement Proposals on Sprinkler System and Standpipe / Wet Riser / Hydrant
System in Tall Buildings
Building Journal Hong Kong China, February 1997, pages 66 - 69

R-4

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Cheung K.P. and Chan H.W. (1997)
A Discussion on the Design Principles o f Fire Hydrant Systems and Sprinkler
Systems in Refuge-floor Based Fire Safety Design in Tall Buildings
Proceedings o f National Fire Conference, Chinese Association for Advancement of
Science and Technology, Zhuhai, China, October, 1997

Cheung K.P. and Chan H.W. (2003)

High Water - Discussion on Methods o f Improving the Water Supply to Sprinkler
and Hydrant Systems in Tall Buildings
Proceedings o f Fire Prevention - Fire Engineers Journal, Fire Protection Association,
UK, May 2003, pages 65 - 67

Chin K.S. (2002)

Development o f Bench-scale Testing o f Sprinkler and Smoked Detector Activation/
Response Time
Fire Engineering Research Report, School o f Engineering, University o f Canterbury,
February, 2002

Chow W.K. and Ho P.L. (1990)

Thermal Response o f Sprinkler Head
Building Services Engineering Research and Technology, Volume 11, Number 2,
pages 3 7 - 4 7

Chow W.K. and Ho P.L. (1992)

Thermal Response o f Sprinkler Head in Hot Air Stream with Speed less than 1 ms-1
Fire Science and Technology, Volume 12, Number 1, pages 7 - 2 2

Chow W.K., (1995)

Use o f the ARGOS Fire Risk Analysis Model for Studying Chinese Restaurant Fires
Fire and Materials, Volume 19, Number 4, pages 171 —178

Chow W.K. and Wong L.T. (1998)

Fire Safety Codes for Hong Kong: Inadequacy for Atrium Design
Building Services Engineering Research and Technology, Volume 19, Number. 2,
pages 93 - 99

R-5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Chow W.K. (1998)
Numerical Studies on Recent Large High-rise Building Fire
ACSE Journal o f Architectural Engineering, Volume 4, Number 2, pages 65 —74

Chow W.K. (1999)

A Preliminary Discussion on Engineering Performance-based Fire Codes in the
Hong Kong Special Administrative Region
International Journal on Engineering Performance-Based Fire Codes, Volume 1,
Number 1, pages 1 —10

Chow W.K. (2001a)

Review on Fire Safety Management and Application to Hong Kong
International Journal on Engineering Performance-Based Fire Codes, Volume 3,
pages 52-58

Chow W.K. and Lui G.C.H. (2001)

A Fire Safety Ranking System for Karaoke Establishments in the Hong Kong
Special Administrative Region
Journal o f Fire Sciences, Volume 19, Number 2, pages 1 0 6 -1 2 0

Chow W. K. (2001b)
Support on Carrying out Full-scale Burning Tests for Karaokes
International Journal on Engineering Performance-Based Fire Codes, Volume 3,
Number 3, pages 104 - 112

Chow W.K. (2002)

Review on Heat Release Rate o f Burning Furniture
International Journal on Engineering Performance-Based Fire Codes, Volume 4,
Number 2, pages 54 —59

Chow W.K. and Zou G.W. (2003a)

Experimental Verification o f the Fire Field Model Fire Dynamics Simulator
Journal o f Fire Protection Engineering, June

Chow W.K., Gao Y., Dong H., Zou G.W., Luo Z. and Meng L. (2003b)
Will Water Mist Extinguish a Liquid Fire Rapidly?
Architectural Science Review, Volume 46, Number 2, pages 1 3 9 -1 4 4

R-6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Chow W.K., Gao Y., Dong H., Zou G. W., Luo Z. and Meng L. (2003c)
Experimental Studies on Minimum Heat Release Rates for Flashover with Oxygen
Consumption Calorimetry
Architectural Science Review, Volume 46, Number 3, pages 291 - 296

Chow W.K., Zou G. W., Dong H. and Gao Y. (2003d)

Necessity o f Carrying out Full-scale Burning Tests for Post-flashover Retail Shop
Fires
International Journal on Engineering Performance-Based Fire Codes, Volume 5,
Number 1, pages 20 - 27

Chow W.K. (2004a)

Smoke Control for Retail Shops with Cabin Design
ASHRAE Transactions, Volume 110, Part 2, pages 417 —423

Chow W.K. (2004b)

A Short Note on Fire Safety for New Architectural Features
International Journal on Architectural Science, Volume 5, pages 1 - 4

Chow W. K. (2004c)
Special Issue on Full-scale Burning Tests
International Journal on Engineering Performance-Base Fire Codes, Volume 6,
Number 3, pages 88 - 92

Chow W.K. and Meng L. (2004d)

Analysis o f Key Equations in a Two-layer Zone Model and Application with
Symbolic Mathematics in Fire Safety Engineering
Journal o f Fire Sciences, Volume 22, Number 2, pages 9 7 - 1 2 4

CIBSE (2003)
CIBSE Guide E: Fire Engineering, 2nd edition
The Chartered Institution o f Building Services Engineers

Cox G. (1995)
Combustion Fundamentals o f Fires, pages 6 1 - 8 2
Academic Press

R-7

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Custer R.L.F. and Meacham J.B. (1997)
Introduction to Performance-based Fire Safety, pages 102 —116
National Fire Protection Association

Dailey W. (2000)
A guide to Fire Safety Management, pages 36 - 50
Perpetuity Press

Della-Giustina and Daniel (1999)

The Fire Safety Management Handbook, 2nd Edition, pages 1 3 3 -1 4 2
American Society o f Safety Engineers

Downie B. and Polyneropoulos C. (1995)

Interaction o f Water Mist with a Buoyant Methane Diffusion Flame
Fire Safety Journal, Volume 24, Number 4, pages 359 - 381

Eaton C. (1991)
Microeconomic Reform: Fire Regulation, Appendix A — Draft National Building
Fire Safety Systems Code
Building Regulation Review Task Force, Department o f Industry Technology and
Commerce

Edwards N. (1998)
A New Class o f Fire
Fire Prevention 310, 8

Information Services Department (1998)

Effective Building Management & Fire Safety Rest with Owners
Hong Kong Special Administrative Region
Website: http://www.info.gov.hk/gia/general/199806/26/0626098.htm

Engineering Guide (2002)

Evaluation o f the Computer Fire Model DETACT -Q S , The SFPF Task Group on
Computer Model Evaluation
Society o f Fire Protection Engineers

R-8

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 England J.P., Yound S.A., Hui M.C. and Kurban N. (2000)
Guide for the Design o f Fire Resistant Barriers and Structures
Building Control Commission, Melbourne

HMSO (1952)
Fire grading o f buildings Part III Personal safety, Post-war Building Studies No. 29,
pages 2 - 5
London

The Stationery Office (1997)

Fire Precautions (Workplace) Regulations
London

Fire Safety Symposium (1989)

Proceedings o f Fire Safety and Engineering International Symposium, The Warren
Centre for Advanced Engineering, The University of Sydney

Fire Safety Engineering (1989)

Fire Safety and Engineering Technical Papers - Book 1 and Book 2, The Warren
Centre for Advanced Engineering, The University of Sydney

Fire Services Department (1998)

Code o f Practice for Minimum Fire Service Installations and Equipment, Fire
Services Department, Hong Kong

Fitzgerald R.W. (1997)

Structural Integrity during Fire
Fire Protection Handbook 18th Edition, Section 7 Chapter 4, pages 7-50 —7-76
National Fire Protection Association

Fleischmann C. and Buchanan A. (2002)

Analytical Methods for Determining Fire Resistance o f Concrete Members
SFPE Handbook o f Fire Protection Engineering, Section 4 Chapter 10, 3rd edition,
National Fire Protection Association

Forney G.P. and Moss W.F. (1994)

Analyzing and Exploiting Numerical Characteristics o f Zone Fire Models
Fire Science & Technology, Volume 14, Number 1, pages 49 - 60

R-9

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Grant G., Brenton J and Drysdale D (2000)
Fire Suppression by Water Sprays
Energy and Combustion Science, Volume 26, pages 79 - 130
Gupta, A.K. (2001)
Modeling the Response o f Sprinklers in Compartment Fires
Fire and Materials, Volume 25, pages 117 - 121

Hadjisophocleous G.V., Benichou N. and Tamim A.S. (1998)

Literature Review o f Performance-based Fire Codes and Design Environment
Journal o f Fire Protection Engineering, Volume 9, Number 1, pages 12 - 40

Hall J.R. (2000)

On People and Chance: the ‘Hard’ Facts about the ‘Soft’ Branches o f Fire Safety
Science
Proceedings o f the Sixth International symposium, pages 23 - 40
IAFSS

Hartford C.T. (1998)

Report on Special Services Investigation o f Operation Times for Duct Mounted,
Corrosion Protected Sprinklers and the Influence of Sprinkler Spray on Heated Air
Flow through Ducting, File NC 1838, Project 97NK29176, HSB Industrial Risk
Insurers, Hartford, CT, 5 June.

Heskestad G. and Smith H.F. (1976)

Investigation o f a New Sprinkler Sensitivity Approval Test: The Plunge Test
Factory Mutual Research Coporation

Heskestad G. and Smith H.F. (1980)

Plunge test for Determination o f Sprinkler Sensitivity
Factory Mutual Research Corporation

Heskestad G. (1984)
Engineering Relations for Fire Plumes
Fire Safety Journal, Volume 7, pages 25 - 32

R-10

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Heskestad G. and Bill R.G. Jr. (1988)
Quantification o f Thermal Responsiveness o f Automatic Sprinklers Including
Conduction Effects
Fire Safety Journal, Volume 14, Number 1 & 2, pages 113 - 125

Heskestad G. and Bill R.G. Jr. (1989)

Modelling o f Thermal Responsiveness o f Automatic Sprinklers
Journal o f Fire Safety Science, pages 603 - 612

Office o f the Licensing Authority of the Home Affairs Department, Home Affairs
Department (2005)
Hotel and Guesthouse License, Hotel and Guesthouse Accommodation Ordinance,
Chapter 349 o f Laws o f Hong Kong, Hong Kong Special Administrative Region

Huggett C. (1980)
Estimation o f rate o f heat release by means o f oxygen consumption measurements
Fire and Materials, Volume 4, Number 2, pages 6 1 - 6 5

Hui M.C. (2004)

Quantification on the Total Fire Safety Concept - The Holy Grail o f Engineered Fire
Safety Design
Proceedings o f the Fire Conference 2004 - Total Fire Safety Concept, 6 — 7
December, Hong Kong, Volume 1, pages 1 7 7 -1 8 5

International Standards Organization (1993)

ISO 9705: Fire tests - Full-scale Room Test for Surface Products

Joyeux D. (2002)
Experimental Investigation o f Fire Door Behaviour During a Natural Fire
Fire Safety Journal, Volume 37, pages 605 —614

Krasuy J.F., Parker W J. and Babrauskas V. (2001)

Fire Behaviour o f Upholstered Furniture and Mattress, pages 1 5 - 2 0
William Andrew Publishing

R -ll

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Klaene B.J. and Sanders R. (2003)
Fireground Operations, Section 7 Chapter 23, Fire Protection Handbook, 19th
Edition, pages 7-333 - 7-341
National Fire Protection Association

Klote J.H. and Milke J.A. (2002)

Principles o f Smoke Management, pages 5 0 -9 5
ASHRAE and SFPE

Knoxville T.N. (1994)

Water Mist Fire Suppression Systems
Proceedings o f Technical Symposium on Halon Alternatives, Society o f Fire
Protection Engineers and PLC Education Foundation, 27 —28 June

Kung H.C., Spaulding R.D. and Stavrianidis P. (1991)

Fire Induced Flow Under a Sloped Ceiling
Fire Safety Science - Proceedings o f the Third International Symposium, pages
271 - 280
Elsevier Applied Science

Leung C.W., and Chow W.K. (2001)

Survey on Partition Walls Commonly Used in Hong Kong and Estimation o f the
Heat Release Rates during Fire
Architectural Science Review, Volume 44, Number 4, pages 379 - 390

Liu Z., Kim K., and Su Z. (2001)

Examination o f Performance o f Water Mist Fire Suppression Systems under
Ventilation Conditions
Journal o f Fire Protection Engineering, Volume 11, Number 3, pages 164 - 193

Liu Z., Kim A. and Carpenter D. (2004)

Extinguishment o f Cooking Oil fires by Water Mist Fire Suppression System
Fire Technology, Volume 40, pages 309 - 333

Loss Prevention Council (1990)

High Rise Sprinkler Systems
Technical Bulletin, TB13:1990

R-12

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Ma H. (1995)
Illustrative Description to the Fire Safety Code o f Highrise Public and Private Use
Building Designs GB 50045 - 95, Public Co. China - in Chinese

M aH . (1997)
Summary o f Fire Alarm System Standards and Department o f Fire Extinguishing
System Standards
Proceedings o f Mainland - Hong Kong Engineering and Construction Standards
Exchange Seminar ’97, 14-15 April 1997, Beijing, China Association for
Engineering Construction Standardization, Hong Kong Institutes o f Engineers,
pages 38 —40

Malhotra H.L. (1987)

Fire Safety in Buildings
Building Research Establishment Report, Department o f the Environment, Building
Research Establishment, Fire Research Station, Borehamwood, Herts, WD6 2BL

Mawhinney J.R. (1993)

Characteristics o f Water Mists for Fire Suppression in Enclosures
Proceedings o f Halon Alternatives Technical Working Conference, 1 0 — 11 May,
Albuquerque, NM, USA

Mawhinney J.R., Dlugogorski B.Z. and Kim K. (1994)

A Closer Look at the Fire Extinguishing Properties of Water Mist
Fire Safety Science - Proceedings o f the Fourth International Symposium, pages
4 7 -6 0

McKinnon G.P. (2003)

What Time has Crystallized into Good Practice
Fire Protection Handbook, Volume II, Appendix C, 19th Edition, pages C l - C5
National Fire Protection Association

Melinek S J. (1988)
Thermal Response o f Sprinklers - a Theoretical Approach
Fire Safety Journal, Volume 13, Number 2 & 3, pages 169 - 180

Ministry o f Public Security China (1987)

Fire Safety Code for Building Construction Design, GBJ 16-87, China - in Chinese

R-13

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Ministry o f Public Security China (1982)
Fire Safety Code o f Highrise Public and Private Building Designs, GBJ 45-82,
China - in Chinese

Ministry o f Public Security China (1997)

Code for Fire Protection Design o f Tall Buildings GB 50045 —95, Beijing, C hina-
in Chinese

Milke J.A. (2002)

Analytical Methods for Determining Fire Resistance o f Steel Members
SFPE Handbook o f Fire Protection Engineering, Section 4, Chapter 9, 3rd Edition,
National Fire Protection Association

Morgan H.P. and Gardener J.P. (1990)

Design Principles for Smoke Ventilation in Enclosed Shopping Centres
HMSO, UK.

Morgan H.P. and Marshall N.R. (1975)

Smoke Hazards in Covered, Multi-level Shopping Malls: An Experimentally-based
Theory for Smoke Production
BRE Current Paper 48, British Research Establishment

NFPA 14 (1993)
Standpipe and Hose Systems
National Fire Protection Association

NFPA 101 (1994)

Life safety code
National Fire Protection Association

NFPA 101A (1995)

Guide on alternative approaches to life safety
National Fire Protection Association

NFPA 550 (1995)

Guide to the Fire Safety Concepts Tree
National Fire Protection Association

R-14

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 NFPA (1997)
Fire Protection handbook, 18th edition
National Fire Protection Association

Nelson H. E. and Mowrer F.W. (2002)

Emergency Movement
SFPE Handbook o f Fire Protection Engineering, Section 3, Chapter 14, 3rd Edition,
pages 3-367 - 3-379
National Fire Protection Association

Ohlemiller T.J. (2002)

Smouldering Combustion
SFPE Handbook o f Fire Protection Engineering, Section 2, Chapter 9, 3rd Edition,
pages 2-200 - 2-209
National Fire Protection Association

Peacock R.D. and Babrauskas V. (1991)

Analysis o f Large-scale Fire Test Data
Fire Safety Journal, Volume 17, pages 387 - 414

Peacock R.D., Bukowski R.W., Jones W.W., Reneke P.A., Babrauskas V. and
Brown J.E. (1994)
Fire Safety o f Passenger Trains: A Review o f Current Approaches and o f New
Concepts
NIST Technical Note 1406, National Institute o f Standards and Technology

Proulx G. (2002)
Movement o f People: the Evacuation Timing
SFPE Handbook o f Fire Protection Engineering, Section 3 Chapter 13, 3rd Edition,
pages 3-342 - 3-364
National Fire Protection Association

Putorti, A.D., Belsinger, T.D. and Twilley, W.H. (1995)

Determination o f Water Spray Drop Size and Velocity from a Low Pressure, High
Momentum, Water Mist Nozzle
Report o f Test FR 4000, May

R-15

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Richardson J.K. and Seaton M. (1997)
Performance-based Fire Codes
NFPA Journal, Volume 91, Part 1, pages 72 —77

R.L.P. Custer and Meacham B.J. (1997)

Introduction to Performance-based Fire Safety, pages 42 —78
Society o f Fire Protection Engineers

R.P.E. Cote (1997)

Life Safety Code Handbook, 7th Edition
National fire Protection Association

Sako S. (1989)
Response Time o f Automatic Sprinklers below a Confined Ceiling
Journal o f Fire Safety Science, pages 613 —623

Schifiliti R.P., Meacham B.J. and Custer R.L.P. (2002)

Design o f Detection Systems
SFPE Handbook o f Fire Protection Engineering, Section 4, Chapter 1, 3rd Edition,
pages 4-1 - 4-43
National Fire Protection Association

Society o f Fie Protection Engineers (2001)

SFPE Engineering Guide to Performance-based Fire Protection Analysis and Design
o f Buildings
Society o f Fie Protection Engineers

Shields T. J. and Silcock G. W. H. S. (1987)

Buildings and Fire, pages 35 - 78
Longman Scientific & Technical

Shields T.J. and Proulx G. (2000)

The Science o f Human Behaviour: Past Research Endeavours, Current
Developments and Fashioning a Research Agenda
Fire Safety Science —Proceedings o f the Sixth International Symposium, pages 95 -
114, IAFSS

R-16

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Shipp M. (1998)
Fire Safety Management in Fire Safety Engineering
Proceedings o f Eurofire’ 98, 1 1 - 1 3 March, Brussels, Belgium

Smith D.A. and Shaw K. (1999)

The Single Burning Item (SBI) Test, the Euro Classes and Transitional Arrangement
Proceedings o f Interflam’ 99, 29 June - 01 July 1999, Edinburgh, UK, Volume 1,
pages 1 —9

SundstrOm B. (1995)
Fire safety o f Upholstered Furniture
Final report on the CBUF Research Programme
Interscience Communication Ltd, London, UK

Tamura, T.G. (1994)

Smoke movement & Control in high- rise Buildings, pages 40 - 46
National fire Protection Association

The Chartered Institution o f Building Services Engineers (1995)

Technical Memoranda TM19: Relationships for Smoke Control Calculations

Theobald C.R. (1987)

Thermal Response o f Sprinklers Part I: FRS Heated Wind Tunnel
Fire Safety Journal, Volume 12, Number 1, pages 5 1 - 6 3

Theobald C.R., Westley, Sarah.A. and Whitbread, S. (1988)

Thermal Response o f Sprinklers Part II: Characteristics and Test Methods
Fire Safety Journal, Volume 13, Number 2 & 3, pages 99 —114

Thome P.F., Theobald C.R. and Melinek S.J. (1988)

The Thermal Performance o f Sprinkler Heads
Fire Safety Journal, Volume 14, Number 1 & 2, pages 89 - 99

Tsui P.H., Cheung, T.C. (2004)

Passive Fire Protection in Hong Kong
Proceedings o f the Fire Conference 2004 - Total Fire Safety Concept, 6 - 7
December, Hong Kong, pages 1 - 8

R-17

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Tubbs B. (1999)
How Far have We Come? The ICC’s Perspective
Proceedings o f the Second Conference on Fire Safety Design in the 21st Century,
9 - 1 1 June, Worcester, Massachusetts, USA, pages 6 1 - 7 2
UL33 (2005)
Heat responsive links for fire protection service

Voelkert C (1998)
Out o f the flying pan
Fire Prevention, Volume 314, pages 24 - 26

Watts J.M. (2003)

Fundamentals o f Fire - Safe Building Design
Fire Protection Handbook, Section 2, Chapter 2, 19th Edition, pages 2-37 - 2-49
National Fire Protection Association

Wilson E. Bright. (1990)

An Introduction to Scientific Research, pages 220 - 226
Dover Publications

White, R.H. (2002)

Analytical Methods for Determining Fire Resistance o f Timber Members
SFPE Handbook o f Fire Protection Engineering, Section 4, Chapter 11, 3rd edition,
pages 4-257 - 4-271
National Fire Protection Association

Wu, H.H. Chow W.K. Gao Y., Zou G.W. and Dong H, (2005)
Performance evaluation o f fire services installation in kitchen o f big hotels
BUEE 2006, eighth International Symposium on Building and Urban Environmental
Engineering, 1 0 - 1 3 July, Tokyo, Japan

Yang, K.T. (1999)

Role o f fire field models as a design tool for performance-based fire - code
implementation
International Journal on Engineering Performance-based Fire Codes, Volume 1,
Number 1, pages 1 1 - 1 7

R-18

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 Yao B., Fan W and Liao G. (1999)
Interaction o f water mists with a diffusion flame in a confined space
Fire Safety journal, Volume 33, Number 2, pages 1 2 9 - 1 3 9

Yao B and Chow W.K. (2001)

Overview on water mist fire suppression system
Journal o f architectural Engineering, Volume 7, Number 3, pages 59 - 60

Yao C. (1988)
The development o f the ESFR sprinkler system
Fire Safety Journal, Volume 14, Number 1 & 2, pages 65 —73

Zukoski E.E. (1994)

Mass flux and fire plumes
Fire Safety Science - Proceedings o f the Fourth International Symposium, pages
137-147, IAFSS.

R-19

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 P U B L IC A T IO N S A R IS IN G F R O M T H IS T H E S IS

1. Chow W.K. and Wu H.H. (1998)

Fire Safety aspects on hotels
Proceedings o f First International Symposium on Engineering Performance -
Based Fire codes, 8 September 1998, Hong Kong, pages. 261 —266

2. Chow W.K. and Wu H.H. (2000)

Ventilation for kitchens in a big hotel
Proceedings o f The 7th International Conference on air distribution in room.
Volume II, Reading, UK, pages. 1177 - 1182

3. Wu H.H. (2003)
First Safety Engineering in a Hotel
International Journal on Engineering Performance - Based Fire Codes, Volume 5,
Number 4, pages 189 - 193

4. Wu H.H. (2003)
Fire Safety Management in Hotel
Proceedings o f International Conference on Sustainable Development in Building
and Environment, SET2 (Sustainable Energy Technology), 24 - 27 October 2003,
Chongqing, China and Journal o f Chongqing University (English Edition)
Volume 2, Special Issue, October, pages 236 - 241

5. Chow W.K. and Wu H.H. (2004)

Hotel kitchen fire
Proceedings o f the Fire Conference 2004 — Total Fire Safety Concept, 6 - 7
December, Hong Kong, Volume 2, Section 28

6. Wu H.H. (2005)
Hotel kitchen fire
P ro c eed in g o f th e 4 th E d u cation S y m p o siu m o n A d v a n c e d F ire R esea rch , 13 —16
June 2005, Hefei, Anhui, China, pages 37 - 44

p-i

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
 7. Chow W.K. and Wu H.H. (2005)
Hotel Kitchen Fire
Proceedings o f Hubei - Hong Kong Joint Symposium on Building services
Design, 1 —2 July Wuhan, China, pages 6 1 - 6 9

8. Wu H.H., Chow W.K., Gao Y„ Zou G.W. and Dong H. (2006)

Performance evaluation o f fire services installation in kitchens o f big hotels
Proceedings o f 8th International symposium on Building and Urban
Environmental Engineering, 1 0 —13 July, Tokyo, Japan

P-2

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.