FEASIBILITY AND DESIGN CONSIDERATIONS -...
Transcript of FEASIBILITY AND DESIGN CONSIDERATIONS -...
FEASIBILITY AND DESIGN CONSIDERATIONS
FOR THE USE OF LIFTS AS AN EMERGENCY
EXIT IN APARTMENT BUILDINGS
BY
Than Singh Sharma
BSc, BE (Fire), MSc (Disaster Mitigation), CFES, MIFE (UK)
Queensland University of Technology
Brisbane, Australia
A THESIS SUBMITTED TO THE SCHOOL OF URBAN DEVELOPMENTS
QUEENSLAND UNIVERSITY OF TECHNOLOGY IN PARTIAL
FULFILLMENT OF REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
March 2008
iii
STATEMENT OF ORIGINAL AUTHORSHIP
The work contained in this thesis has not been previously submitted for a degree or
diploma for any other higher education institution to the best of my knowledge and
belief. The thesis contains no material previously published or written by another
person except where due reference is made.
Signed:
THAN SINGH SHARMA
Date: 20.03.2008
v
ABSTRACT
Emergency evacuation in high-rise apartment building is a challenge for fire safety
professionals. Lift evacuation is a controversial issue because the safe operation of
lift is not ensured under the existing design and operating conditions. Lifts are not
permitted for public evacuation in apartment buildings during fire emergencies as per
the provisions of building codes and regulations. However, the concept of using lifts
for emergency evacuation has been gaining considerable attention during recent
years.
The lift evacuation can be considered as an alternative facility if it is efficient,
reliable and readily accessible. It can also provide a safer means of evacuation for the
aged and disabled persons, who may not be able to evacuate promptly, efficiently
and unassisted using the exit stairs during fire emergencies. Moreover, lifts can
enable building corporate management to easily and promptly access the fire-
affected floor and commence fire fighting.
The work on the use of lifts for emergency evacuation was initiated in the early
1990s at the National Institute of Standards and Technology (NIST, USA) in which
pros and cons were analysed in order to develop suitable guidelines. This research
project examines the feasibility of using lifts along with design modifications as an
alternative facility for a safer and more efficient emergency evacuation. The scope of
this research is limited to apartment buildings where occupant load is low and fire
load is generally confined to dwelling compartment units.
This research project analysed the important issues in relation to the use of lifts for
emergency evacuation. The issues were divided into three categories: human
behavioural response, fire hazards and lift operational mechanism. Output variables
relating to human behavioural response were modelled and analysed as a stochastic
process. Residents’ choice for using evacuation routes was determined using a
survey. The issues of fire hazards (fire, smoke and toxic gases) were analysed for
occupant safety under variable conditions using the concept of fire safety index. The
issues of lift operational mechanism such as lift malfunctioning due to excessive
vi
temperature, electric power failure and water damage were considered for developing
probabilistic models. An integrated approach of risk assessment for the issues of
human behavioural response and fire hazards (such as ‘decision uncertainty’, ‘panic’,
‘nonfatal and fatal injuries’) was developed based on the Multi-Objectives Decision
Analysis method. The results for lift and stair systems were compared and the
feasibility of using lift with design modifications was analysed for alternative designs
and evacuation strategies.
The outcomes of this research have shown that using lifts with a protected lobby for
up to one-fourth of the building population (who may be aged and disabled) has huge
potential as an alternative evacuation facility with enhanced level of safety. Lifts
with protected lobby for one-fourth of the building population showed an improved
level of fire safety from exposure to fire effluents. The reliability of lift operational
mechanism is also improved in protected lift shafts. Lifts with protected lobby for up
to one-fourth of the building population and stairs for up to three-fourth of the
building population showed an improved evacuation safety. The risks in combined
evacuation systems (protected lifts and stairs) are found to be lower when compared
to using stairs or protected lifts. Lifts with double lobby protection (for example, two
levels of compartmentation with fire and smoke doors for lift lobby) showed further
improvements.
This research has proposed alternative designs for lifts and developed models for
analyzing evacuation effectiveness based on risks related to human behaviour, fire
hazards and operational mechanism. It has shown that a combined use of lifts and
stairs has significant advantages. The performance based lift evacuation system is
achievable in apartment buildings. These research findings are based on uncertainty
analysis, which can be further extended to other types of buildings in the future.
vii
ACKNOWLEDGEMENT
I take this special opportunity to express my deep sense of gratitude to Prof. Mahen
Mahendran, Dr. Yaping He, A/ Prof. Gopi Chattopadhya and Mr. B. (Jack)
Williamson, my supervisors, whose invaluable guidance, support and encouragement
have nourished this research project. My wholehearted thanks to my supervisors.
The time spent with Dr. Yaping He at University of Western Sydney is also a
priceless treasure for me, who so patiently and cheerfully goaded and helped me to
complete this research project.
I am particularly thankful to Mr. Christopher R. Odgers, Principal Fire Safety
Engineer from Fire Check Consultants, Queensland for his guidance and immense
assistance related to computer modelling and literature in all phases of this research
project. I also convey my special thanks to Mr. David Mason, University of Southern
Queensland for providing guidance on developing ARENA simulation models. The
officers from NSWFS and QFRS involved in the survey and interviews were fully
supportive at all times.
I wish to acknowledge Mr. R. C. Sharma, Chief Fire Officer for approving my study
leave from Delhi Fire Service for undertaking this research project. I also
acknowledge Mr. S. K. Dheri, Retired Chief Fire Officer, Delhi Fire Service, for
providing me support and guidance.
I express my gratitude and appreciation to my father and my mother for their
unlimited love, support and sacrifices. I acknowledge and recognize my wife Manju
for her moral support and encouragement in all phases of my life and my sons
Aayush and Archit for whom I couldn’t spare my time on many important occasions
during this research tenure.
viii
CONTENTS
STATEMENT OF ORIGINAL AUTHORSHIP .................................................... III
ABSTRACT ................................................................................................................V
ACKNOWLEDGEMENT ...................................................................................... VII
CONTENTS ............................................................................................................VIII
LIST OF FIGURES................................................................................................XIII
LIST OF TABLES..................................................................................................XVI
ABBREVIATIONS/ DEFINITIONS ....................................................................XIX
NOMENCLATURE ...............................................................................................XXI
1. INTRODUCTION .................................................................................................1
1.1 BACKGROUND ................................................................................................1 1.2 PROBLEM STATEMENT....................................................................................3 1.3 RESEARCH STATEMENT (HYPOTHESIS)...........................................................5 1.4 OBJECTIVES OF RESEARCH .............................................................................5 1.5 SCOPE .............................................................................................................6 1.6 RESEARCH METHODOLOGY ............................................................................7 1.7 THESIS CONTENTS ..........................................................................................8
2. LITERATURE REVIEW ...................................................................................10
2.1 CURRENT REGULATIONS IN RELATION TO THE USE OF LIFTS FOR
EVACUATION .............................................................................................................11 2.2 RECENT DEVELOPMENTS IN THE SUBJECT AREA ..........................................13 2.3 STATISTICS ...................................................................................................15 2.4 HUMAN BEHAVIOUR DURING FIRE ...............................................................19 2.5 FIRE AND SMOKE HAZARDS..........................................................................22 2.6 TOXICITY OF FIRE EFFLUENTS ......................................................................29
2.6.1 Mass Loss Models ...................................................................................29 2.6.2 Toxic Gas Models....................................................................................30 2.6.3 Human Incapacitation Model ..................................................................32 2.6.4 Visibility ..................................................................................................33
2.7 LIFT OPERATIONAL SYSTEMS .......................................................................33 2.7.1 Overview .................................................................................................33 2.7.2 Lift Dispatch Control...............................................................................34 2.7.3 Concerns of Lift Operational Mechanism ...............................................35 2.7.4 Lift Protection Measures .........................................................................36 2.7.5 Sprinklers in Residential Buildings .........................................................36 2.7.6 Evaluation of Lift Evacuation Time ........................................................38
2.8 STAIR EVACUATION SYSTEM........................................................................39 2.8.1 Overview .................................................................................................39 2.8.2 Evaluation of Stair Travelling Time........................................................39
2.9 RISK ASSESSMENT METHODS .......................................................................41 2.10 APPLICATION OF COMPUTER MODELS IN THE RESEARCH STUDY .................44 2.11 DISCUSSION AND SUMMARY.........................................................................46
ix
3. RESEARCH METHODOLOGY....................................................................... 49
3.1.1 Risk Identification ................................................................................... 50 3.1.2 Analytical Hierarchical Process for Risk Priorities................................. 53 3.1.3 Acceptable Level of Risk ........................................................................ 62 3.1.4 Hypothetical Building and DTS Provisions ............................................ 63 3.1.5 Selection of a Fire Scenario..................................................................... 65 3.1.6 Concept Design Options.......................................................................... 68 3.1.7 Risk Quantification.................................................................................. 71 3.1.8 Risk Assessment...................................................................................... 71 3.1.9 Selection of Design Options.................................................................... 72
3.2 RESEARCH WORK......................................................................................... 72 3.3 CONCLUSION ................................................................................................ 75
4. STOCHASTIC MODELS OF BUILDING EVACUATION .......................... 76
4.1 INTRODUCTION ............................................................................................. 76 4.2 PILOT SURVEY.............................................................................................. 77
4.2.1 Pilot Survey Overview ............................................................................ 77 4.2.2 Pilot Survey Results ................................................................................ 78 4.2.3 Discussion and Conclusion ..................................................................... 81
4.3 INTERVIEWS ................................................................................................. 83 4.4 ANALYSIS OF BUILDING EVACUATION PERIODS........................................... 85
4.4.1 Methodology ........................................................................................... 85 4.4.2 Discrete Event Simulation....................................................................... 89
4.5 MODEL FRAMEWORK ................................................................................... 91 4.5.1 Hypothetical Building and Parameters.................................................... 91 4.5.2 Lift Supervisory Controller for Lift Simulation Model........................... 93 4.5.3 Lift Simulation Variables ........................................................................ 95 4.5.4 Stair Simulation Variables..................................................................... 100
4.6 SIMULATION MODELS ................................................................................ 103 4.6.1 Lift Simulation Model ........................................................................... 103 4.6.2 Stair Simulation Model ......................................................................... 104
4.7 SIMULATION RESULTS ................................................................................ 105 4.7.1 Lift Simulation Model ........................................................................... 106 4.7.2 Stair Simulation Model ......................................................................... 108
4.8 ANALYSIS OF RESULTS ............................................................................... 110 4.8.1 Lift Waiting Time.................................................................................. 110 4.8.2 Lift Transportation Time and Stair Travelling Time............................. 112 4.8.3 Lift Pre-Evacuation Time and Stair Pre-Evacuation Time.................... 114 4.8.4 Lift Evacuation Time and Stair Evacuation Time................................. 116 4.8.5 Number of Evacuees in Queue .............................................................. 118 4.8.6 Findings ................................................................................................. 119
4.9 MODEL VERIFICATION ............................................................................... 120 4.10 CONCLUSION .............................................................................................. 121
5. FIRE HAZARD MODELS OF LIFE THREATENING CONDITIONS .... 123
5.1 INTRODUCTION ........................................................................................... 123 5.2 ANALYSIS OF FIRE HAZARDS ..................................................................... 124
5.2.1 Effects of Fire Effluents and Evaluation Criteria .................................. 124 5.2.2 Safety Index........................................................................................... 129 5.2.3 Safety Index for Three Locations.......................................................... 131
x
5.2.4 Field Model ‘FDS’.................................................................................133 5.3 MODEL FRAMEWORK AND VARIABLES.......................................................133
5.3.1 Hypothetical Building Model ................................................................133 5.3.2 Concept Designs ....................................................................................136 5.3.3 FDS Model Boundary Conditions .........................................................136 5.3.4 Fire Simulation Scenarios......................................................................138
5.4 FDS MODEL SET UP...................................................................................139 5.4.1 Conventional Domain and Grid System................................................139 5.4.2 Smoke Leakages/ Openings...................................................................141
5.5 FDS OUTPUT ..............................................................................................142 5.6 FDS RESULTS .............................................................................................149
5.6.1 Concept Design A (Unprotected Lift Lobby)........................................149 5.6.2 FDS Results Analysis ............................................................................165
5.7 FED OF SMOKE, ASPHYXIANT AND HEAT ..................................................170 5.7.1 Concept Design A (Unprotected Lift Lobby)........................................170 5.7.2 Summary of FED Results and Analysis ................................................177
5.8 DETERMINATION OF SAFETY INDEX ...........................................................181 5.8.1 Strength Variables ASET ......................................................................181 5.8.2 Load Variables RSET............................................................................182 5.8.3 Safety Index...........................................................................................183 5.8.4 Safety Index Analysis............................................................................183
5.9 CONCLUSION ..............................................................................................184
6. RELIABILITY OF LIFT OPERATIONAL MECHANISM ........................185
6.1 INTRODUCTION ...........................................................................................185 6.2 ANALYSIS OF LIFT OPERATIONAL MECHANISM..........................................186 6.3 METHODOLOGY ..........................................................................................187 6.4 LIFT MALFUNCTIONING DUE TO EXCESSIVE TEMPERATURE RISE...............189 6.5 ELECTRIC POWER FAILURE.........................................................................195
6.5.1 System Descriptions ..............................................................................195 6.5.2 System Boundary Conditions ................................................................196 6.5.3 Data and Statistics .................................................................................196 6.5.4 Fault Tree Analysis................................................................................198 6.5.5 Analysis of Results ................................................................................201
6.6 PROBABILISTIC ANALYSIS OF WATER DAMAGE .........................................202 6.6.1 Water Spread from Fire Protection and Fire Fighting Measures...........203 6.6.2 Complex Parallel and Series System for Water Spread ........................205 6.6.3 Water Spread Result Analysis ...............................................................209
6.7 OUTCOMES .................................................................................................210 6.8 INFLUENCE ON HUMAN BEHAVIOURAL RESPONSE .....................................210 6.9 CONCLUSION ..............................................................................................211
7. RISK ASSESSMENT OF EVACUATION ROUTES....................................212
7.1 INTRODUCTION ...........................................................................................212 7.2 RISK ANALYSIS OF BUILDING EVACUATION SYSTEM .................................213
7.2.1 Assumptions ..........................................................................................213 7.2.2 Methodology – Multi-Objectives Decision Analysis ............................214
7.3 RISK ASSESSMENT......................................................................................216 7.3.1 Identify Concept Design Options and Evacuation Strategies................216 7.3.2 Evaluation Considerations and Evaluation Measures ...........................216
xi
7.3.3 Specify Weights .................................................................................... 218 7.3.4 Value Functions..................................................................................... 222 7.3.5 Sensitivity Analysis ............................................................................... 229
7.4 ANALYSIS OF RESULTS ............................................................................... 231 7.5 CONCLUSION .............................................................................................. 232
8. FEASIBILITY AND DESIGN CONSIDERATIONS.................................... 234
8.1 INTRODUCTION ........................................................................................... 234 8.2 FEASIBILITY OPTIONS................................................................................. 235
8.2.1 Lifts with Protected Lobby to evacuate 25% of the Building Population235 8.2.2 Double Protected Lift Lobby for the Entire Building Population ......... 236
8.3 REDUNDANCY MEASURES .......................................................................... 237 8.3.1 Common Lift and Stair Lobby .............................................................. 237 8.3.2 Refuge Area........................................................................................... 238 8.3.3 Scattered Design of Lift System............................................................ 239 8.3.4 Pressurization ........................................................................................ 240 8.3.5 Smoke Seal in Lift Landing Door ......................................................... 240
8.4 FIRE PROTECTION MEASURES FOR LIFT SYSTEM........................................ 241 8.5 STRATEGIC PLANNING................................................................................ 242 8.6 CONCLUSION .............................................................................................. 242
9. CONCLUSIONS................................................................................................ 244
9.1 SUMMARY .................................................................................................. 244 9.2 RESEARCH FINDINGS .................................................................................. 247
9.2.1 Advancements in Systematic and In-Depth Risk Analysis ................... 248 9.2.2 Contribution to Building Evacuation Strategy ...................................... 248
9.3 SCOPE FOR FUTURE WORK ......................................................................... 249
REFERENCES ........................................................................................................ 251
APPENDIX A ............................................................................................................ 263 International Listing of Major Fires, where Lifts were used............................. 263
APPENDIX B ............................................................................................................ 264 Risk Priorities and Matrix Consistency Ratio ................................................... 264
APPENDIX C ............................................................................................................ 266 Survey Questionnaire ........................................................................................ 266
APPENDIX D ............................................................................................................ 271 Interview............................................................................................................ 271
APPENDIX E............................................................................................................. 275 Occupant Response and Coping Times............................................................. 275
APPENDIX F............................................................................................................. 276 SIMAN Language ............................................................................................. 276
APPENDIX G ............................................................................................................ 283 ARENA Results ................................................................................................ 283
APPENDIX H ............................................................................................................ 304 Verification of ARENA Model ......................................................................... 304
APPENDIX J ............................................................................................................. 308 Building Characteristics, HRR and Temperature.............................................. 308
APPENDIX K............................................................................................................ 311 Occupants’ Movement Causing Door Opening and Closing ............................ 311
APPENDIX L............................................................................................................. 313 Visibility Determination at a Focal Point.......................................................... 313
xii
APPENDIX M............................................................................................................315 Species Concentration and Fractional Effective Doses of Smoke, Gases and Heat ...............................................................................................................315
APPENDIX N ............................................................................................................367 Calculation of Fractional Effective Doses of Smoke, Gases and Heat .............367
.
xiii
LIST OF FIGURES
Figure 1-1 – Fire and Rescue from Pallister Plaissance Apartments, USA .................3
Figure 2-1 – Fire Shaft and Stairs Protected with Lift Lobby (NBS, 2000) ..............12
Figure 2-2 – Fire Fighter Lift is utilized up to one level below the Fire-Affected
Floor (CEN, 2003)..............................................................................................13
Figure 2-3 – Factors Causing Death during Residential Fires (Miller, 2005)............18
Figure 2-4 – Stress Model (Proulx, 1993) ..................................................................20
Figure 2-5 – Fear induced Panic Behaviour ...............................................................21
Figure 2-6 – Piston Effect: Lift acts as a Piston for Smoke Movement .....................25
Figure 2-7 – Smoke Infiltration and Heat Exposure to Lift Landing Door................26
Figure 2-8 – Lift Door after the Fire Test (Bennetts et al., 1999) ..............................26
Figure 2-9 – Temperature inside the Lift Shaft (Bennetts et al., 1999) .....................27
Figure 2-10 – LMR Protected with Sleeves (smoke dissipates through vent) ...........28
Figure 3-1 – Risks involved in the Lift Evacuation System.......................................52
Figure 3-2 – Risks at Three Hierarchical Levels........................................................57
Figure 3-3 – Hierarchical Relationship for the Evaluation of Risk Priorities ............59
Figure 3-4 – Global Risk Priorities ............................................................................61
Figure 3-5 – Typical Floor of a Hypothetical Building (57 m ×××× 20 m)......................63
Figure 3-6 – Positive Pressurisation and Smoke Lobby in a Fire Isolated Exit .........65
Figure 3-7 – Event Tree Analysis for a Worst Possible Path (or Fire Scenario)........67
Figure 3-8 – Comparison of Stairs and Lifts ..............................................................69
Figure 3-9 – Three Concept Designs for risk analysis ...............................................70
Figure 3-10 – Research Work Flow Diagram ............................................................73
Figure 4-1 – Residents’ Awareness of Emergency Evacuation Procedure ................80
Figure 4-2 – Residents’ Experience during Fire Drill ................................................81
Figure 4-3 – Flow Diagram for Analysing the Output Variables of Models .............88
Figure 4-4 – SIMAN Flow Diagram ..........................................................................91
Figure 4-5 – Typical Floor of a Hypothetical Building (57m ×××× 20m)........................92
Figure 4-6 – Lift Controller Logic Diagram used for ARENA Simulation Model....94
Figure 4-7 – Poisson Distribution for Occupant Arrival (Lifts).................................99
Figure 4-8 – Poisson Distribution for Occupant Arrival (Stairs) .............................101
Figure 4-9 – Triangular Distribution for Occupants’ Stair Travelling Time............102
xiv
Figure 4-10 – Various Zones in Lift Simulation Model (two floors only)...............103
Figure 4-11 – Various Zones in Stair Simulation Model (three floors only) ...........104
Figure 4-12 – Animation of Lift and Stair Simulation Models at 300 seconds........105
Figure 4-13 – Lift Waiting Times during the Fire Occurrences at Three Levels.....110
Figure 4-14 – Lift Waiting Times.............................................................................112
Figure 4-15 – Lift Transportation and Stair Travelling Times.................................114
Figure 4-16 – Lift Pre-Evacuation Times.................................................................115
Figure 4-17 – Lift and Stair Evacuation Times ........................................................117
Figure 4-18 – Number of Evacuees in Queue in Lift System ..................................119
Figure 5-1 – Flow Diagram for Calculating Safety Index........................................130
Figure 5-2 – Load Variables for Safety Index for the locations of Lift Lobby, Lift
Shaft and Stair Shaft .........................................................................................132
Figure 5-3 – Typical Floor of a Hypothetical Building for a Generalised Fire
Scenario (dimensions not to scale) ...................................................................134
Figure 5-4 – Wind Speed Profile..............................................................................138
Figure 5-5 – Computational Domain for FDS Model ..............................................140
Figure 5-6 – Three Grid Sizes used in the FDS Model (38th floor view).................141
Figure 5-7 – Smoke Leakage Openings in the Lift Shaft Wall ................................142
Figure 5-8 – Output points in the Fire Compartment, the Lift Lobby, the Lift Shaft
and the Stair ......................................................................................................143
Figure 5-9 – Snapshots of Smoke View and Temperature Contour (Fire Scenario 1)
..........................................................................................................................144
Figure 5-10 – Snapshots of Temperature Contour and Vector slice (Fire Scenario 3)
..........................................................................................................................145
Figure 5-11 – Slice Snapshot of Visibility in the Lift Lobby (Fire Scenario 5).......146
Figure 5-12 – Snapshot of Temperature Contour at 720 seconds (Fire Scenario 5) 147
Figure 5-13 – Slice Snapshots in a Vertical Plane in the Lift Lobby at 600 seconds
(Fire Scenario 5) ...............................................................................................148
Figure 5-14 – Smoke, Gases and Heat in the Lift Lobby, the Lift Shaft and the LMR
(Fire Scenarios 1 to 6) ......................................................................................164
Figure 5-15 – Temperature and CO on the 38th Floor for Unprotected Lift Lobby .166
Figure 5-16 – Temperature in the Lift Shaft (with and without wind).....................167
Figure 5-17 – Temperature in the LMR (without wind) ..........................................168
Figure 5-18 – FED in Fire Scenarios 1 to 6..............................................................176
xv
Figure 6-1 – Probability of Excess Temperature Rise in LMR (Fire Scenarios 1 to 6)
..........................................................................................................................192
Figure 6-2 – Probability Distribution and Consequences of Excess Temperature in
LMR .................................................................................................................194
Figure 6-3 – Typical Electrical System for Essential and Non-Essential Supplies..195
Figure 6-4 – Fault Tree Analysis for Electric Fire in SOU ......................................198
Figure 6-5 – Fault Tree Analysis for Electric Power Failure in Unprotected Lift
Lobby................................................................................................................199
Figure 6-6 – Fault Tree Analysis for Electric Power Failure in Protected Lift Lobby
..........................................................................................................................200
Figure 6-7 – HRR during Three Stages of Water Application.................................203
Figure 6-8 – Complex Parallel-Series System for Probability of Water Spread......207
Figure 6-9 – Complex Parallel-Series System for Quantity of Water Spread..........208
Figure 6-10 – Quantity of Water Spread and Building Evacuation .........................209
Figure 7-1 – Multi-Objectives Decision Analysis Methodology .............................214
Figure 7-2 – Influence Diagram of Building Evacuation Risk Model .....................217
Figure 7-3 – Disadvantages of High Rise Living (Mori and UHK, 2002)...............220
Figure 7-4 – A Value Tree for the Parametric Global Weights ...............................222
Figure 7-5 – Value Functions for Building Evacuation Times ................................224
Figure 7-6 – Sensitivity Analysis for Different Weights Placed on Panic...............231
Figure 7-7 – Risk Values for Concept Design Options............................................232
Figure 8-1 – Evacuation Option 1: One-Fourth of the Building Population using
Protected Lifts and the rest using Stairs ...........................................................235
Figure 8-2 – Evacuation Option 2: Double Protected Lift Lobby............................236
Figure 8-3 – Rectangular Arrangement for a Common Lift Lobby .........................238
Figure 8-4 – Lifts located in the Refuge Area..........................................................239
Figure 8-5 – Smoke Seal in the Lift Landing Door and Wall Frame.......................241
xvi
LIST OF TABLES
Table 2-1: Summary of High-Rise Building Fires in the US for 1998 (Hall, 2001) ..15
Table 2-2: Extent of Fire and Smoke Damage in Apartment Buildings for 1994-98
(Hall, 2001).........................................................................................................16
Table 2-3: Locations of Fires in Apartment Buildings for 1994-98 (NFPA, 1999)...16
Table 2-4: Australian Fires, Fatalities and Nonfatal Injuries for 1993-94 (King, 1997)
............................................................................................................................17
Table 2-5: New Zealand Fire Fatalities for 1997-2002 (Miller, 2005) ......................17
Table 2-6: Causes of Death in New Zealand for 1997-2002 (Miller, 2005) ..............17
Table 2-7: People living in High-Rise Apartment Buildings, Australia (ABS, 2004)
............................................................................................................................18
Table 2-8: Effects of Visibility distance on Turning Back during Smoke in Corridors
(Pauls, 1995).......................................................................................................23
Table 2-9: Performance of Automatic Sprinkler System in Residential Buildings
(Marryatt, 1988)..................................................................................................37
Table 2-10: Risk Assessment Methods ......................................................................44
Table 3-1: The 9-Point Scale (Saaty, 1980) ...............................................................54
Table 3-2: Random Inconsistency Indices (Source: Saaty, 1980)..............................56
Table 3-3: Matrix (2 × 2) for Priorities of Lift Evacuation ........................................60
Table 3-4: Matrix (3 × 3) for Priorities of Psychological Impact...............................60
Table 3-5: Matrix (3 × 3) for Priorities of Physiological Impact ...............................60
Table 3-6: Global Risk Priorities................................................................................61
Table 3-7: Consistency Tests of Matrices ..................................................................62
Table 3-8: Lift and Stair Systems for Comparison.....................................................69
Table 4-1: Residents’ Age Distribution......................................................................78
Table 4-2: Residents’ Use of Lifts as Normal Access and Egress Routes .................79
Table 4-3: Residents’ Preferred Exit Routes during Fire Emergencies......................80
Table 4-4: Model Parameters .....................................................................................93
Table 4-5: Lift Time Periods and Number of Evacuees in Queue (2nd floor fire)....106
Table 4-6: Lift Time Periods and Number of Evacuees in Queue (19th floor fire) ..106
Table 4-7: Lift Time Periods and Number of Evacuees in Queue (38th floor fire) ..106
xvii
Table 4-8: Lift Time Periods and Number of Evacuees in Queue (2nd floor fire) –
25% population.................................................................................................107
Table 4-9: Lift Time Periods and Number of Evacuees in Queue (19th floor fire) –
25% population.................................................................................................107
Table 4-10: Lift Time Periods and Number of Evacuees in Queue (38th floor fire)–
25% population.................................................................................................108
Table 4-11: Stair Time Periods (2nd floor fire).........................................................108
Table 4-12: Stair Time Periods (19th floor fire) .......................................................109
Table 4-13: Stair Time Periods (38th floor fire) .......................................................109
Table 4-14: Stair Time Periods (2nd
floor fire) – 75% population ...........................109
Table 4-15: Stair Time Periods (19th floor fire) – 75% population ..........................109
Table 4-16: Stair Time Periods (38th floor fire) – 75% population ..........................109
Table 4-17: Means and Standard Deviations of Output Variables...........................120
Table 4-18: Verification of ARENA model for Lifts and Stairs..............................121
Table 5-1: Description of Fire Simulation Scenarios ...............................................139
Table 5-2: Time to Exceed Tenability Limits in Lift Lobby....................................177
Table 5-3: Time to Exceed Tenability Limits in Lift Shaft and LMR .....................178
Table 5-4: Time to Exceed Tenability Limits in Stair Shaft ....................................179
Table 5-5: Means and Standard Deviations of ASET (lift lobby)............................181
Table 5-6: Means and Standard Deviations of ASET (lift shaft) .............................181
Table 5-7: Means and Standard Deviations of ASET (stairs) ..................................182
Table 5-8: Means and Standard Deviation of RSET................................................182
Table 5-9: Safety Indices for Lift and Stair Evacuation...........................................183
Table 6-1: Temperatures and their Impact on Lift Systems.....................................190
Table 6-2: Probability of Excess Temperature Occurrence in LMR in Fire Scenarios
..........................................................................................................................193
Table 6-3: Unavailability of Lifts to Building Population .......................................194
Table 6-4: System Boundary Conditions .................................................................196
Table 6-5: Impact on Lift System.............................................................................201
Table 6-6: Probability of Water Spread at Three Levels..........................................206
Table 7-1: Concept Design Options and their illustrations ......................................216
Table 7-2: Risk related Parameters ..........................................................................218
Table 7-3: Parametric Values and Weights relating to Building Evacuation...........221
xviii
Table 7-4: Matrix (3 × 3) for Priorities Risk Factor (p3) .........................................225
Table 7-5: Parameter Strengths for Concept Design Options ..................................228
Table 7-6: Summary of Assigned Values.................................................................228
Table 7-7: Total Risk Values for Concept Design Options......................................229
Table 7-8: Total Risk Values from Analyses based on a Zero Weight for Panic.....230
Table 7-9: Total Risk Values from Analyses based on 100% Weight for Panic......230
xix
Key Words: Apartment building, Aged, Disabled, Emergency exit, Evacuation, Fire,
Fire safety, High-rise building, Hazard, Injury, Life safety, Lift, Modelling, Panic,
Risk analysis, Reliability, Smoke, Stochastic model, Tenability limits, Toxicity,
Uncertainty.
ABBREVIATIONS/ DEFINITIONS
Aged: Occupants more than 65 years in age.
Alternative evacuation facility: Alternative evacuation facility such as lift is to
comply with the performance requirements of BCA, other than the DTS provisions
such as stairs.
Apartment building: A building containing two or more sole occupancy units, each
being a separate dwelling (BCA classification).
AHP: Analytical Hierarchy Process.
ASET: Available Safe Evacuation Time.
BCA: The Building Code of Australia 2005. {A reference of BCA 2005 is quoted in
this thesis as this research was undertaken in the year 2004. However, BCA 2007 is
currently enforced and there is no substantial change in evacuation facilities.}
Coping time: Time for occupants’ coping activities those resulting from an occupant
perceiving that fire poses an actual threat to the point where the occupant initiates
evacuation or avoidance activities.
Disabled: Occupants physically incapable of evacuating building using stairs.
DTS: Deemed-to-satisfy Provision of BCA 2005.
EEP: Emergency evacuation procedure.
Evacuation and egress: The term “evacuation” is used for emergency escape from
the building whereas “egress” is used for non-emergency exit only.
Fractional effective dose (FED): FED refers to incapacitation, lethality or endpoint.
When not used with reference to a specific asphyxiant, the term FED represents the
summation of FEDs for all asphyxiant in a combustion atmosphere.
Fire door (FD): Door designed to contain fire for a nominated period while
facilitating emergency egress. Fire doors are rated in terms of structural adequacy,
integrity and insulation.
High-rise buildings: High-rise buildings are 25 m in effective height or above. The
effective height is measured from the lowest level of fire brigade access to the
xx
highest occupied level (BCA classification). The term “mega high-rise” is used for
buildings having 50 storeys or more.
Incapacitation: State of physical inability to accomplish a specific task.
Lift and elevator: Lift is the UK nomenclature; Elevator is the US nomenclature.
LMR: Lift Machine Room.
MODA: Multi-Objectives Decision Analysis.
PDF: Probability Distribution Function.
POE: Probability of Occurrence.
Protected shaft and unprotected shaft: Protected shaft is provided with protected
lift lobby having fire resistive construction and fire door whereas unprotected shaft is
unprotected lift lobby without fire resistive construction and fire doors.
Refuge area: Refuge area is provided in high-rise building as a temporary shelter for
occupants during fire emergencies.
Response time: Time for occupants’ response from occurrence of detectable cue to
activities involved in responding to those cues.
Round trip time (RTT): RTT is the time taken by lift after stopping at entrance
lobby for performing the entire trip after serving the most probable number of floors.
RSET: Required Safe Evacuation Time.
Smoke alarm: A smoke alarm complying with Australian Standard AS 3786. This
alarm is self-contained including an audible warning device and is not connected to
the building fire detection and alarm system.
Smoke detector: A smoke detector complying with AS 1670 and connected to the
building fire detection and alarm system.
Smoke door (SD): Door with a closely fitting leaf or leaves, furnished with smoke
seals and an approved sensing device, release mechanism and closing mechanism, to
protect openings in partitions against the passage of smoke during a fire.
SOU: Sole Occupancy Unit – a unit in an apartment building.
Uncertainty: Two type of uncertainties – uncertainty due to lack of fundamental
knowledge about specific factors or pathways; uncertainty due to variability
(stochastic uncertainty, randomness) in data.
Tenability limits: Tenability limits are the safe environment conditions during
which occupants can perform designated actions. When the limit exceeds, occupants
are considered incapacitated due to hazardous conditions or exposure.
Vertical transportation: Vertical transportation means lift transportation.
xxi
NOMENCLATURE
AE Total leakage area from the lift shaft m2
aij Element (or attribute) of matrix -
cp Specific heat of gas kJ/kg.K
Cs Extinction coefficient 1/m
D Physical fire diameter m
D* Characteristic fire diameter m
Fs Nominal evacuee flow in stairs person/m/s
g Acceleration due to gravity m/s2
h Height above the ground for wind m
J Number of lifts -
k Exponent for the wind velocity -
L Stochastic parameter for load variable -
m Number of lift round trips -
n Number of floors -
N Number of people per floor person
p Parameter measure -
PE Pressurisation level in the pressurized space Pa
q Radiant heat flux kW/m2
Q Heat release rate kW
Q* Heat release rate kW
QE Air supply to the pressurized space m3/s
Q Water discharge of fire protective system litre
RA Acceptable Risk -
S Stochastic parameter for strength variable -
t1 Egress time (congestion) second
ta Elevator evacuation start up time second
tn Egress time (free walk) second
to Travel time from the elevator lobby to the outside second
tr,j Time for round trip second
ts Walking time between adjacent floors second
xxii
tLPM Lift pre-movement time second
tLW Lift waiting time second
tLT Lift transportation time second
tLI Lift intermittent floor movement time second
tFD Fire detection time second
tLM Occupant movement time for lift second
tLPE Lift pre-evacuation time second
tSPM Stair pre-movement time second
tST Stair travelling time second
tSI Stair intermittent floor movement time second
T Air temperature C
T∞ Ambient temperature K
TLMR Temperature in LMR C
vi Parametric value -
V Visibility distance m
Vh Wind speed at a height m/s
Vr Wind speed at the reference height m/s
W Effective width of the stairs m
wi Parameter weight -
x Number of evacuees person
Greek symbol
Mµ Mean of safety margin -
Mσ Standard deviation of safety margin -
β Safety index -
volβ Degree of voluntariness -
α Fire growth coefficient kW/s2
η Lift trip inefficiency %
λ Occupant arrival rate person
λ Eigenvalue -
ρ∞ Density at ambient temperature kg/m3
1
1. INTRODUCTION
1.1 Background
In the past stairs was the only emergency evacuation facility in high-rise buildings.
Alternative safe evacuation facility has been of interest to many researchers. Over the
past two decades, the works of Klote (1982) and many others (Pauls, 1977, Pauls et
al., 1991) have led to a better understanding of many problems associated with the
use of lifts during fire emergencies. Subsequent to the WTC twin tower collapse on
September 11, 2001, researchers have focused their efforts on the use of lifts as an
alternative evacuation facility in buildings (Kuligowski, 2003, Kuligowski and
Bukowski, 2004). Analysis of 480 first-person accounts in WTC showed that 81%
used the stairs, 6% used the lifts, and 13% used a mix of lifts and stairs (Bill, 2002).
Two-third of the survivors from floors above the 78th floor sky lobby (an area where
people changed lifts) and one-third from floors between the sky lobbies on the 44th
and 78th floors used lifts. A total of 2605 lives was lost. Some researchers believe
that if adequate emergency escape lifts were available at the WTC, more occupants
could have been evacuated in a shorter period of time and the extent of the fatality
might have been greatly reduced (So and Yu, 2003). There were incidents where
stairs became unsafe due to the presence of smoke, and occupants therefore used lifts
as a faster mode of evacuation (Bill, 2002). This caused a great concern among the
fire safety community and generated the need to review the relevance of existing
evacuation strategy for high-rise buildings.
Use of lifts is not recommended for building occupants for evacuation purposes due
to some of the major incidents. However, past records show that the occupants do
use lifts for emergency evacuation (see Appendix A). In some cases, evacuating
occupants were trapped in the lift car and exposed to hot smoke resulting in
unconsciousness and deaths (Barnett et al., 1992). One of the worst cases was in the
MGM Grand Hotel, Las Vegas on November 21, 1980, where fire from ground floor
restaurant rapidly spread to the upper floors through air systems, stairways and lift
shafts. The fire caused 85 fatalities due to smoke inhalation and many were found in
lifts (Bryan, 1982). The occupants’ decision to use lifts can be perceived as an
alternative during life threatening conditions as the occupants may feel that their
2
main exit routes are closing rapidly due to smoke and toxic gases and their search for
main exit route would be a waste.
Whilst alternative evacuation facilities are explored for safe evacuation from
buildings, evacuation using lifts during fire emergencies is a controversial issue
because the safe operation of normal lift is not always ensured. The major concerns
are believed to be loss of power supply, smoke infiltration, lift passing through the
fire-affected floor, lift car being exposed to heat radiation, water run off from fire
fighting operation, malfunctioning of lift equipment, electric short circuit of re-call
button, occupant panic and overloading (Klote, 1982). The use of lifts for an
emergency exit is viewed as an alternative in high-rise buildings since the occupants
can have options for an alternative evacuation route. Building management will also
use the lifts for prompt access to the fire-affected floor.
The benefits of using combined lift and stair evacuation have been demonstrated by
many researchers. In a study conducted by Klote et al., (1993a), it was found that the
use of both lifts and stairs can reduce evacuation times by as much as 50% over the
use of stairs alone. If lifts are used in mega high-rise buildings during an emergency
situation, evacuation times can be reduced to 15-30 minutes instead of 2-3 hours
(Siikonen and Hakonen, 2003). Greater evacuation efficiency occurs as the height of
the building increases. A similar result was found by Andersson and Wadensten
(2000) in their simulation of the One Canada Square building at Canary Wharf in
London. Evacuation time was reduced with the use of stairs and lifts. It was also
noted that the response time and coping time during an emergency was up to two
thirds of the total time to evacuate a building. With the use of stairs and lifts, the
building can be evacuated promptly with significantly reduced evacuation time.
However, this requires extensive research work in view of the uncertainty in fires,
life threatening conditions arising from fire effluents and effects on human
physiological and psychological behavioural response.
This research project began by identifying issues relating to the lift safety and was
aimed at determining the feasibility of using lifts (with suitable design modifications)
as a safe and efficient evacuation facility in apartment buildings during fires and
other emergencies. The research can be extended to other types of buildings.
3
1.2 Problem Statement
Australian building fire statistics show that there were 12.9% apartment fires
resulting in 12.5% fatal injuries and 18% nonfatal injuries of total residential fires
during the year 1993-94 (King, 1997). A total of 1547 fires were reported in
apartment buildings causing 8 fatalities and 133 nonfatal injuries. A recent study
showed that 32.1% of the residential fire deaths consisted of aged less than 15 years
and 25.9% consisted of aged 60 years or above during the period of 1997 to 2003
(Miller, 2005). Thirteen percent of the population living in high-rise apartment
buildings in Australia are aged 65 years or above (ABS, 2004). This pattern is also
reflected in other international studies. A study in Japan found that aged (65 years
and above) accounted for 47.8% of residential fire deaths (Sekizawa, 1991). The risk
for disabled is five times higher than the average population. In New Zealand, 26.5%
of the residential fire deaths consisted of aged persons during the period of 1991 to
1996 (Duncanson et al., 2000). Twelve percent of the population are aged in New
Zealand (Dunstan and Thomson, 2006). These data show that children (less than 15
years) and aged persons (more than 60 years) are more prone to risks in residential
fires than those from other age groups.
Figure 1-1 – Fire and Rescue from Pallister Plaissance Apartments, USA
(Fire caused 3 causalities in a 12 storey building fire (188 units); occupants were
rescued from windows due to heavy smoke in evacuation routes; April 2000, Source:
The Detroit News)
The aged and disabled persons are slower to evacuate through narrow stairs with an
average walking speed of 0.43 m/s (Proulx, 1995). They might adversely affect the
evacuation of other occupants. Due to their slow movement, the problem of using
4
stairs, such as bottleneck, queuing and stampede during emergency evacuation can
be encountered. Uncertainty and anxiety can be expected amongst the aged and
disabled occupants during building evacuation (Proulx, 1995). In one of the buildings
studied by Proulx (1995), many elderly and mobility impaired residents were
unsettled by the fire alarm, not knowing what to do. A woman in a wheelchair
entered the main staircase blocking the way for descending occupants, threatening
that she would go down the stairs with her wheelchair. Residents took her back to
the corridor and stayed with her until the arrival of fire fighters.
Subsequent to the WTC bomb explosion on February 26, 1993, Juillet (1993)
conducted an interview of 27 disabled and impaired persons in one of the towers.
The evacuation time with the assistance of other occupants or emergency personnel
was reported to range from 40 minutes to over 9 hours with an average of 3.34 hours.
Some of the permanently disabled residents were not able to evacuate the building
without aid. Pauls (1977) determined that 3% of occupants in high-rise buildings in
Canada were unable to use stairs due to their permanent or temporary immobility.
Evacuation time increases with the number of stair flights. Long evacuation times
might endanger the life of evacuees due to tiredness, dizziness, slipping on surfaces
or becoming less capable physically. Evacuees normally experience fatigue after
about five minutes of travelling downstairs (So and Yu, 2003). Research has also
shown that the evacuees will begin to suffer fatigue when they have travelled about
18 storeys (So and Yu, 2003).
Building occupants often add complexity and danger to their evacuation process by
remaining in the building until the danger becomes severe. Also, prior experiences
with false alarms result in occupants attempting to validate the alarm before starting
the evacuation (Allen, 1995). Precious time is lost during actual fire emergencies.
Sometimes hazardous conditions arise quickly and in such circumstances occupants
may make irrational decisions when searching for the escape route (Barnett et al.,
1992).
Fire fighters also need a safe, fast and reliable mode of transport system because they
are not expected to climb stairs to much higher levels with heavy gear. Fire fighters’
5
movement can be delayed due to evacuees’ using stairs (Kuligowski, 2003).
Presently fire fighters’ lifts are permitted to be used up to one level below the fire-
affected floor depending upon the safety of lift evacuation system.
There is an increasing trend of high-rise apartment building construction in Australia
(ABS, 2005). The 323 m high Q1 tower with 78 storeys was built in Australia
during 2005 and is the tallest residential tower in the world (Emporis, 2005).
Another such tower in Australia is the 297 m high Eureka tower with 88 storeys
(Emporis, 2005). Although stairs are the main evacuation facility in high-rise
buildings, the use of lifts is to be considered as an alternative evacuation facility for
the aged, motion impaired and physically weak people in this study.
1.3 Research Statement (Hypothesis)
Occupants in high-rise apartment buildings need safe and efficient evacuation
facilities during fire emergencies. Currently, stairs are the only evacuation facility
permitted by building regulations for high-rise buildings because lifts are not
considered safe in fire emergencies. The risks associated with the use of lifts are
manageable. The use of lifts for emergency evacuation is feasible. If lifts are found
to be safe for emergency evacuation, occupants including the aged and disabled will
have an alternative evacuation system and can safely evacuate the buildings. This
will reduce the injuries considerably.
1.4 Objectives of Research
This research explores the feasibility of using lifts as a safe alternative evacuation
facility in apartment buildings. The main objective of this study is:
To study if lifts (elevators) provide an acceptable means for evacuation in apartment
buildings greater than 25 m in effective height during fire and other emergencies.
The specific objectives of this research are:
6
1. To identify the risks associated with the use of lifts for emergency evacuation and
develop an inter-relationship among the risks.
2. To establish a research strategy for an acceptable level of risk and consider
suitable design options and evacuation strategies.
3. To investigate a suitable risk assessment approach for lift evacuation system.
4. To gain a better understanding of residents’ preferred access and egress routes
and expert’s opinion on the use of lifts.
5. To develop a model for the building evacuation under uncertainties associated
with human social, behavioural and physical movement (with a priori heuristics
of the lift domain) and determine the probable time for safe evacuation.
6. To develop a model for the fire and toxic hazards under variable conditions and
determine the probable time when evacuees are predicted to become
incapacitated during exposure of fire effluents in evacuation routes.
7. To determine the reliability of lift operational mechanism and the feasibility of
reliability improvement for design options.
8. To assess the risks in building evacuation systems for a comparative analysis.
9. To study the feasibility of using lifts as an alternative evacuation facility and
propose redundancy measures for the safe and efficient lift evacuation system.
1.5 Scope
The scope of this research is to determine the feasibility of lifts as an alternative
evacuation facility in apartment buildings greater than 25 m (high-rise) in effective
height. This system could also be applied to non-fire emergency evacuation (like
bomb threat, gas leakage or similar calamities). The existing and modified lift
designs are to be considered in determining the safe lift evacuation system.
The scope of this research does not include buildings less than 25 m (low-rise) since
the benefits of lift evacuation will be far outweighed due to cost considerations and
occupants will feel comfortable going down the stairs rather than waiting for a lift
(Smith, 2003).
7
1.6 Research Methodology
The risk in lift evacuation system involves complexity, represented by multiple
attributes and requiring diverse sources of evidence to demonstrate its achievement.
The research methodology used in this project involves the following components:
• To identify all the significant risks in the lift evacuation system and develop a
relationship among the risks.
• Rank all the risks (risk priorities) in terms of likelihood of occurrence and
expected impact on the building evacuees.
• Establish a research strategy for an acceptable level of risk.
• Identify risk control design options and evacuation strategies for evaluating
risks.
• Quantify consequences with the models and techniques.
• Conduct risk assessment with a suitable method/ technique.
• Select appropriate risk control design options.
A relationship is developed among the risks related to the use of lifts for emergency
evacuation. The relationship identifies the key issues to be addressed. The issues of
human behavioural response, fire hazards and lift operational mechanism give rise to
three risks i.e. decision uncertainty, panic and injuries (nonfatal and fatal). The risks
of “decision uncertainty, panic and injuries” may be interrelated, which may
ultimately lead to psychological or physiological impact. These related risks form a
complex process for risk assessment. The research strategy involves risk
management by reducing the level of consequences to an acceptable level. The risk
consequences are reduced by considering design options and evacuation strategies.
The risks are quantified from building evacuation simulation models, fire hazard
models and probabilistic risk models.
Stochastic model of building evacuation, fire hazard model of life threatening
condition and probabilistic analysis of reliability are developed for quantifying the
parametric values of risks. The technical parameters are multi-dimensional. The
direct measurement of risk using the traditional approach (i.e. Risk = Probability of
8
Occurrence × Severity of Consequences) involves complexity. An indirect evaluation
procedure based on Multi-Objectives Decision Analysis (MODA) approach is used.
This risk decision analysis approach constitutes the process of analysing and scoring
the parameters. The parameters are given weights based on data generated from the
statistics and simple analytical techniques are used. The priorities of conflicting key
issues (risks) are assigned with the help of Analytical Hierarchy Process (AHP). The
results of the analysis are theoretically sound and justified for making suitable
decisions. This research methodology is elaborative and requires identification of
various key issues from the literature review. Therefore, full details of the research
methodology are presented in a separate chapter (see Chapter 3).
1.7 Thesis Contents
This thesis consists of nine chapters followed by references and appendices.
Followed by the Introduction chapter, the contents of the thesis are:
Chapter 2 reviews the literature on the current trends in fire emergency evacuation
systems for high-rise buildings and human behaviour in case of fire, tenability limits
of fire and smoke, risk assessment techniques and related models.
Chapter 3 presents the details of the research methodology used in this project.
Details of the research methodology based on Multi-Objectives Decision Analysis
(with Analytical Hierarchy Process) on a comparative basis are given. Inter-
relationships of identified risks are presented. Design options and evacuation
strategies are identified. Research strategy based on reducing risk consequences to an
acceptable level is described. (Objectives 1, 2 and 3)
Chapter 4 presents the stochastic models for building evacuation systems (lifts and
stairs). The stochastic models are developed using SIMAN ARENA software. Lift
time periods, stair time periods and existence of queuing occupants are derived as
output variables for risk assessment. The results of a survey and interview are also
presented. (Objectives 4 and 5)
Chapter 5 analyses the fire and smoke hazards under variable conditions. Tenability
limits relating to temperature, smoke, toxic gases and visibility are considered in
9
modelling. The concepts of ‘fractional effective dose’ and ‘safety index’ are
presented for evaluating evacuees’ safety. (Objective 6)
Chapter 6 analyses the reliability of lift operational mechanisms. Risks associated
with water spread, malfunctioning of lifts and electric power failure are modelled and
analysed using probabilistic techniques, i.e. complex event tree and fault tree
analyses. (Objective 7)
Chapter 7 provides an integrated model for risk assessment based on the Multi-
Objectives Decision Analysis method. The model is proposed to comprehensively
evaluate all the risks associated with the lift system for the design options and
evacuation strategies. (Objective 8)
Chapter 8 demonstrates the feasibility of designs for lift evacuation systems based on
conceptual design options and evacuation strategic planning. Redundancy measures
are proposed. (Objective 9)
Chapter 9 presents the summary, research findings and scope for future work.
10
2. LITERATURE REVIEW
Over the last decade or so there has been a proliferation of interest on the topic of lift
evacuation system. Lifts used in the early of 20th century lacked many fire safety
features. For example, many lifts were located in open shafts, which acted as a
chimney during fires (Martin, 2003). Some lifts were not provided with fire resistive
doors and car cabins were exposed to risk areas. Lift doors were not fire rated and
they could transmit heat and smoke to the lift shafts and to other floors. These lifts
were considered unsafe for fire evacuation from the viewpoint of operational safety.
Major concerns of such lifts were loss of power supply, smoke infiltration, lifts
passing through danger zones and lift car being exposed to heat radiation, lift
software failure, electric short circuit of lift re-call button, damage by water run off
from fire fighting operation, occupant panic and lift overloading (Klote, 1982).
Whilst most lifts adhere with the national and international safety norms, minor
modifications in lift design can place lifts as the next available option of safe
evacuation facility in emergencies. Research and development work are already
under way in the US. Evidences from literature and preliminary research work
demonstrated that design of a lift evacuation system for a small number of people is
feasible (Klote et al., 1995). A lift evacuation system would be most beneficial for
the disabled in office buildings and for all residents in luxury apartment buildings
due to low occupancy load (Klote et al., 1995). However, the use of lifts for large
number of people was not thought practical at the time, due to the required
complexity of the system (Klote et al., 1995). Pauls et al. (1991) suggested that
evacuation via lifts should only be an option for occupants who can not use the stairs.
A literature survey was conducted to investigate the risks associated with the use of
lifts and need for the lifts as an alternative emergency evacuation facility in
buildings. The literatures were obtained from various search engines and the
National Institute of Standards and Technology (NIST) website. The use of lifts for
emergency evacuation is a multifaceted topic and therefore the results of the
literature review are organised into different categories as given next.
11
2.1 Current Regulations in Relation to the Use of Lifts for Evacuation
Lifts can be broadly classified as unprotected (normal) lift and protected emergency
(fire fighter) lift. Internationally recognized ASME A17.1 (2000) does not
recommend the use of normal lifts in fire emergencies and it also does not elaborate
precautionary measures against fire. However, the US NFPA 101 (2000) Life Safety
Code includes the provision of a normal lift as a secondary means of evacuation
facility for air traffic control towers only. The provision is permitted due to the small
footprint of the building, where the construction of two remote stairs is not possible
and moreover the building is not accessible to the general public. An example of a
structure that uses lifts as the secondary means of evacuation is the Stratosphere
Tower in Las Vegas, USA.
In a survey conducted by the International Organization of Standardization (ISO/TR
16765, 2003), it was determined that there is a specific requirement for fire fighter
lifts, in at least 12 countries including the US, UK, Japan and Australia. The fire
fighter lifts are recommended for use by fire fighters and not by the general public
during fire emergencies. This is particularly important in high-rise buildings, where
the carriage of heavy fire fighting gear takes time and uses valuable resources
(Degenkolb, 1991). An example of this was the 62 storey First Interstate Bank fire in
1988 in Los Angles, where lifts were not used, and a fire on the 12th floor required
100 fire service men to carry equipment up the stairs (Degenkolb, 1991) and took
more than 3 ½ hours to control the fire. The scenario might have been worse if the
fire had occurred on the upper levels.
The Building Code of Australia (ABCB, 2005) mandates the installation of
emergency lifts in all the buildings which have an effective height of more than 25m
and buildings in which patient care areas are located at a level that does not have
direct egress to a road or open space. The emergency lifts contain the provisions for
fire brigade control and facilities for people with disabilities.
The US Building Code mandates the installation of fire fighter lifts for fire fighting
and rescue purposes. ASME A17.1 (Safety Code for Elevators and Escalators)
recommends lift operation in two phases during fires (ASME A17.1, 2000). In
12
Phase-1, smoke detector senses the fire and sends a signal to the lift control panel.
All the lifts move to the ground floor and halt, resulting in no lift operation by the
public. In Phase-2, the fire brigade arrives at the building and manually overrides the
lift operation for rescue and fire fighting operations. Fire brigade personnel can use
the fire fighter lifts for evacuating the aged and disabled persons and use them for
access and egress for fire fighting purposes.
The UK Building Code (NBS, 2000) requires fire fighting shaft for buildings, which
have occupied space at more than 18 m above and/or 10 m below the fire brigade access
level. Firefighting shafts incorporate a firefighting lift that opens into the lobby (see
Figure 2-1). The lift has a back-up electrical supply and car control overrides. The
primary function of the lift is to transport firefighting personnel and their equipment
to the scene of a fire with the minimum amount of time and effort. They enable
firefighting operations to start quickly and in comparative safety by providing a safe
route from the point of entry to the floor where the fire has occurred. It may also be
used to help evacuate less mobile people in the event of fire, provided the evacuation
is supervised and managed.
Figure 2-1 – Fire Shaft and Stairs Protected with Lift Lobby (NBS, 2000)
The CEN (2003) allows the use of lifts one level one level below the fire affected floor
(see Figure 2-2). In such circumstances, fire fighters must be aware of the location of
fire affected area.
13
Figure 2-2 – Fire Fighter Lift is utilized up to one level below the Fire-Affected
Floor (CEN, 2003)
2.2 Recent Developments in the Subject Area
During fires, lifts are taken out of service and people are advised not to use lifts.
Deliberations over the years resulted in varying philosophies toward the use of lifts
for emergency evacuation. The American Society of Mechanical Engineers (ASME
International) has organised several symposiums, where lift industries responded to
the inquiries of using lifts for emergency evacuations. Based primarily on influences
from the fire fighting communities, the emphasis has always been placed on the use
of lifts for fire fighting only, and not for general public use (Koshak, 2003).
A workshop on the “Use of Elevators in Fires and Other Emergencies” was held in
March 2004 in Atlanta, Georgia (NIST Special Publication 983, 2003). This
workshop was co-sponsored by ASME International, NIST, International Code
Council (ICC), National Fire Protection Association (NFPA), US Access Board, and
the International Association of Fire Fighters (IAFF). The goal of the workshop was
to develop concrete proposals after considering ‘pros and cons’ for the use of lifts
during fire emergencies. The ‘pros’ include the advantages of lift systems, whereas
‘cons’ are the risks associated with their use during fire emergencies.
14
The consensus of workshop attendees was that the lift operation should work only
until Phase-1 goes into effect. However, the lift operation is not feasible to control
prior to Phase-1. The workshop attendees were of the opinion that the building codes
should have scope for this operation. They also made recommendations to the
ASME A17 Emergency Operation Subcommittee for further research addressing the
technical issues and develop performance requirements for lift evacuation system
during fire emergencies.
NIST funded a research project on “Analysis of the life safety consequences of
smoke migration through elevator shafts” conducted by Klote (2003). The study was
confined to the smoke movement in office buildings with the current infrastructure of
lift system and was conducted with the help of fire and smoke simulation modelling
with CFAST (Peakcock et al., 2004) and CONTAMW (Walton, 1993). The results
showed that unsafe conditions arise on upper levels through lift shafts during a fully
developed fire. However, sprinkler controlled fire reduced the unsafe conditions in
the buildings. The occupants’ evacuation aspects were not considered in this study.
The author also stated that future research is needed to evaluate the extent to which
compartmentation failure would impact smoke flow through lift shafts (Klote, 2003).
He also stated that the approach of the computer models (CFAST and CONTAMW)
was cumbersome and yielded questionable results for scenarios involving reverse
stack effect.
The City University of Hong Kong has also taken initiatives to explore the possibility
of using lifts in super high-rise buildings. In an article published in the Fire
Prevention Journal, So and Yu (2003) stated that all the problems could be solved
without too much difficulty by increasing the capital cost of the system but the
complex psychological reaction of the evacuees could be a major obstacle. They
argued that lifts could be used for emergency escape provided evacuation must be as
quick as possible, residents must feel safe to wait for lift services and at every stop a
lift car must have space to accommodate waiting passengers. However, it would be
difficult in practice to satisfy all the stated premises simultaneously.
A performance-based fire engineering approach was used in the design of the 88-
storey Eureka Tower, Melbourne (Aloi and Rogers, 2002). A lift evacuation strategy
15
was used in the building design for emergency exit. The lift arrangement was
stacked into vertical evacuation zones. This led to an additional evacuation facility
for the occupants, who can evacuate via stairs within the fire-affected zone until they
reach the next lift transfer level (on levels 24 and 52). At transfer floors, occupants
may use express lifts to the ground floor. Occupants with disabilities are assisted by
fire fighters in their dedicated lifts within the zone of fire-affected floor. (In tall
buildings, lift zoning is made for a group of floors for providing efficient lift service
to the occupants). This gave a scope for further research for extending the lifts for
other occupants within zones of fire-affected floors.
2.3 Statistics
A series of significant fires over the years has demonstrated the danger to the
occupants in high-rise apartment buildings. In the United States, over 80% of all fire
deaths occur in residential occupancy and about 20% of these fatalities occur in
apartment buildings. Aged people are mainly at threat in high-rise apartment
buildings. Sekizawa (1991) studied the death pattern of residential fires and found
that people 65 years and older account for 47.8% fire deaths. The risk of death in a
residential fire for disabled residents is 5 times higher than the average population.
The risk for a person greater than 65 years of age and bedridden is 41 times higher
than the average population. According to Hall (2001), apartment occupancies have
experienced the highest frequency of fires, deaths and injuries in high-rise buildings.
The statistics for fires by the type of high-rise occupancy in the US during the year
1998 are shown in Table 2-1.
Table 2-1: Summary of High-Rise Building Fires in the US for 1998 (Hall, 2001)
16
Apartment buildings pose the biggest fire problem in the US and therefore the
occupants of these buildings are at the highest risk. Due to the presence of fire
fighting systems, the fire itself can be limited to the room of origin, however, smoke
spreads far beyond and presents risk to the occupants. Researchers found that 14%
of fires in apartment buildings resulted in smoke propagation beyond the origin of
the fire-affected floor during the year 1994-98 (see Table 2-2).
Table 2-2: Extent of Fire and Smoke Damage in Apartment Buildings for 1994-98
(Hall, 2001)
Brennan’s (1999) analysis of US NFIRS fire statistics over a period of 10 years
estimated that the percentage of fatalities outside the room of origin in evacuation
routes was approximately 15%. Occupants were reported to have died due to smoke
in stairs and lifts. The typical area of origin of fires in high-rise apartment buildings
is shown in Table 2-3. Most fires occurred in the kitchen, followed by bedrooms.
Table 2-3: Locations of Fires in Apartment Buildings for 1994-98 (NFPA, 1999)
Limited data available on Australian building fires shows that fire in residential
properties accounted for 62.4% of all the structural fires during the year 1993-94.
Among residential fires, 12.9% were apartment/ flat fires resulting in 12.5% of total
fatalities and 18% of total nonfatal injuries as shown in Table 2-4.
17
Table 2-4: Australian Fires, Fatalities and Nonfatal Injuries for 1993-94 (King, 1997)
The details of fire related fatalities in New Zealand for the period 1997-2002 are
shown in Table 2-5. Average annual residential fire fatalities are 21.8.
Table 2-5: New Zealand Fire Fatalities for 1997-2002 (Miller, 2005)
The causes of deaths for 115 victims in New Zealand for the period 1997-2002 are
shown in Table 2-6 (Miller, 2005). The cause of death relates directly to three
general areas of fatal effects – consequences of exposure to fire (burns, thermal
injuries to airways and incineration), inhalation of toxic products of combustion
(smoke, CO2, CO, and other poisonous gases, hypoxia and asphyxia) and shock from
injuries that precipitate death from pre-existing health conditions (cardiac failure and
respiratory diseases).
Table 2-6: Causes of Death in New Zealand for 1997-2002 (Miller, 2005)
18
It can be noted that the death can be attributed to more than one cause – 79 (60.8%)
died from a single cause, 47 (36.2%) from two causes and 4 (3.1%) from three
causes. Miller (2005) also indicated that 28.5% were found dead while attempting to
evacuate or escape and later died of their injuries. Aged and disabled persons have a
‘high risk’ level than others. The causes of deaths due to pre-existing health
conditions will be more in the ‘high risk’ group. The data indicated that death from
pre-existing health condition was 5.95% (see Figure 2-3).
Figure 2-3 – Factors Causing Death during Residential Fires (Miller, 2005)
In Australia, the number of people living in high-rise apartments rose from
approximately 129,000 in 1981 to around 334,000 in 2001, representing an increase
of roughly double the number of people living in private dwelling units (ABS, 2004).
The socio-demographic profile of people living in high-rise units shows that there is
a decrease in aged population (65 years or above) from 17% to 13% (see Table 2-7).
Table 2-7: People living in High-Rise Apartment Buildings, Australia (ABS, 2004)
The number of fires with the increase of people living in high-rise apartment
buildings illustrates the necessity of reviewing the evacuation strategies and
procedures for prompt exit.
19
2.4 Human Behaviour during Fire
Proulx (2003) mentioned after several studies of evacuation of tall buildings that
normal patterns of behaviour and movement route choices tend to persist during
emergency situations. She further stated that occupants often ignore the fire cues or
spend time investigating, seeking information about the nature and seriousness of the
situation, which delay the evacuation time. With the ambiguous information and
short time for decision making, people are likely to choose their most familiar way
out of the building. Visitors often use familiar entry and exit routes in emergency
situations and they may not be familiar with the protected evacuation routes.
Sime’s (1983) study provided strong evidence of human behaviour as affiliative
behaviour during building evacuation. The affiliative model exhibits occupants’
strong tendency toward familiar people and familiar places (such as usual entrance
route). Individuals respond quickly to ambiguous cues, whereas intact individuals in
groups did not begin to evacuate until there was a clear sign of the fire threat.
In a 20-storey apartment building (Japan), a major fire occurred on October 28, 1996,
which engulfed from 9th floor to 20
th floor within 30 minutes. Subsequently, a survey
was conducted in the building, which showed that 47% residents used lifts, 42%
residents used stairs and 7% residents used both. Not even a single resident used
stair on 18th to 20th floors (Sekizawa et al., 1996). In the Forest Lane fire, 40% of the
respondents said that they used lifts (Proulx, 1995). This included people who were
assisted by rescue personnel and occupants who were not successful in escape
through stairs. Occupants may prefer to use lifts, instead of stairs, for evacuation due
to the physical exertion of walking down several flights of stairs (Klote et al., 1993a).
The principal variables influencing an occupant’s decision to move through smoke
tends to be recollecting the exit location and ability to estimate the travel distance
required; secondary variables are the perception of the severity of the smoke and heat
(Bryan, 1983). Occupants try to evacuate the building through evacuation facilities
if the smoke and hot gases are within the tenability limits. Due to poor visibility,
occupants’ walking speed also reduces to a greater extent. Jin and Yamada (1989)
also reported that occupant mental capability reduces with the increase of smoke
20
density and increased radiant heat exposure. In such circumstances occupants
sometimes take wrong decisions about their safe evacuation route.
Proulx (1993) developed a stress model to demonstrate the stress levels induced in
people while making a decision in fire conditions (see Figure 2-4). The stress model
starts with the perception of ambiguous information, which is interpreted in the
processing system resulting in denial of information and non-reactive response.
Occupants tend to ignore the information with the frequency of false fire alarms. A
state of uncertainty prevails with the repeated ambiguous information that induces a
feeling of stress. Overloading of information leads to fear during the emergency
situation, inflicting an increased level of stress and concern of safety. Irrelevant
information induces high levels of stress causing worry about self concern for own
performance in overcoming the emergency situation. Irrelevant information further
manifests a state of confusion since an individual puts in more effort to resolve the
problem and that may result in fatigue and inefficiency.
Figure 2-4 – Stress Model (Proulx, 1993)
21
Schultz (1968) has defined panic type of behaviour a fear-induced behaviour which
is non-rational, non-adaptive, and non-social, which serves to reduce the escape
possibilities of the group as a whole. The perception of fear can further induce in to
a panic behaviour. An example of the way in which the concept of panic is attributed
to people, on the basis of the outcome of a large scale fire tragedy occurred on 28
May, 1977, is provided by a comparison of the extensive report on the Beverly Hills
Supper Club fire, USA. Report concluded that ‘Panic is not considered a major
contributing factor to the large loss of life, but such behaviour probably did occur
when people knew they could not escape’ (Canter, 1990). While the evidence for
panic occurring after people knew they could not escape is inconclusive, the fact is
clear that there was no panic while there was reasonable access to the exits (see
Figure 2-5).
Figure 2-5 – Fear induced Panic Behaviour
There has been an argument by another school of researchers that panic is an event of
rarity during emergency evacuations (Bryan, 2002). Researches have shown that
human behavioural response is consistent during fires and people take rational
decisions. Ramachandran (1991) also found people generally act rationally and
appropriately and they do not panic, which is due to the fact that information is
available to people regarding the existence, size and location of the fire. If the
evacuation routes are not available during the immediate threat of fire, the issue of
irrational behavioural may arise. However, literature survey (Canter, 1990) remains
inconclusive about the perception of issues of human behavioural response during
the non-availability of evacuation route (such as stairs or lifts). Human behavioural
response may vary during lift waiting time and it is likely that a stage of ‘decision
Availability of information
Non-availability of information
such as evacuation
routes
Fear induced in limited time
(or danger)
Non-panic
behaviour
Panic behaviour
...inconclusive?
22
uncertainty’ may prevail amongst them. When building occupants are subjected to
perceived life threatening conditions and there is a hope of survival, there may be an
urge of doing something, which may lead to panic. Therefore, it is possible that the
phenomena of ‘decision uncertainty’ and ‘panic’ could occur, no matter how rare
they are.
Whilst occupants waiting for a lift, concern for their life safety may also arrive,
although, primary means of evacuation such as stairs are available in the building for
evacuation. Moreover if occupants are waiting for lifts and lift arrives at a later stage
with a full occupant load from upper levels, the waiting population may adopt
competitive behaviour. This may cause the lift car to stop and remain at the floor
(Klote et al., 1993b). Competitive behaviour may depend on several factors that may
include lift waiting period, number of evacuees in queue, building features, number
of lifts, fire protection system and level of the fire-affected floor. Therefore the
aspects of human behavioural response such as ‘decision uncertainty’ and ‘panic’
require an analysis in the research.
2.5 Fire and Smoke Hazards
Vertical fire spread in a building could occur through vertical shafts such as garbage
chute, electrical, communication or/ and plumbing shafts as these shafts are normally
located in public corridor and near to residential unit (risk area). Vertical shafts like
stairs and lifts in high-rise buildings are usually required to be isolated with fire rated
doors (ABCB, 2005) and therefore less prone to fire spread but can be a major path
of smoke spread in buildings. US statistics also demonstrate that smoke spread
through vertical shafts accounts for about 95% of the upward movement of smoke in
typical high-rise buildings (Tamura, 1994). Sixty five percent of the vertical
migration of smoke in buildings occurs through the lift doors and shafts whereas
other building system combined together contribute to the remaining 35% (Tamura
and Shaw, 1976). Seventy five percent of the reported incidents show that the smoke
migrated in the buildings where there is no lift lobby (unprotected lift). Smoke also
migrated in 25% of the reported incidents where there was a lift lobby (protected
lift). These statistics were complied by Smoke Guard Corp (USA) using raw data
23
from the US Fire Administration. Other vertical shafts can also act as a passageway
for smoke transfer.
The NFPA data for 1993-97 reveals the location of the victims, which indicates 74%
were intimate with the fire, 20% were on the same floor and 6% were in other
locations (Proulx, 2000). Brennan’s (1999) analysis of US National Fire Incident
Reporting System (NFIRS) fire statistics over a period of 10 years (1983-1993
except 1986) found the number of victims outside the room of origin was
approximately 308 with a maximum of 478 (48 per year). The total number of
victims in apartments was 3,126. Therefore it is estimated that 15% of the victims are
due to occupants attempting to escape. An example of smoke spread in the building
is the MGM Grand (85 fatalities), where all the 61 victims succumbed to smoke
inhalation and asphyxiation. Of the 61 victims, 25 were found in rooms, 22 in
corridors, 9 in stairways and 5 in lifts (NFPA, 1982). Incidents were also reported
where 60% of the hotel guests had moved in to the smoke filled environment by a
distance of more than 21 m. Half of those guests moving through smoke estimated
that visibility was only 1.2 m or less. Two thirds reported turning back when
visibility was 1.5 m (Pauls, 1995). Pauls (1995) also reported that the British
population turned back, when the visibility distance decreased (see Table 2-8). The
visibility distance has been rounded up after converting from British units to SI units.
Table 2-8: Effects of Visibility distance on Turning Back during Smoke in Corridors
(Pauls, 1995)
Smoke can travel at 0.5 m/s - 2.5 m/s under fire conditions. While escaping the fire
affected unit, residents often leave their SOU door in open position (Willey, 1973).
This allows a significant amount of smoke to enter the public escape path. Residents
24
have lost their lives as a result of doors being in the open position. In the Baptist
Towers Home for the Senior Citizens fire (Willey, 1973), the door of a fire-affected
unit was left open, which resulted in 10 causalities in the building. Similarly in the
Rockefeller Park Towers Fire (Bell, 1983), the door was left open resulting in five
causalities. The estimated reliability of passive protection is 90% for construction
with openings (with self-closers like unit doors) under pre-flashover and flashover
conditions (FCRC, 1996). In reality, the door closure mechanism may also fail due to
which the fire door may be in open position. Therefore, the estimated reliability of
door would be much lower value.
Lift shafts are considered to be one of the major paths of smoke migration to other
floors. Although lift cars are fire rated, but the gaps provided to allow trouble free
operation of the doors may result in large quantities of smoke leakage. ASME A17.1
(2000) and AS 1735.1 (2003) permit a maximum gap of 6.5 mm between lift landing
door and frame, and the leakage area calculated is in the range of 0.045 to 0.065 m2
per door. The leakage area of the lift doors is the primary factor in causing smoke to
migrate to upper floors in a building. Construction openings also contribute to the
spread of smoke to upper floors. Smoke movement is further influenced by wind
speed, stack effect and piston effect. All these factors are related to the presence of
leakage area, thus smoke can move considerable distance in buildings.
The action of wind is an important feature in the movement of smoke through lift
shafts. The wind speed is a function of height above the ground at a time, being
nearly zero at ground level and gradually increasing with height. Hence, a high-rise
building will have the major volume of air to follow its path, which causes positive
and negative wind pressures on either side of the building. Window glass breakage
may aggravate the scenario. Roytman (1969) noted that a room gas temperature of
around 300°C is needed for glass breakage to occur.
The stack effect occurs whenever there is any temperature differential between
exterior and interior of a building. Usually temperature difference exists between
interior and exterior atmosphere during fires and stack effect plays a vital role in the
25
smoke movement in lift shafts. The stack effect is significant in high-rise buildings
although it occurs in small buildings too.
Lift movement causes piston effect, which could increase the smoke spread instantly
in the lift shaft. Due to piston effect, smoke may be pulled into and pushed out of the
shaft. Experiments were conducted in a 15 storey hotel in Mississauga, Ontario to
investigate the piston effect and evaluate the model (Klote and Tamura, 1986). The
maximum pressure differential between floor level and lift shaft was measured to be
16 Pa at floor level of the top floor, which gradually decreased as the lift car
approached the ground floor. This value indicated a flow from the building interior
through the lift lobby and into the lift shaft. Analysis of the experimental data
together with modelling results yielded the conclusion that lift piston effect was of
significance only for single car shaft (see Figure 2-6) and could be ignored in the
case of a multiple car shaft due to open peripheral space.
Figure 2-6 – Piston Effect: Lift acts as a Piston for Smoke Movement
Lift cars may be exposed to fire. Lift doors may be in the risk area and may be
damaged in the event of a fire. The doors can warp and do not open and close freely
resulting in loss of the lift for evacuation from fire-affected floors. If the lift lobbies
are not enclosed, smoke and hot gases can flow into the lobby and travel in the lift
shafts (see Figure 2-7).
26
Figure 2-7 – Smoke Infiltration and Heat Exposure to Lift Landing Door
In one of the tests conducted by Bennetts et al. (1999), a fire was ignited in front of
the lift shaft to observe the temperature inside the lift shaft in a two storey building.
Peak heat release rate was approximately 9 MW in 18 minutes after the ignition. The
maximum temperature of the door reached about 1,000°C. The temperature inside
the lift shaft reported to be not more than 160°C. The test determined the maximum
temperature inside the lift shaft for providing the fire rating to the steel components.
Other parameters such as smoke leakage and spread of toxic gases were not
measured. The condition of lift door after the fire test is shown in Figure 2-8 whereas
the temperature inside the lift shaft for air and rear wall is shown in Figure 2-9.
Figure 2-8 – Lift Door after the Fire Test (Bennetts et al., 1999)
27
Figure 2-9 – Temperature inside the Lift Shaft (Bennetts et al., 1999)
Steel distortion at elevated temperature could increase the door gap and may increase
the flow of hot smoke. Tamura and Shaw (1976) measured the air leakage rate
through lift and stair doors and found that at a pressure differential of 75 Pa the air
leakage through lift door was determined to vary approximately linearly with the
width of the crack between the door and doorframe. For a crack width of 2.0 mm, the
air leak rate per door was measured as 0.1 m3/s. For a crack width of 7.0 mm, it was
measured as 0.45 m3/s.
The Building Code of Australia (ABCB, 2005) prescriptive requirements specify
pressurisation for the fire isolated exits in high-rise apartment buildings. The most
common forms of smoke control in apartment buildings are the pressurisation of stair
shafts (this provision is not for lift shafts). A study of smoke control reliability by
Zhao (1998) through a fault tree analysis found that zoned smoke control system has
a reliability between 52% and 62% for buildings between 5 and 20 storeys and stair
pressurisation system has a reliability of about 90%. Smoke may also be controlled
with public corridor pressurisation. In a fire that occurred in Carlyle Apartment
(Taylor, 1975), the door from fire unit to the corridor was burnt. Pressurisation
system was so effective, evacuating residents were able to walk past the apartment
and observed the burning inside the unit.
During the normal operation, the temperature generated inside the lift machine room
is governed by the lift code AS 1735.1 (2003). This code requires that the
temperature in the lift machine room (LMR) should not exceed 43°C (although
28
commercial chips are rated to 70°C). During a fire occurrence, the lift shaft may
carry hot smoke to LMR though floor openings and causes rise in temperature of
electronic components. The higher temperature may cause lift software failure,
which may inadvertently bring lift car to the fire-affected floor. Hence, the provision
of venting in lift shafts is generally incorporated in the codes. Many codes state that
holes in the machine room floor are only permitted for the passage of ropes, cables or
other moving lift equipment and are limited so as to provide no greater than 51 mm
clearance on all sides. Some codes permit venting by means of floor grates into the
machine room with mechanical ventilation to the outside. ASME A17.1 (2000)
recommends the provision for protecting the LMR; cable slots and other openings
between the LMR and lift shaft are to be sleeved from the machine room floor to a
point less than 30 cm below the lift shaft vent (see Figure 2-10).
Figure 2-10 – LMR Protected with Sleeves (smoke dissipates through vent)
It is noteworthy that ASME A17.1 (2000) provides adequate resistance to smoke
spread to LMR, but it does not provide adequate smoke protection for main lift
shafts. Without protecting the lift shafts for safe use of lifts, the necessity of LMR
protection is of little use. If lifts remain functional during hazardous conditions in the
29
lift shafts, in such circumstances if evacuees use lifts, they may be exposed to smoke,
heat and toxic products.
The standards require the provision of protected lift lobbies for emergency lifts only.
The emergency lifts are permitted to be used by fire fighters only, but not by general
public. This aspect needs further research and analysis for providing lifts as an
emergency exit for general public. Influences of wind and stack effect on smoke
spread through lift shafts need to be addressed as random variables in the risk
assessment of lift systems.
2.6 Toxicity of Fire Effluents
Toxicity of fire effluents is expressed in terms of time-additive values. Time-additive
values account for the effect of exposure to a particular gas (or gases) over a period
of time rather than an instantaneous exposure. Toxicity of fire effluents is measured
in terms of the fractional effective dose (FED). The FED for a constant concentration
(C) of a toxicant product is the dose received up to exposure time (t) divided by the
dose required to cause incapacitation or death (Ct). The FED are calculated by the
(1) mass loss models; or (2) toxic gas models.
2.6.1 Mass Loss Models
Mass loss rate is determined either by direct measurement of material in the fire test
or by mathematical modelling. The tests, operated under condition relevant to those
in the fire, supply the lethal mass loss exposure dose expressed in g min/m3. The
FED is calculated for each small time interval. Continuous summation of the FED is
carried out to calculate the total accumulated exposure dose of a toxic species. If
several materials are involved in a fire, the FEDs of each material are summed. A
number of methods for applying this approach have been developed and two of them
are given below:
• Purser mass loss FED model (Purser, 2002) is based on mass loss burning
rate (kg/min) and dispersal volume (kg/m3). The FED for toxicity by mass
30
loss is the summation of the exposure dose of toxic gases, based on an
average smoke toxicity lethal concentration time product of 300 g min/m3.
• Hazard I model (Peacock, et al., 1991) assumes that the smoke toxicity of the
vast majority of combustible materials is virtually the same. The FED for
toxicity by mass loss is the summation of the exposure dose of toxic gases,
based on an average smoke toxicity lethal concentration time product of 900
g min/m3. ISO TS 13571 states that this value is valid for well-ventilated pre-
flashover fires and that half of that value is valid for vitiated post-flashover
fires.
However, these two models do not distinguish between the different effects of
individual fire gases and derives an estimate of toxic potency from the overall lethal
effects of a toxic effluent mixture.
2.6.2 Toxic Gas Models
This method is based on the composition of combustion products and its toxic effects
as a function of time. Generally, a small number of combustion products is
considered for calculating the FED for each toxicant product. A number of methods
for applying this approach has been developed:
• Hartzell toxic gas FED model (Hartzell, 2001) is based on the concept of
upper and lower limits of exposure on occupants to toxic fire gases.
Exposure in excess of an upper limit would be expected to result in serious
harm to a significant number of occupants, while exposure below a lower
limit should ensure that essentially all occupants would be safe from harmful
effects. FED is described as a cumulative effect of exposure to asphyxiant
(or narcotic) gases and is expressed as (Hartzell, 2001):
( )t
Ct
CFED i
n
i
t
t
∆= ∑∑= 11
2
1
2. 1
31
where Ci is the concentration of the asphyxiant gas i in ppm and (Ct)i is the
specific exposure dose in ppm-min required to produce incapacitation.
Equation 2.1 can be written in terms of CO and HCN as:
[ ] [ ]∑∑
∆+
∆=
2
1
2
1)(
t
t HCN
t
t CO Ct
tHCN
Ct
tCOFED 2. 2
Fractional Effective Concentration “FEC” approach is used to calculate the
risk associated with irritant gases (Hartzell, 2001). A total FEC for effects
due to irritant gases, being cumulative, is shown in the following equation
(Hartzell, 2001):
[ ] [ ] [ ] [ ] [ ] [ ]i
i
NOSOHFHBrHCl IC
Irritant
IC
NO
IC
SO
IC
HF
IC
HBr
IC
HClFEC ++++++= ..........
22
22 2. 3
It is noteworthy that the setting of safe exposure criteria at higher FED values
does not provide much additional evacuation time in a rapidly growing fire
scenario. An FED value of 1.0, at which point many occupants are likely to
be overcome by smoke, is reached within only about 2.4 min, following the
0.1 FED safe criterion (Hartzell et al., 1985).
• N-gas model (Levin et al., 1987) addresses the lethal interactions in rats of
four gases (CO, CO2, HCN and low O2). It does not allow for the integration
of changing concentrations with time. It is used largely for 30 minutes
exposures to a constant concentration. It is useful to determine the extent of
which, rats lethality can be explained in terms of the four common gases.
• Human incapacitation model (Purser, 2002) is applied to actual physiological
uptake function and to the effects of major toxic fire gases. It is designed to
predict toxic hazard in terms of exposure dose and time to incapacitation for
humans in fires. This model is intended for use in the current research.
32
2.6.3 Human Incapacitation Model
Purser (2002) addresses tenability limits for smoke toxicity by asphyxiant and heat or
other thermal effects. In his approach, asphyxiant was separated from irritants.
Asphyxiant is addressed by summation of the exposure dose of the individual toxic
gases, based on their individual concentration at each time period. The safe escape
criterion based on the asphyxiant toxicants CO and HCN would most appropriately
be one-tenth of the dose known from experiments to incapacitate non-human
primates. The hypoxia low concentration of oxygen (< 10%) and high level of CO2
(>5%) exacerbate the effect of asphyxiant. CO2 increases the rate of uptake CO and
HCN (hyper-ventilation). People suffering from cardiac failures and respiratory
diseases exhibit greater sensitivity to the effect of asphyxiant (Purser, 2002).
Asphyxiant hydrogen cyanide (HCN) is important if the burning material contains
nitrogen. However, the ultimate effects of the both the asphyxiant are similar, the
pattern of toxicity during the early stages is different (Purser, 2002). The onset of CO
intoxication is slow and insidious, HCN intoxication is rapid and dramatic (Purser,
2002). However, HCN is not routinely measured as a part of post-mortem process.
The effects of irritant gases are determined from the mass loss of material divided by
the volume of air into which the material is dispersed. Irritant fire products cause
painful effects to the eyes and upper respiratory tract and sometimes to lungs.
Irritants are generally not regarded as presenting an initial threat to escape. The
effects of low concentrations of irritants can best be considered as adding to the
obscuration effects (visibility) of smoke by producing mild eye and upper respiratory
tract irritation (Purser, 2002).
Purser (2002) uses the tenability criteria for radiant heat and for convective heat.
Tenable conditions within the building are assumed to be maintained, provided the
hot smoke layer remains at 2.1 m above the floor. Radiant heat flux greater than 2.5
kW/m2 signals the onset of hazardous conditions for occupants. An average hot layer
temperature of 200°C for shallow layers is used to determine criteria for signal that
the heat flux value of 2.5 kW/m2 has exceeded. A hot layer height less than 1.8 m
above the finished floor level of temperature 60°C is the onset of unsafe condition
33
(FCRC, 1996 and Purser, 2002). Radiant heat and convective heat due to temperature
are dose related and FEDs are determined.
2.6.4 Visibility
Lack of visibility does not have a physiological effect. However, even a small
quantity of smoke reduces visibility, and the reduced visibility can lead occupants
failing to find the way out. Thus, occupants may be trapped and suffer the effects of
other fire effluents. Exit sign of light emitting type or light-reflecting type plays a
vital role in visibility. The empirical relations have been established to determine the
visibility distance (V) for exit signs (Jin, 2002):
)()10~5(m
CV
s
= for a light-emitting sign 2. 4
)()4~2(m
CV
s
= for a light-reflecting sign 2. 5
where Cs is extinction coefficient (1/m). The visibility of objects such as walls or
long corridor varies depending on the interior of its contrast condition; however,
minimum value for light reflecting signs may be applicable. The extinction
coefficient should not be higher than 0.5 m-1 for a visibility distance of 4 m.
It is assumed that smoke visibility and heat have no effect on FED for asphyxiant.
Although some effects are likely, no quantitative information is available. Beyler
(2004) concluded that visibility criterion is reached before carbon monoxide hazards
arise in most egress situations and soot & CO yields are well correlated.
2.7 Lift Operational Systems
2.7.1 Overview
Two types of lifts are normally used viz. hydraulic and electric. The hydraulic lift
pushes the car up on a shaft filled with compressed oil whereas the electric motor is
used for lifting car up or down in electric type lift. Hydraulic lifts are slower than
34
electric lifts, having a speed less than of 0.5 m/s, whereas electric lifts operate at
faster than 1 m/s (or in excess of 5 m/s) (Straskosch, 1998). Hence, hydraulic lifts are
suitable for low-rise buildings while traction type electric lifts are suitable for mid or
high-rise buildings.
The traction type lifts suspend the car by cables from the pulley system, using
counterweights to minimize energy expenditure. The counter weights are normally
equal to 50% of full load car capacity. The traction pulley may be attached to
driving motor directly or through gear according to which the lift is called gearless or
geared type. Gearless design is employed in high speed systems (2 m/s and above)
whereas gear type lift is used in slow speed systems (0.125 m/s to 2.3 m/s)
(Straskosch, 1998).
To control the weight capacity, load sensors located under the floor can weigh each
car after stops and when the car has reached its maximum capacity, the controller
will often ignore incoming calls until enough passengers leave. The capacity limits
are intended more for passenger comfort and the adjustment of counterweights. Once
the controller senses an unplanned release from the main cables, a set of emergency
brakes is immediately deployed. The car will not move from that location by itself.
Even if the brakes did fail mechanically, the counterweights would assist to slow the
car's descent (ASME A 17.1, 2000).
Building codes control the maximum number of lifts permitted in a single shaft in
order to limit the potential of a fire disabling all lift services in a structure. Most
building codes permit three or fewer lifts in the same shaft enclosure. When there
are four lifts, they must be in at least two separate shaft enclosures. When there are
more than four lifts, not more than four can be in the same shaft enclosure (AS
1735.1, 2003).
2.7.2 Lift Dispatch Control
The problem of down peak traffic has been addressed by efficient lift dispatching
strategies, which decide where to go and which should be served first. In practice, lift
dispatchers are designed heuristically and evaluated on simulated buildings.
35
Passenger arrivals are modelled as discrete, stochastic events, with arrival rates
varying frequently over the course of a simulated day. It was possible to simulate the
evacuation by lifts using control with intelligence for the evacuation. However, it
will be a major challenge to develop a dispatch control policy for evacuating the
occupants systematically and intelligently above the fire-affected floor. The lift
group controller implements dispatch rules that decide where the cars should go and
stop. The currently used modern approach of genetic algorithms based lift group
control system is utilized to multi-objective optimization in a dynamically changing
process control environment (Tyni and Ylinen, 2006). The chromosome is built up
by taking the landing calls one by one and inserting them into the chromosome as
‘‘call genes’’. This coding approach offers a natural way to meet the requirement that
each call should be responded to only once by one of the lifts. The genetic algorithm
allows a better performance attending to the system waiting times than the traditional
duplex algorithms. The system can be effective during the heavy traffic conditions,
which can invariably be seen during fire emergencies. The latest technological
advancements such as neural networks, fuzzy logic and genetic algorithms provide
better performance (Cortes et at., 2003) and can be useful for lift emergency
evacuation.
2.7.3 Concerns of Lift Operational Mechanism
If lifts are considered for building evacuation, they must be reliable for safe
operation. But the most common problem is the temporary electric power failure.
Statistics reveals that power failure rates in urban Australian locations are
approximately three outages of 10 minutes duration per annum, i.e., a failure rate of
5.7 × 10-5 per annum (Lacey, 2000). (The failure rate is the duration of power failure
per annum). However, power failure may occur during fires since fire fighters cut
the power supply for fire fighting operations in order to avoid possible electrocution.
Power failure may occur due to fire in the electrical system or water damage in the
lift system. The water damage occurs due to fire fighting operation and water based
fire fighting installation. The risks relating to lift operational mechanism need to be
evaluated during fire emergencies.
36
2.7.4 Lift Protection Measures
The lifts are provided with protection measures to avoid electric supply failure.
Emergency power supply of the same voltage characteristics of normal power via the
normal feeder to run the lift system is required (ASME A 17.1, 2000). Transfer
switch is provided with an adjustable time delay of approximately 20-60 seconds for
pre-transfer signal in either direction. The electric supply for the lifts is provided on
separate circuit from the main switch rooms and is taken through armoured cable
separately through respective lift shafts in fire protected route. Automatic Rescue
Device (ARD) meant for the purpose of bringing the car to the nearest landing door
up or down (depending upon the load condition), are also available. ARD normally
activates rescue operation within 10 seconds of normal power supply failure and
operates with the help of battery back up. For a large number of lifts, there is a
winch in the lift machine room at the top of the building that the fire service use to
lower the stopped car to the next floor down. Even with the gearing involved, this is
a very strenuous manual activity and the fire fighters have to take turns at it (HMSO,
1993).
Protected lift lobbies are required according to ASME A17.1 Safety Code to avoid
water damage. Lifts are designed so that water entering hall fixtures will not shut
down the lift when on Fire Phase II operation, which is generally resolved by
providing water resistant components and preventing water entering the lift shaft.
ASME A17.1 also requires the shutdown of power to the lift upon or prior to the
application of water in lift machine rooms or hoist ways (though fire in lift machine
room or lift shaft shut may not require complete evacuation of building). This
shutdown can be accomplished by a detection system with sufficient sensitivity that
operates prior to the activation of the sprinklers (NFPA 72, 2002).
2.7.5 Sprinklers in Residential Buildings
Based on numerous studies, the reliability of fire sprinkler systems has been
documented. The fire sprinkler system reliability ranges from 81.3% to 100%
representing a significant range of performance. The following table summarizes the
37
performance of automatic sprinklers in residential occupancy group during the period
1886 to 1986 (Marryatt, 1988).
Table 2-9: Performance of Automatic Sprinkler System in Residential Buildings
(Marryatt, 1988)
Thirty three fires were reported in apartment type buildings that are sprinkler
protected. Marryatt reports that 100% of fires that occurred in single-family unit
dwellings were controlled by sprinklers. Marryatt reports an average of 1.22
sprinklers in operation for this type of occupancy with no fatality reported. Whilst
the scope of the research does not solely relate to ‘no fatality’ scenario, it is
considered that there would be nonfatal and fatal injuries in apartment buildings due
to the presence of smoke and asphyxiant gases. It must be noted that the higher
reliability of fire sprinklers reported by Marryatt of 100% reflect fire sprinkler
systems where inspections, testing and maintenance were exceeded normal
expectations and applicable generally to installations in Australia and New Zealand.
The study by Bukowski et al. (1999) found the reliability of sprinklers to be 96.6%
for residential occupancy and 94.6% for overall occupancy. The NFPA statistics
(Rohr, 2001) for ten years reporting period from 1989 to 1998 indicates the
operational reliability of automatic sprinkler systems for apartment buildings is
87.6%. Bukowski et al. (1999) and NFPA findings demonstrate the sprinkler
performance on “real” fires that have occurred inside apartment type buildings due to
which there may be nonfatal and fatal injuries in apartment building.
38
2.7.6 Evaluation of Lift Evacuation Time
The performance of lift systems depends on average waiting time and handling
capacity. Handling capacity is the number of people served in a given period during
a round trip time (RTT) for a lift. Handling capacity is calculated for traffic up-peak
in five minutes and represented as HC. RTT is the average time taken for a single
lift to complete its cycle, i.e. receiving passengers on the ground floor, releasing
passengers on the upper floors and returning to the ground floor. RTT is used as the
basis for estimating average waiting time. The evacuation time consists of the sum
of all the RTT divided by the number of lifts plus the time needed to start up the lift
evacuation and the travel time from the lift lobby to the outside (or to another safe
location). Accounting for inefficiencies of lift operation, this evacuation time can be
expressed as:
t J
) + ( + t + t = t j r,
m
=j
oae ∑1
1 η 2.6
where
tr,j is the time for round trip j
m is the number of round trips
J is the number of elevators
η is the trip inefficiency (typically 0.1 for rounding off of probable stops, door
operating time, door starting and stopping time and the unpredictability of people)
ta is elevator evacuation start up time
to is the travel time from the elevator lobby to the outside or to another safe location.
The round trip time tr is can be written as:
t + t = t sTr 2 2.7
where
ts is the standing time
tT is the travel time for one way of the round trip
The standing time is the sum of the time to open and close the lift doors twice, the
time for people to enter the lift and the time for people to leave the lift. The trip
inefficiency accounts for trips to empty floors and trips to pick only a few stragglers.
39
ELVAC (Klote et al., 1991) model is used to determine the lift evacuation time (see
Section 2.10). Traffic condition varies with time in apartment buildings too. Down-
peak traffic substantially increases in the morning when people are going to work
and up-peak traffic can be noticed during evening hours. Down peak traffic would be
similar to a fire condition, if lifts are used for evacuation. The down-peak traffic is
1.5-1.8 times more efficient since the occupants will be using the same lift.
Generally acceptable waiting time for the apartment buildings is 50 to 80 seconds
with 5% to 8% of population handled during a 5 minute period.
2.8 Stair Evacuation System
2.8.1 Overview
Evacuation time depends on the availability of number of stairs in the buildings. The
number and width of stairs depend on the floor population and travel distance
requirement and mainly governed by the building codes. Normally two fire-isolated
stairs are the minimum requirement for a high-rise building. The BCA 2005 DTS
provisions state that the fire-isolated stairs are required to be protected with an open
access balcony, a smoke lobby or to be pressurised. The stair evacuation facility can
be protected with dual system i.e. double compartmentation (with smoke lobby) or
compartmentation and pressurisation.
2.8.2 Evaluation of Stair Travelling Time
The travel time for building evacuation is calculated for the evacuees in the building
since it depends on alarm, sign of danger and time of day. The computer models for
stair evacuation give optimistic results since they assume that stairs do not become
overcrowded. When the occupant density is more than 3.5 persons per square meter,
flow is very congested and slow (Buchanan, 2001). Melinek and Booth (1975)
estimated the evacuation scenarios in two categories. In one case, stairs are
congested and there is virtually no flow. In another case, occupants walk freely. The
stair travelling time is determined using the following equations (maximum value).
40
s
s
tWF
nNt +=1 2.8
s
s
n ntWF
Nt += 2.9
where
t1 is the egress time (congestion)
tn is the egress time (free walk)
n is the number of floors
N is the number of people per floor and exit
Fs is the nominal occupant flow on stairs (persons/meter/second)
W is the effective width of the stairs
ts is the walking time between adjacent floors
Melinek and Booth (1975) recommenfed a typical value of 16 seconds for ts and 1.1
persons/m/s for Fs. However, ts is greatly increased in the presence of elderly
persons. Their walking speed in the stairs is generally 0.5-0.8 m/s. Therefore ts is 30
seconds for a combined population of aged and young occupants. The value of Fs is
0.5 persons/m/s. The expressions indicated above are used for verification purposes.
The difference between the use of stairs and lifts for occupant decent ranged between
15 to 30 minutes. This is a large difference in time lost to travel by stairs, especially
when a fire can grow significantly in a matter of minutes. In 15 minutes, the
environment can be less toxic for the occupants, smaller fires and the property less
damaged. Nelson and MacLennan (1995) indicate that, for stair evacuation, actual
evacuation time can be two or even three times as long. The stairs is the fastest way
of evacuation for low-rise and mid-rise buildings. But the evacuation time by lifts
becomes more favourable for building with more than 100 persons per floor and with
more than 50 floors (Siikonen and Hakonen, 2003).
41
2.9 Risk Assessment Methods
Larsson (2000) classified fire risk analysis into three categories i.e. regulations and
checklists, ranking methods and probabilistic (or quantitative) methods. Regulations
and checklists involve limited risk assessment techniques. A review of ranking and
probabilistic methods is given below:
1. Ranking methods have been developed with the purpose of simplifying the risk
assessment process. Ranking method involves identification of every single
factor that affects the level of safety and data gathering. The importance of
each factor has to be decided by assigning a value (or called weight). This
value is based on the knowledge and the experience of experts. Assigned
values are operated by combination of arithmetic functions to achieve a single
value. The value is a measure of the level of risk and it is possible to compare
this to other similar designs and to a stipulated minimum value. The following
ranking methods are commonly used:
• The Multi-Objectives Decision Analysis (MODA) provides a systematic
approach that addresses several factors that complicate the selection
decision in any simulation design situation. The MODA is also called
Multi-Attributes Decision Analysis. There is not much work done on the
MODA approach in fire safety risk engineering. However, this approach is
used in economics, nuclear energy and resources, policy analysis, scientific
research management, industrial management, manpower planning and
medical diagnosis and defence. The applications in a variety of areas
demonstrate that decision analysis continues to be a widely used approach
for a variety of strategic and tactical decisions (Keefer et al., 2002). The
MODA approach is used in this research project (see Chapter 7).
• The Analytical Hierarchy Process (AHP) operates by using pair wise
comparison judgment to consider factors which are not effectively
quantified. This process is similar to MODA. Factors subject to
uncertainty, ill-defined parameters, conflicting objectives and inexactness
in measurement may be considered with the judgmental process (Saaty,
42
1980). However, the judgment can be highly variable and difficult to work
with. One may study the consistency of judgment and its validity under
certain contexts. This approach is used in a variety of areas, as referred in
MODA. The AHP is used in combination with MODA in this research
project (see Chapters 3 and 7).
• The AHP evolved to an application of Fire Risk Index Method (Watts,
2002). The University of Edinburgh developed this approach initially for a
study to improve the evaluation of fire safety in the U.K. hospitals through
a systematic method of appraisal (Watts, 2002). Risk Index can be used to
describe the impact on the building structure. The Delphi Panel awarded
weights to various related attributes (the Delphi Panel, consists of a group
of experts, who never meet physically and all communication is through a
group controller). The approach yields an effectiveness of a specific design
solution for a given objective and provides a rational basis for possible
design improvements. Earlier reports have described the development of
the Fire Risk Index method and demonstrated that the Fire Risk Index
method can be a very useful tool (Larsson, 2000, Karlsson, 2000, Hultquist
and Karlsson, 2000). However, the Fire Risk Index analysis may not be
appropriate, where greater sophistication is required (Watts, 2002).
2. Probabilistic methods provide quantitative values, typically produced by
methods that can be traced back through explicit assumptions, data and
mathematical relationships to the underlying risk distribution.
• An event tree is a graphical logic model that identifies and quantifies
possible outcomes following an initiating event. The tree structure is
organized by temporal sequence. Probabilities can be calculated and
consequences are assigned to the end states along the tree. Each path
through the event tree defines a scenario. Various outcomes for state of
functioning/ non-functioning of event/ system can be shown. Complex
process can be analysed by modifying event tree into a hybrid combination
of parallel and series system (a complex parallel and series system). Basic
event tree approach is used in fire safety engineering under varying
43
conditions of fire detection, sprinkler operation, door opening and closing
and fire brigade intervention for determining the probability of fire
extinguishment. This approach is not suitable for modeling the building
evacuation under the dynamic scenario (temporal fire and smoke spread).
• Fault-tree provides a simple graphical model based on circuit diagram that
can be used to analyse potential errors in a design. Fault tree is constructed
from events and gates. Fault tree begins with basic events, which represent
the underlying failures that lead to an accident, to top event (outcome).
Numerical probabilities of occurrence are entered and propagated through
the tree to evaluate probability of the foreseeable, undesirable event. This
approach is widely used in electrical engineering for risk assessments in
electrical power stations and provides credible results.
• Safety Index Method involves a complex way of evaluating the level of
fire risk. The advantage is the precision of the results. The risk orientated
analysis started with the selection of potential fire hazards, which could
endanger occupants inside the building. The safety index is described in
terms of escape time margin. The disadvantage with the safety index β
method is that it only considers the escape time of the last person reaching
the safe area for describing the probability of failure of evacuation route
(Frantzich, 1997a). The probability is only addressing the fact that persons
are unable to evacuate safely, i.e. the time margin is less than zero
(negative safety index). No information is available of the probability that
exactly one or two or more are unable to escape (Frantzich, 1997a).
However, this method is effective for a comparative design analysis.
• Monte Carlo technique is used for the risk assessment having stochastic or
probabilistic basis. The Monte Carlo analysis can provide the distributions
of the output variables and their sensitivities to the input variables. Typical
outputs are, for example, the times of component failure, fire detection and
flashover (Hostikka and Rahkonen, 2003).
44
Table 2-10 gives an overview of the common risk assessment methods listed above
and their advantages, disadvantages and ability to meet the research requirement.
Table 2-10: Risk Assessment Methods
Method Advantages Disadvantages Meets the Research
Requirement
Multi-Objectives Decision Analysis
Comparable; Used in multi-element system; resolve conflicting issues
Lack of statistical resources
Yes; Used for addressing the risks
Analytical Hierarchy Process
Comparable; Used in multi-element system; resolve conflicting issues
Hierarchical relationship; Lack of statistical resources
Partially used in-conjunction with MODA
Risk Index Method
Comparable; Used in multi-element system
Workforce requirement (Delphi Panel)
No
Event Tree Analysis
Easy to demonstrate state of functioning/ non-functioning
Not suitable for modelling dynamic scenarios
Yes; Modified form of complex parallel and series system is used.
Fault Tree Analysis
Used in multi-element system/ process; Identifies all possible causes of a specified undesired event
Lack of statistical resources; Not suitable for modelling dynamic scenarios
Yes; Used for reliability study
Safety Index Method
Comparable; Provide precise results
Negative safety index as it considers the entire population for safe margin
Yes; Used for addressing the unsafe conditions in buildings
Monte Carlo Simulation
Comparable; Provide precise results
Time consuming; Lack of statistical resources
Yes; Used for addressing the various probability distribution functions
2.10 Application of Computer Models in the Research Study
Computer models are mainly identified as zone model, field model and egress model.
Field and egress models are used in this research study. A brief is given below.
The zone model ‘CFAST’ (Peakcock et al., 2004) divides the compartment into two
zones and solves the conservation equations within the individual zones, whereas a
field model ‘Fire Dynamic Simulator’ FDS (McGrattan et al., 2004) divides the
compartment into a large number of volumes and solves the conservation equations
within the individual volume and therefore is more complex. The zone model
45
‘CFAST’ is used for peak temperature and smoke transport calculations and it
assumes that the compositions of layers are uniform, thus, the temperature and other
properties are the same throughout each layer (Peakcock et al., 2004). This
assumption is less valid for very large spaces or long narrow spaces such as stairs
and lift shafts.
A computer simulation developed by the NIST called CONTAMW (Walton, 1993)
was used to analyse the effect of compartmentalization strategy in buildings. The
CONTAMW (a multi-zone, multiple floor airflow network analyser) is able to model
wind and stack effects and can also predict smoke movement. The CONTAMW
simulations used in combination with zone model (CFAST) showed that smoke
movement was significantly reduced by the compartmentalization strategy (Klote,
2003). Klote (2003) stated that future research is needed to evaluate the extent to
which compartmentation failure would impact smoke flow through lift shafts.
However both softwares had limitations in considering non-deterministic variables
with the flow of fire, smoke and hot gases in long shafts. With the intent of further
extending the research, field model ‘FDS’ is used in the current study.
ELVAC (Klote et al., 1991) presents the analysis of people movement by lifts in the
buildings, which incorporates more details about lift motion and lift loading and
unloading. The model is developed based on mathematical derivations described in
lift engineering (see Section 2.7.5). However, ELVAC model does not incorporate
horizontal components (deterministic and non-deterministic movement) on upper
levels. Software Building Traffic Simulator (BTS), which is an advanced version of
Advanced Lift Traffic Simulation (ALTS) (Siikonen, 1989), is also used in lift
industries for simulating passenger traffic in the buildings (Hakonen, 2003). BTS is
designed with the purpose to analyse the performance of a lift system, to demonstrate
lift systems for customers and to test lift group control software, but this software has
also the same limitations as in ELVAC.
EVACNET (Kisko et al., 1998) can model building evacuations using stairs and lifts.
The program accepts a network description of a building and information on its
initial contents at the beginning of the evacuation. From this information, EVACNET
produces results that describe an optimal evacuation of the building. Each evacuation
46
is optimal in the sense that it minimizes the time to evacuate the building. Model
incorporates several service levels for walkway and stairway. The level of service can
be assumed to be 50-95% of maximum capacity in the stairs, where speeds are
restricted, passing is virtually impossible and reverse flows are severely restricted.
Both the models (ELVAC and EVACNET) used in evacuation modelling are
deterministic models and these models can not predict the realistic situations such as
occupants’ pre-movement and movement activities. The problem is resolved by
using industrial application SIMAN ARENA (Rockwell, 2000) discrete event
simulation model. Discrete event simulation is a type of simulation where occupants’
pre-movement and movement activities can be incorporated by probability
distributions. This model is used in determining the lift waiting time, lift evacuation
time and the number of occupants in queue. Simulation models for lifts and stairs
have been prepared so that the issues relating to human behavioural response (such
as uncertainty and panic) and life threatening conditions can be adjudged and
compared. However, this software does not simulate fire.
For risk assessment, @RISK package (Palisade Corp, 1996) is used based on Monte
Carlo technique, where all the parametric probability distributions are analyzed for
determining the risks. Uncertain input values in spreadsheet are specified as
probability distributions. An input value is a value in a spreadsheet cell or formula
which is used to generate results in spreadsheets. A probability distribution describes
the range of possible values for the input is substituted for its original single fixed
value. Available graphs include probability distributions of possible output variable
values and cumulative probability curves. By dragging the delimiters displayed on a
histogram or cumulative graph, target probabilities can be calculated.
2.11 Discussion and Summary
Stairs are the only evacuation facility acceptable by prescriptive building regulations
for high-rise buildings as lifts are not considered safe during fire emergencies. The
literature review reveals the following points, which warrant the use of lifts as an
emergency evacuation facility in high-rise apartment buildings:
47
• Due to limited physical capability, aged and disabled persons are slow to
walk in stairs. The aged and disabled persons may require lifts for prompt
and safe evacuation from the building during emergency situations.
• People find enormous difficulties in evacuating high-rise buildings. Long
evacuation time may sometimes endanger the life of evacuees or cause
injury due to tiredness, dizziness, slipping on surfaces or becoming less
capable physically.
• Although stair design is compliant with building regulations, the
travelling speeds as a whole may be affected by bottlenecks, turns, and
obstacles in the stairs, all of which can cause crowding at certain points
and hinder timely evacuation of high-rise buildings.
Literature review reveals that people generally act rationally and appropriately
(Proulx, 2003). Normal patterns of behaviour and movement route choices tend to
persist during emergency situations. However, under certain circumstances, evacuees
could become impatient and overcrowd the lifts, which can cause the car to stop
functioning and remain at the floor. Considering that panic may be a rare event and
the outcome of the study of this phenomenon is inclusive, it is assumed that evacuees
may become panic of the dangerous conditions.
Literature review identifies that lifts are exposed to several risks during building fires
that require further works for bringing lifts to a required safety standard (Klote,
1982). Due to a gap between lift door and lift landing frame, smoke can move to lift
shafts and upper levels in buildings. Sixty five percent of the vertical migration of
smoke occurs through the lift doors and shafts, which is further influenced by stack
and wind effects. The lift operational mechanism is also not yet fully reliable. To
provide a safe alternative evacuation system, the risks relating to lift operational
mechanism must be addressed for improving the reliability of lift systems. Literature
review also revealed that a lift evacuation system for a small number of people is
feasible (Klote et al., 1995 and Paul et al., 1991). However, these preliminary
research works have not been quantified.
The issues relating to human response, fire hazards and lift operational mechanism
need to be addressed in an integrated manner to determine the feasibility of lift
48
evacuation systems. The safety of evacuees is to be considered under variable
conditions of fires and human behaviour. If all the fire safety issues relating to lifts
are resolved appropriately, lifts can be considered as an option for emergency
evacuation.
Literature review also reveals that a combined use of stairs and lifts can reduce the
evacuation time considerably during emergencies in mega high-rise buildings (Klote,
et al., 1993a, Siikonen and Hakonen, 2003). Greater evacuation efficiency occurs as
the height of the building increases. However, this aspect does not include the issues
related to the safe use of lifts during fire emergencies. The current research is
performed to address the need to use lifts for evacuation purposes and to answer the
following questions:
• Can protected lifts provide adequate safety for general public during fire
emergencies?
• Is any additional provision required for providing lifts as an emergency
evacuation facility? What can be the acceptable level of risk for lift
systems?
• What is the suitable risk assessment approach for lift evacuation systems?
This research focuses on the above points for exploring the safe use of lifts on a
comparative basis, with stairs, after establishing the inter-relationships among the
potential risks.
49
3. RESEARCH METHODOLOGY
Risk assessment remains a challenging task, especially in view of the uncertainties
related to extreme events exceeding the safety limit criterion of fire safety measures.
Detailed spatial information on risks is extremely important in determining the use of
lifts as an alternative evacuation facility in the buildings. This chapter provides an
overview of research methods, indicating how the subsequent chapters fit together
and demonstrates how an integrated risk assessment method connects to other
evaluative activities. The objectives of this chapter are:
1. To identify the risks associated with the use of lifts for emergency
evacuation and develop an inter-relationship among the risks.
2. To establish a research strategy for an acceptable level of risk and consider
suitable design options and evacuation strategies.
3. To investigate a suitable risk assessment approach for lift evacuation
system.
Risk is defined as the probability of a specific undesirable event occurring in specific
circumstances arising from the realisation of a special hazard (Magnusson, 1996).
Risk is expressed as a function of the probability of an event occurrence and the
consequences of that event occurrence, which can be represented as:
Risk = Probability of Occurrence × Severity of Consequence
Risk increases as a function of the probability of occurrence and the severity of the
consequences. No activity is risk free and therefore a research strategy is adopted to
reduce the probability of occurrence or severity of the consequences or both to
achieve an acceptable level of risk (research methodology for evaluating the risks is
different from this research strategy; research methodology involves selecting
suitable methods whereas research strategy involves appropriate planning).
50
The research methodology involves the following components:
• To identify all the significant risks in the lift evacuation system and develop a
relationship among the risks.
• Rank all the risks (or risk priorities) in terms of likelihood of occurrence and
expected impact upon the building evacuees.
• Establish a research strategy for an acceptable level of risk.
• Identify risk control design options and evacuation strategies for evaluating
risks.
• Quantify consequences with the models and techniques (for example,
stochastic evacuation models, fire hazard models and probabilistic analysis
techniques).
• Conduct risk assessment with a suitable method (Multi-Objectives Decision
Analysis – MODA method is used).
• Select appropriate risk control design option.
The methods and results from this research form the base for comprehensive risk
analysis and risk management strategies. Therefore, an appropriate scale of risks is a
fundamental precondition for a reliable risk assessment. These risks need to be
prioritised and presented in a straightforward, readily understandable format that
shows both the risks and how they are being managed.
3.1.1 Risk Identification
Building regulations worldwide do not permit the use of lifts as a safe mode of
vertical transport system for building occupants during fire emergencies. The
traditional approach of not using lifts as an evacuation facility during fires is mainly
due to the following controversial and unresolved issues (Klote, 1982):
Issues related to human behavioural response
1. Irrational human psychological behaviour
� People may be ‘uncertain’ (or doubtful) of their decision of choosing
lifts for evacuation during lift waiting period; and
51
� People may ‘panic’ in the lift lobby (or in the lift cabin) under certain
circumstances (due to 5, 6, 7 and 8 shown next).
Issues related to fire hazards
2. Smoke and toxic gases spread to lift lobby and vertical lift shaft (vertical
shaft as a major path for smoke and toxic product spread); and
3. Lift passing through fire-affected floor and possible exposure to high
temperature; and
4. Influence of wind speed and stack effect on smoke spread.
� Because of 2, 3 and 4, people may be exposed to ‘life threatening
conditions’ in the lift lobby (or in the lift cabin).
Issues related to lift operational mechanisms
5. Malfunctioning of lift equipment may inadvertently cause the lift to go to the
fire-affected floor; and
6. Effect of water from fire fighting or sprinkler operation on electrical systems;
and
7. Loss of electric power supply during fire; and
8. Lift unavailability due to maintenance.
� Because of 5, 6, 7 and 8, people may ‘panic’ in the lift lobby (or in the
lift cabin).
Figure 3-1 shows the risk cause-effect relationship, developed among the above
issues related to the use of lifts during fire emergencies. The risk issues broadly fall
under three categories i.e. human behavioural response, fire hazards and lift
operational mechanism. These issues can lead to the risks of decision uncertainty (a
psychological impact), panic (mainly psychological impact) and injuries (nonfatal
and fatal – mainly physiological impact).
52
Figure 3-1 – Risks involved in the Lift Evacuation System
53
3.1.2 Analytical Hierarchical Process for Risk Priorities
The Analytical Hierarchical Process (AHP), developed by Saaty (1980), is one of the
more extensively used approaches in multi-objectives decision making methods. The
AHP has been applied to a wide variety of decisions and human judgment processes
(Lee et al., 2001). The AHP involves three basic steps:
1. hierarchical structure; and
2. comparative judgments or defining and executing data to obtain pair-wise
comparison data on elements of the hierarchical structure; and
3. synthesis of priorities or constructing an overall priority rating.
A hierarchy is an abstraction of the structure of the system for studying the
functional interactions of its components and their impact on the entire system. The
abstraction can take several related forms, all of which essentially descend from an
overall objective. Problems under complex conditions are analyzed into a hierarchy
structure. The hierarchical structure is prepared based on previous studies and
empirical experiences. Once a hierarchy has been developed, a pair wise comparison
is needed to determine the relative importance of the elements (or entities) in each
hierarchical level. For a pair wise comparison, the analysis involves
1. developing a comparison matrix at each level of the hierarchy starting from
the second level and working down; and
2. computing the relative weights for each element of the hierarchy; and
3. estimating the consistency ratio to check the consistency of the judgment.
The pair wise comparison at a given level can be reduced to a number of square
matrices A = [aij]nn as:
nnnn
n
n
aaa
aaa
aaa
.........
.
.
........
........
21
22221
11211
54
The matrix has reciprocal property as:
ji
ija
a1
=
In AHP, Saaty (1980) recommended numerical values 1, 3, 5, 7 and 9 for making
subjective pair-wise comparisons (see Table 3-1). Intermediate values between two
adjacent judgments can also be assigned. The increasing numerical values indicate
increasing importance.
Table 3-1: The 9-Point Scale (Saaty, 1980)
The vector weights, w = [w1,w2, . . . ,wn], is computed on the basis of Saaty’s
eigenvector procedure in the following two steps:
a. The pair-wise comparison matrix, A = [aij]nn, is normalized by the
following equation:
∑=
=n
ji
ij
ijij
a
aa
1,
*
3.1
b. The weights are computed by the following equation:
n
a
w
n
ji
ij
i
∑== 1,
*
3.2
55
Then, the weight of all the elements is
11
=∑=
n
i
iw 3.3
If λ is a number and w is a non-zero vector, then w is called eigenvector of A and λ is
the associated eigenvalue. The following equation gives the relation between the
vector weights w and the pair-wise comparison matrix A:
wAw λ= 3.4
This equation can be written as:
( ) 0=− wIA λ 3.5
where I is the identity matrix (with elements on the main diagonal set to 1). Then the
determinant equation is:
0)det( =− IA λ 3.6
The maximum eigenvalue λmax is an important validating parameter in AHP (Saaty,
1980):
n≥maxλ 3.7
The λmax is used as a reference index to screen information by calculating the
consistency ratio (CR) of the estimated vector. When the matrix is perfectly
consistent, λmax equals to n. If the matrix is not perfectly consistent, λmax, is greater
than n. The larger the λmax, the greater is the degree of inconsistency. To calculate
the CR, the consistency index (CI) for each matrix of order n can be obtained from
equation:
56
1
max
−
−=n
nCI
λ 3.8
The CR can be calculated using the following equation:
RI
CICR = 3.9
where RI is the random inconsistency index. The random inconsistency indices for
the matrices of the order of 1 to 10 are given in Table 3-2. If CR < 0.1, the
comparisons are consistent and if CR > 0.1, the comparisons are of inconsistence
judgment.
Table 3-2: Random Inconsistency Indices (Source: Saaty, 1980)
The risks are given priorities from the AHP and a relationship is developed on
numerical scale as given below:
Hierarchical Risk Levels: A triangle-shaped diagram is used to indicate the degree
of hazards associated with risks (see Figure 3-2). The degree of hazard is utilised to
indicate hazard rating. The diagram identifies three colour-coded categories of
hazard for each risk. The risk of decision uncertainty is shown with green, panic
with blue and injuries (fatal and/ or nonfatal) with red colour. The triangle indicates
convergence from low level to high level as the severity of risk consequences
increases to the maximum at the top. However, the area of triangle indicates
divergence from high level to low level as the probability of risk occurrence is the
maximum at the bottom.
57
Figure 3-2 – Risks at Three Hierarchical Levels
Three hierarchical levels are given to the risks for pair-wise risk comparisons. The
risk in hierarchical level represents the dominance of risks at its bottom. The
hierarchic levels are based on the degree of hazards from psychological to
physiological effects on a time based sequence during the evacuation procedure. The
impacts lie on a continuum from little or no effect at low level to relatively severe
incapacitation at high levels, varying in response for different individuals. The risks
at three hierarchical levels and their expected consequences (impacts) are given
below:
• Low Risk (Decision Uncertainty): Uncertainty of making decisions for
using lifts arrives at the time of building evacuation and may be caused by
excessive waiting time for lifts in the lift lobby. Decision uncertainty may be
influenced by the number of evacuees waiting in the queue. Uncertainty is the
state of belief when one is unsure (Reber, 1995) and decision uncertainty
refers to a lack of knowledge about the lift waiting period or pathways.
Decision uncertainty is considered as low level of severity. The expected
impact can be ‘anxiety for information or knowledge’ or ‘mental agony’
amongst the evacuees (psychological effects).
Lev
el o
f Sev
erity
Low
Decision Uncertainty
Pro
bab
ility o
f O
ccurr
ence
Injuries (Nonfatal and/or
fatal)
Panic
Medium
High
Lift Operational Mechanism and/ or Pre-Life Threatening
Condition)
Fire
Hazards
Physiological and/or
Psychological
Cause Level of
Severity
Risk/
Hazard Impact
Psychological and/or
Physiological
Psychological
Lift Waiting Time and/ or Evacuees’ Queue
58
• Medium Risk (Panic): Panic may arrive in the lift lobby and/or lift cabin and
may be due to faulty or unavailability of lift system or visual threat (pre-life
threatening condition). Schultz (1968) defined panic as a fear-induced
behaviour which is non-rational, non-adaptive and non-social, which serves
to reduce the escape possibilities of the group as a whole. Prolux’s (1993)
stress model demonstrated that fear can be induced during emergencies,
which may subsequently convert into panic. Ramachandran (1991) found that
if evacuees are abnormally delayed and they are likely to be exposed to
unsafe conditions, the concern of their life safety is imminent. It could cause
a sense of life threat and inflict panic. Panic is considered as medium level of
severity. Evacuees may adopt competitive behaviour and overload lifts. The
expected impact can be ‘mental agony, physical injuries or deaths
(psychological and/or physiological effects).
• High Risk (Injuries): Life threatening condition may arrive due to hazardous
conditions arising from fire hazards in the lift lobby and/or lift cabin. Injuries
(nonfatal or fatal) are considered as high level of severity. Evacuees may
adopt competitive behaviour and overload lifts. The expected impact of life
threatening condition can be ‘mental agony, physical nonfatal or fatal injuries
due to exposure to smoke, heat and toxic products (psychological and/or
physiological effects).
The levels are conservatively assumed in orders in the lift system. However, panic
may occur first without first causing decision uncertainty among the evacuees.
Likewise, physical injuries may also occur first, without causing decision uncertainty
or panic or no such issues may arise during fire emergencies. The risk levels are
considered independently.
Comparative judgments: The use of lifts during fire emergencies is evaluated in
relation to the psychological and physiological impacts. The psychological and/or
physiological impacts may occur due to the risks from “decision uncertainty, panic
and injuries (nonfatal or fatal)”. However, these risks are related and one risk may
lead to another risk during the lift evacuation procedure. A relationship is developed
among the risks and impacts for a comparative judgement (see Figure 3-3).
59
Figure 3-3 – Hierarchical Relationship for the Evaluation of Risk Priorities
Matrices are developed for assigning the priorities for three levels of risks {decision
uncertainty, panic and injuries (nonfatal and fatal)}. The priorities are the numerical
ranks measured on a ratio scale (Saaty, 1980). The priorities are obtained from the
judgment of a column divided by the sum of all the judgment.
Physiological impact is considered more important than psychological impact for
causing injuries. Miller (2005) indicated that out of 131 victims of residential fires,
15 victims acted in irrational or attention seeking ways (although their intent to cause
fires and to cause harm was ambiguous or unclear). Nearly 11% of the victims were
due to psychological impact. Therefore, physiological impact is assigned the absolute
number 8 in the (2, 1) or second-row first-column (see Table 3-3). This implies that
physiological impact is eight times more risky than psychological impact. The
reciprocal value is shown in (1, 2) as 1/8, which signifies that psychological impact
is eight times lower in risk than physiological impact. The priorities are 0.111 for
psychological impact and 0.889 for physiological impact {calculated for example, 1
divided by (1+ 8) for first column, gives the priority for psychological impact a value
of 0.1111 (≅ 0.111)}.
Physiological Impact
Psychological Impact
Decision
Uncertainty Panic Injuries
(Fatal)
Evaluation
Impact
Risk
Use of Lifts during
Fire Emergencies
Injuries
(Nonfatal)
60
Table 3-3: Matrix (2 × 2) for Priorities of Lift Evacuation
Psychological Impact Physiological Impact Priorities
Psychological Impact 1 1/8 0.111
Physiological Impact 8 1 0.889
λmax = 2; C.I. = 0; C.R.= 0
Panic is considered more risky than decision uncertainty for psychological impact
and is assigned the absolute number 5 in the (2, 1) cell of Table 3-4. Likewise,
nonfatal injury is considered to be of extreme importance than other risks and
therefore assigned intensity of importance to a value of 9. Similarly, 2 is assigned to
nonfatal injury in comparison to panic in (3, 2). The risk priorities are 0.066 for
decision uncertainty, 0.319 for panic and 0.615 for nonfatal injury for psychological
impact. The calculations of risk priorities, λmax , C.I. and C.R. are given in Appendix
B.
Table 3-4: Matrix (3 × 3) for Priorities of Psychological Impact
Decision Uncertainty
Panic Injury (Nonfatal)
Priorities
Decision Uncertainty 1 1/5 1/9 0.066
Panic 5 1 ½ 0.319
Injury (Nonfatal) 9 2 1 0.615
λmax = 3.10; C.I. = 0.01; C.R.= 0.10
Nonfatal injury is considered more risky than panic for physiological impact and is
assigned the absolute number 3 in the (2, 1) cell of Table 3-5. Fatal injury is
considered to be of extreme importance than other risks and therefore is assigned an
intensity of importance value of 9. The risk priorities are 0.077 for panic, 0.231 for
nonfatal injury and 0.692 for fatal injury for physiological impact.
Table 3-5: Matrix (3 × 3) for Priorities of Physiological Impact
Panic Injury (Nonfatal)
Injury (Fatal) Priorities
Panic 1 1/3 1/9 0.077
Injury (Nonfatal) 3 1 1/3 0.231
Injury (Fatal) 9 3 1 0.692
λmax = 3; C.I. = 0; C.R.= 0
61
Synthesis of priorities: The priority vectors from Tables 3-3 to 3-5 are combined
into a single (or global) priority vector for evaluating the risks associated with the
use of lifts in fire emergencies (see Table 3-6).
Table 3-6: Global Risk Priorities
Risk Global Priorities
Decision Uncertainty = {(0.066 × 0.111) + (0 × 0.889)} 0.0073
Panic = {(0.319 × 0.111) + (0.077 × 0.889)} 0.1039
Injury (Nonfatal) = {(0.615 × 0.111) + (0.231 × 0.889)} 0.2736
Injury (Fatal) = {(0 × 0.111) + (0.692 × 0.889)} 0.6152
The risk priorities are 0.0073 for decision uncertainty, 0.1039 for panic, 0.2736 for
nonfatal injury and 0.6152 for fatal injury (see Table 3-6 and Figure 3-4). Fatal
injury has the maximum risk priority for causing risk in the lift system. The
combined risk priority for injury (nonfatal and fatal) is 0.8888 (≅ 0.889) and is used
later for evaluating the risk relating to physiological impact.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Priority
Decision
Uncertainty PanicNonfatal Fatal
Figure 3-4 – Global Risk Priorities
The matrices indicated in Tables 3-3 to 3-5 are consistent, and the consistency ratios
(C.R.) are 0 or less than 0.1 (see Table 3-7).
62
Table 3-7: Consistency Tests of Matrices
Matrix Consistency Ratio Consistency Test
Matrix for Main Objective 0 Accepted
Matrix for Psychological Impact 0.086 Accepted
Matrix for Psychological Impact 0 Accepted
Limitation: For pair wise comparisons in a matrix, the decision maker specifies a
judgment about ‘How much more important is one risk than the other?’ Each pair
wise comparison requires the decision maker to provide an answer to the question:
“how much more important is decision Uncertainty (or Panic) than Panic (or Injury),
relative to the overall objective?” Decision makers often find it difficult to accurately
determine cardinal importance weights for a set of risks simultaneously. As the
number of risks increases, better results are obtained when the problem is converted
to one of making a series of pair wise comparisons.
3.1.3 Acceptable Level of Risk
Risk analysis of lift evacuation system can provide a framework for decision-
making. More challenge is involved with the identification of a level of acceptable
risk, which is more philosophical than technical (Watts and Hall, 2002). Vrijling et
al. (1995) proposed acceptable levels of risk from both individual and societal points
of view. Acceptable risk RA is defined in terms of the policy factor expressing the
degree of voluntariness with which the activity is undertaken and with the benefit
perceived.
410−×= volAR β 3.8
where βvol is the degree of voluntariness with which the activity is undertaken
(varying from 0.01 to 100). The value of 100 is an indicator of voluntary high risk
exposure for its thrill activities like mountaineering and 0.01 for involuntary risk
exposure due to a hazardous structure, as with spatial planning issues (Vrijling et al.,
1995). This scale can be further refined (Vrijling et al., 1995).
63
The acceptable risk can provide a rational basis for single objective attribute, but this
does not reflect the criterion for multi-objective attributes. The present criteria for
risks are not yet fully developed to address the risks in a complex environment. A
universally acceptable level of fire risk does not exist (Watts and Hall, 2002).
Researchers are confronted with the problem of deciding what risks will adhere to
safety that affect life and death of the people taking advantage of its benefits. A value
judgment can be an arbitrarily human perception of risk, which may involve error.
For considering acceptable level of risk in the evacuation route, the risks, involved in
the fire-isolated exit of BCA 2005 Deemed-To-Satisfy (DTS) provision such as
stairs, are considered.
3.1.4 Hypothetical Building and DTS Provisions
Hypothetical Building: A hypothetical building shown in Figure 3-5 with 38 floors
is considered in this research analysis. The building containing 38 floors is selected
with the intent of providing a generic situation for high-rise apartment buildings,
above which, other provisions such as lift stacking or lift zoning are generally
applicable. The typical floor area of the building is 1000 m2 approximately. The
details of enclosures are given in Appendix J. The apartment building contains
twelve dwelling units (three bedrooms, two bedrooms and one bedroom), lifts, stairs
and public corridor. The lifts and stairs run from the ground floor to the top level.
The stairs and lifts are accessible through public corridor.
Figure 3-5 – Typical Floor of a Hypothetical Building (57 m × 20 m)
64
All the access, egress and fire safety features are assumed to be compliant with the
Australian building code (ABCB, 2005).
BCA 2005 DTS Provisions (ABCB, 2005): A fire-isolated stairway (including any
associated fire-isolated passageway) serving any storey above an effective height of
25 m must be provided with:
• an automatic air pressurisation system for fire-isolated exits in accordance
with AS 1668.1 (Table E2.2a); or
• open access balconies in accordance with (Clause D2.5):
- ventilation opening to the outside air; and
- not to be enclosed on its open sides above a height of 1 m except by an
open grille or the like having a fire air space of not less than 75% of its
area.
• If more than 2 access doorways open to a required fire isolated exit in the
same storey (Clause D1.7d),
- a smoke lobby must be provided; or
- exit must be pressurised in accordance with AS 1668.1.
The hypothetical building shown in Figure 3-5 is not provided with open access
balcony (or public corridor). There are more than 2 access doorways opening to the
public corridor in the same storey. Fire-isolated exits such as stairway and public
corridor need to be provided with pressurisation. As the positive pressurisation
would not allow smoke and hot gases to fire-isolated stairway, associated
passageway and lift shaft, positive pressurisation as a mechanical system is not
considered. Only a smoke lobby in the stairway is considered. The practice of
smoke lobby is also adopted by international codes and regulations (NBS, 2000 and
NFPA 101, 2000). The provisions of pressurisation and smoke lobby are shown in
Figure 3-6.
65
Figure 3-6 – Positive Pressurisation and Smoke Lobby in a Fire Isolated Exit
The risk involved in the fire escape stair with smoke lobby is considered as a
minimum acceptable requirement for a comparative analysis.
3.1.5 Selection of a Fire Scenario
The variables related to the building spatial environment, fire dynamics environment
and human activities are important for evaluating the use of lifts for emergency
evacuation. The variables are discussed in the subsequent chapters. However, a
deterministic approach is used for deciding a worst possible fire scenario. After
selection of a worst possible fire scenario, non-deterministic approach is considered
for detailed analysis.
Event tree analysis (ETA) is used for the evaluation of the worst possible fire
scenario. ETA is based on binary logic, in which an event either has or has not
happened. It is valuable in analyzing the consequences arising from an undesired
event. An event tree begins with an initiating fire (see Figure 3-7). The consequences
of the event are followed through a series of possible paths. Each path is assigned a
probability of occurrence and the probability of the various possible outcomes can be
calculated. A smoke detector either detects the smoke or it does not. A sprinkler
either works or it does not. Sole occupancy unit (SOU) door either closes correctly
or it does not. Shaft pressurisation either works or it does not. The probabilities of
success values of these events are given below:
FD
Legend: FD – Fire Door
SD – Smoke Door
Enclosed public corridor with
number of SOU access door Enclosed
Stair
Smoke
lobby
FD SD Enclosed public corridor with
number of SOU access door Enclosed
Stair
Blower to pressurise fire-
isolated exits
Provision of a smoke lobby
66
• Smoke alarms: The estimated reliability of smoke alarms in residential
buildings is 77.8% and the estimated reliability in apartment buildings is
69.3% (Bukowski et al., 1999). Smoke alarms (AS 3786) are therefore
assigned a probability of success value of 0.693.
• Sprinklers: Marryatt (1988) has quoted sprinklers as being 100% reliable
where inspection, maintenance and testing activities were well documented.
The study by Bukowski et al. (1999) found the reliability of sprinklers to be
96.6% for residential occupancy and 94.6% for overall occupancy. The
NFPA statistics (Rohr, 2001) for ten years reporting period from 1989 to
1998 indicates the operational reliability of automatic sprinkler systems for
apartment buildings is 87.6%. Hence, sprinkler systems are assigned a
probability of success value of 0.876 for apartment buildings.
• SOU fire door: The estimated reliability of passive protection is 95% for
construction with no openings and 90% for construction with openings (with
self-closers) under pre-flashover and flashover conditions (FCRC, 1996).
Residents within the fire affected unit are assumed to escape and leave their
SOU in one of two conditions i.e. entry door open or closed. These scenarios
simulate real life conditions where building occupants have lost their lives as
a result of doors being in the open position and occupants have decided to
flee at this time. Hence, SOU entry door is assumed to be either open or
closed. The door is assigned 0.9 in the closed and 0.1 in the open position.
• Shaft pressurization: The shaft pressurisation system is assigned a
probability of success value of 0.9 (Zhao, 1998) (see Chapter 2, Section 2.5).
67
Figure 3-7 – Event Tree Analysis for a Worst Possible Path (or Fire Scenario)
(Second worst possible path is analysed in this research program; remaining paths have less severity and therefore they are not analysed)
68
Probabilities of events are shown on the tree diagram. For each path (branch), a
probability of final outcome is calculated. Assuming the events are independent, the
probabilities along each path are multiplied. High severity of consequences is
involved in the evacuation routes, if sprinkler, fire door and pressurisation are not
working or none of the events are working (for example, failure of detection,
sprinkler, fire door and pressurisation). High severity of consequences may cause
psychological and/or physiological impact (for example, mental agony, anxiety,
panic, nonfatal injuries and/or fatal injuries). The probability of occurrence for high
severity of consequences is 0.0004 and is considered very low for failure of all fire
safety measures. Next higher probability of occurrence for high severity of
consequences is 0.0009 for failure of all fire safety measures except smoke alarm.
Other paths have comparatively lower severity of consequences and have high
probabilities of occurrence. However, it can not be interpreted that the severity of
consequences leading to mental agony, anxiety, panic, nonfatal injuries and/or fatal
injuries would not occur in other paths. It can also not be interpreted that the severity
of consequences leading to mental agony, anxiety, panic, nonfatal injuries and/or
fatal injuries would definitely be caused in the high severity of consequent paths.
Keeping in view the second worst possible path (or branch), where sprinkler, fire
door and pressurisation are not working and smoke alarm is working, the path is
analysed with variables related to the issues of human behavioural response and fire
hazard.
3.1.6 Concept Design Options
Risk analysis for lift system is conducted on a comparative basis with the minimum
requirement (stairs with smoke lobby). Three concept design options along with
additional evacuation strategy (lift evacuation for 25% of the population) are
considered for risk analysis (see Figure 3-8).
69
Figure 3-8 – Comparison of Stairs and Lifts
Table 3-8 gives the purpose of comparison and strategic planning between lift and
stair systems.
Table 3-8: Lift and Stair Systems for Comparison
S. No. Lift system Stair system Purpose
1. Unprotected lift lobby
(100% population evacuation)
Stairs
(100% population)
Feasibility consideration
2. Protected lift lobby
(100% population evacuation)
Stairs
(100% population)
Feasibility consideration
3. Protected lift lobby
(25% population evacuation)
Stairs
(75% population)
Feasibility consideration
and strategic planning
4. Double protected lift lobby
(100% population evacuation)
Stairs
(100% population)
Feasibility consideration
The details of the five concept design options are given below:
Concept Design ‘A’ (Lifts with Unprotected Lobby): Generally lifts are not
protected in high-rise apartment buildings. Therefore, concept design ‘A’ is
considered without any protective measure (see Figure 3-9 a).
Evacuation routes in
buildings
Stairs (Main evacuation
route)
Lifts (Alternate evacuation
route)
Lifts with unprotected lobby for
the entire population
Lifts with protected
lobby
Lifts with double protected lift lobby for
the entire population
Lifts for the entire
population
Lifts for 25% of the
population
Stairs with double protected smoke
lobby
Stairs for the entire
population
Stairs for 75% of the
population
Parts of one evacuation strategy
70
Concept Design ‘B’ (Lifts with Protected Lobby): To restrict heat exposure to the
lift cars, concept design ‘B’ is considered with a lift lobby. The lift lobby is enclosed
with the fire resistive walls, floors and self closing fire doors (see Figure 3-9 b).
Figure 3-9 – Three Concept Designs for risk analysis Concept Design ‘C’ (Lifts with Double Protected Lobby): The concept design
‘C’ is similar to the concept design ‘B’, except one additional door is provided in the
protected lift lobby. This provision is similar to the DTS provision for stair smoke
lobby (smoke lobby between two sets of doors). The lift lobby is enclosed with the
fire resistive walls and floors, one set of door is self closing fire doors and other can
be magnetically operated sliding door or self closing smoke door (see Figure 3-9 c).
Concept Design ‘D’ (Stairs): The concept design ‘D’ is an option for which the
associated risk is deemed acceptable. This option was included as a reference for the
comparative study.
Concept Design ‘E’ (Stairs and Lifts): The concept design ‘E’ is an option for the
use of protected lift system to evacuate 25% of the population and the remaining
population by stairs.
Unprotected lift lobby Double protected lift
lobby
Protected lift lobby
Smoke lobby
FD Public corridor
(a) (b) (c)
SD
Legend: FD – Fire Door
SD – Smoke Door
71
3.1.7 Risk Quantification
In order to determine the risks associated with the use of lifts, risks are quantified
and evaluated. The risks are quantified with the help of following models:
a. Stochastic models for evacuation using lifts and stairs; and
b. Fire hazard computational models for the determination of time for fire
hazards to occur; and
c. Probabilistic risk analysis for lift reliability issues.
Stochastic evacuation models estimate the building evacuation times. The model is
developed within the context of occupant load and building space, where the
evacuation time would tend to be constrained by human factors (social, physiological
and psychological characteristics) and limited by flow rate capacities of evacuation
routes. The fire hazard computational model determines the time to exceed the
tenability limits, or the available safe egress time (ASET) (ABCB, 2005b). The
results are compared with the safe evacuation time, or the required safe egress time
(RSET) (ABCB, 2005b), for determining the impact on building occupants and the
concept of ‘safety index’ is applied. The reliability of lift operational mechanism is
determined from standard risk assessment techniques. The results from building
evacuation models, fire hazard computational model and probabilistic risk analysis
form an integrated base for risk assessment.
3.1.8 Risk Assessment
The risk assessment is basically a structured approach to decision making under
uncertainty (Watts and Hall, 2002). To determine expected impact upon the building
evacuees, Multi-Objectives Decision Analysis (MODA) based on Analytical
Hierarchy Process (AHP) is used.
The parameters causing the risks are given weights (importance or merit). They are
selected from the literature review based on professional judgments and past
experience (for example, decision uncertainty may be caused by long waiting time
for lifts and the number of evacuees standing in queue). Weights are given by a
72
simple analytical method based on survey and statistics (for example, weights are
given to each component for causing deaths from fire, toxic gases or pre-existing
health conditions). After fixing the weights to the parameters, the value of each
parameter is obtained from the stochastic evacuation model, fire hazard model and
probabilistic risk analysis model for each concept design and strategy. These
parametric values may be different for each concept design based on risk involved.
The value of each parameter is multiplied by its weight and all the weighted values
are added to give a final risk value. The weighted values represent the risks
associated with the individual concept design on a comparative basis (see Chapter 7).
The risk analysis does not cover fire-cost expectation (FCE). The FCE includes the
capital cost for the passive and active fire protection measures, the maintenance cost
of active fire protection measures, and expected loss from the fire (Meacham, 2002).
The FCE quantifies the fire cost associated with the particular fire safety system
design.
3.1.9 Selection of Design Options
After determining the weighted risk values, the selection of concept design can be
made in light of the acceptable risk. The selection is rational, transparent and
systematic for decision making.
3.2 Research Work
Figure 3-10 outlines the research work for achieving the objectives. It presents the
proposed research activities based on the following chapters.
In Chapter 4, the residents’ choice for using evacuation routes is modelled using a
pilot survey. This determines the residents’ willingness and acceptance of the use of
lifts for emergency evacuation. Data from the pilot survey are used for risk
assessment. Interviews were conducted with the fire brigade personnel. An
illustrative case of 38 storey hypothetical apartment building is prepared for risk
analysis. A stochastic model for lift evacuation is developed for determining the lift
time periods and the number of evacuees in queue. These output variables are
73
determined for the entire population and one-fourth of the population (25%) for lift
evacuation. The output variables for one-fourth of the population are determined for
evaluating a safe and efficient lift evacuation strategy. The proportion of 25%
population is based on the literature review and survey findings (see Chapter 4,
Section 4.2.3). Stochastic model for stair evacuation is also developed for
determining the stair time periods and the number of evacuees in queue. These
output variables are determined for the entire population and three-fourth of the
population (75%) in the stairs.
Figure 3-10 – Research Work Flow Diagram
Risk identification
for the use of lifts
Issues of fire hazard Issues of lift operational
mechanism
Issues of human
behavioural response
Design considerations
and strategy evaluation
Stochastic evacuation models for lifts and stairs
Fire hazard modelling for lifts
and stairs
Risk assessment
(An integrated approach)
Pilot survey and
interviews
Probabilistic risk
Analysis
Establish available safe evacuation time and parametric analysis
Establish required safe evacuation time for lifts
and stairs
Safety indices for
the lifts and stairs
Parametric analysis for ‘Decision Uncertainty’
Parametric analysis for
‘Panic’
Parametric analysis for
‘Injuries’
74
In Chapter 5, models for fire and smoke hazards in evacuation routes are proposed
for determining the time to exceed the tenability limits relating to visibility, CO,
CO2, O2 (low oxygen-causing hypoxia), temperature and radiant heat flux. Twenty
four fire scenarios are analysed after incorporating uncertainties relating to wind and
vertical location with FDS model (McGrattan et al., 2004). The safety indices are
evaluated for the lift and stair systems. The safety index estimates the probability that
the escape time would exceed the available time. The fire hazard model does not
take into account self closing device of SOU door or operation of sprinkler.
In Chapter 6, the reliability of lift operational mechanism for water damage, lift
malfunctioning and electric power failure are considered for probabilistic risk
analysis. The analysis is based on statistics and standard risk assessment techniques.
The techniques include a complex parallel and series system (a modified form of
event tree analysis) and fault tree analysis. Output variables from FDS models are
used for determining the lift malfunctioning.
In Chapter 7, the Multi-Objectives Decision Analysis method is used for risk
assessment. The parameters are assessed and statistics from various sources are used
for an integrated risk assessment approach. Each parameter is given a degree of
importance (weight). For decision uncertainty, survey data and statistics reports are
analysed for giving weights. The output variables are obtained from the stochastic
evacuation models. For panic, statistics relating to unavailability of evacuation routes
and visual threat (during pre-life threatening condition) are considered for giving
weights. For life threatening conditions, statistics relating to fire deaths are analysed
for giving weights. The risk priorities were derived from analytic hierarchy process.
After giving parametric weights and values, risk assessment is conducted for design
options and evacuation strategies. Sensitivity analysis is conducted for adjudging the
importance of variables. In Chapter 8, the feasibility and design considerations for
lift evacuation system are determined. The redundancy measures are proposed.
75
3.3 Conclusion
The research methodology involves the following main steps:
(a) risk identification, risk priorities and expected impact upon the building
evacuees.
(b) establishing acceptable level of risk for design options and evacuation
strategies.
(c) quantifying risks with the stochastic models for building evacuation, fire
hazard models for determining time to exceed tenability limit and
probabilistic risk analysis for lift reliability issues.
(d) conducting risk assessment using Multi-Objectives Decision Analysis
method and determining suitable design options and evacuation strategies.
Risk cause-effect relationship identifies the key issues to be addressed. The issues of
human behavioural response, fire hazards and lift operational mechanism give rise to
three hierarchical levels of risks i.e. decision uncertainty, panic and injuries (nonfatal
and/ or fatal). These risks are interlinked, multi-dimensional and require a complex
process for risk assessment. The research strategy involves risk management by
reducing the risk level to an acceptable level using various concept design options
and evacuation strategies. The risks are quantified by using building evacuation
simulation models, fire hazard models and probabilistic risk models. The Multi-
Objectives Decision Analysis method is proposed for risk assessment. The priorities
of the three risks are assigned on a ratio scale. Risk assessment is conducted based
on parametric values obtained from the models. The feasibility of alternative design
model is determined in light of the acceptable risk.
This chapter has addressed risk identification, expected impact upon the building
evacuees, concept design options and stairs as an acceptable level of risk. The next
chapter is related to stochastic building evacuation model.
76
4. STOCHASTIC MODELS OF BUILDING EVACUATION
4.1 Introduction
Required safe evacuation time (RSET) is an important parameter in fire safety
engineering (ABCB, 2005b). RSET is defined as the time period, subsequent to fire
alarm, required for safe occupant evacuation. RSET depends on several factors, such
as the physical dimensions of evacuation path (length and width of corridors, stairs
and lifts), fire detection and alarm, the occupant density, number and distribution,
and the occupants’ social, physiological and psychological characteristics. The
physical and human factors play a vital role in the evacuation procedure and are
often not given due consideration. The building evacuation is generally affected by
uncontrollable and random arrival pattern of occupants and causes the output to be
random as well. A considerable period of time is lost in pre-evacuation activities.
Earlier researches have been conducted to predict the building evacuation time
(Sekizawa et al., 1996 and Kuligowski, 2003). The approaches of these researches
were deterministic and did not include physical and human factors. However, the
results have demonstrated that using both stairs and lifts could provide a better
performance (Sekizawa et al., 1996 and Kuligowski, 2003). The approach used in
determining the evacuation times should include human factors for realistic
scenarios.
This chapter is focused on developing stochastic models for determining RSET using
lift and stair systems. Stochastic models reflect uncertainty due to variability or
randomness in data. The discrete event simulation using ARENA (Rockwell, 2000)
gives the capability to develop realistic time based models for evacuation using lifts
and stairs. The model simulates the lift cabin operation and computes output
variables relating to RSET (such as lift waiting time, lift transportation time and lift
evacuation time) and the number of evacuees waiting in queue. Such performance
parameters are also developed for the stair system for comparison with the lift
system. The output variables are determined for the entire building population and
for one-fourth of the building population. The output variables for one-fourth of the
building population are determined with the intent of reducing RSET so that
77
evacuees are not subjected to a long waiting period and exposed to toxic hazards of
fire effluents (RSET is reduced with the evacuation of one-fourth of the building
population). RSET is compared with the available safe evacuation time (within safe
conditions) in Chapter 5. The use of lifts for a partial population can help in
developing operational strategies.
Before evaluating the output variables relating to RSET, occupants’ attitude towards
the use of lifts need to be considered as the occupants should have inherent sense of
confidence in the mechanical evacuation facility. In Hong Kong, a survey was
conducted for high-rise apartments with the objectives of determining comforts of
high-rise living and any effect it might have on children (Mori and UHK, 2002). In
response to a question relating to the disadvantages of high-rise living, 36% reported
fire escape, 20% reported lift breakdown, 2% reported strong wind, 2% reported heat
and 4% reported lack of play areas. This indicates that a significantly large
proportion of high-rise residents are concerned of fire emergency evacuation
followed by lift breakdown. The confidence on lift systems requires further analysis.
A pilot survey was therefore conducted to gain an understanding of the occupants’
attitude towards the use of lifts. Interviews were also conducted to determine the
attitude of professional fire fighters toward the use of mechanical lift evacuation
system.
4.2 Pilot Survey
4.2.1 Pilot Survey Overview
A pilot survey was conducted for a 38-storey apartment building in Brisbane. The
pilot survey was conducted using a questionnaire form, which was distributed to 250
residents in the building, out of which 20% responded. The questionnaire form and
data collected are given in Appendix C. The objectives of the survey were to:
1. determine the residents’ preferred access and egress routes in normal
circumstances; and
2. determine the residents’ preferred exit route during fire emergencies; and
78
3. determine the residents’ awareness towards the ‘Emergency Evacuation
Procedure’ (EEP) ; and
4. determine the residents’ priority and willingness to accept lift as an
alternative evacuation facility.
The residents were adjudged for the following parameters:
• age distribution in the high-rise apartment building
• residents’ inclination toward the use of lifts for emergency evacuation
• residents’ awareness for emergency evacuation procedure
• residents’ inability to use stairs and hence they need to use lifts for prompt
evacuation
The parameters were used in the analysis to determine the necessity of lifts as an
evacuation facility in high-rise apartment buildings.
4.2.2 Pilot Survey Results
Age Distribution: Residents living in high-rise apartment building were analysed for
their age distribution. This parameter is helpful in ascertaining the necessity of lifts
as an evacuation facility for aged and disabled persons. The results presented in
Table 4-1 show the aged distribution of residents. The results show that the majority
of the baby boomers (aged from 65 to 73 years) were living above the 20th floor
level.
Table 4-1: Residents’ Age Distribution
Under 30
years
30 to 44
years old
45 to 64
years old
65 years
and older
1 to 10 storeys 18% 18% 64% -
11 to 20 storeys 13% 54% 26% 7%
21 to 28 storeys 24% 46% 15% 15%
29 to 38 storeys 27% 18% 46% 9%
Overall average 20.5% 34% 37.75% 7.75%
79
Normal Entry and Exit: Residents were surveyed for their use of entry and exit
routes in the building. This helped in determining residents’ inclination toward the
use of evacuation routes. The survey results presented in Table 4-2 show lifts as
normal access and egress routes. The pilot survey results showed that 92.75% of the
residents always use lifts, 3% of the residents use lifts sometimes, and 4.25% of the
residents use them rarely on average, for normal building access and egress. These
values were obtained from the average of grading given to individual data collected
for the use of lifts and stairs during normal entry and exit. From these results, it is
clear that residents were inclined towards the use of lifts for everyday access and
egress. Some of the residents have the tendency of using stairs.
Table 4-2: Residents’ Use of Lifts as Normal Access and Egress Routes
Storey Always Sometimes Rarely Never
1 to 10 storeys 92% 4% 4% -
11 to 20 storeys 85% 4% 11% -
21 to 28 storeys 96% 2% 2% -
29 to 38 storeys 98% 2% - -
Overall average 92.75% 3% 4.25% -
Emergency Exit: Residents were surveyed for their intended use of emergency exit
routes. Residents were asked about the use of lifts or stairs during fire emergencies.
The survey results presented in Table 4-3 show that 84% of the residents, living on
the 1st to 10th storeys, 93% living on the 11th to 20th storeys, 92% living on the 21st to
28th storeys and 92% living on the 29th to 38th storeys, intend to use stairs during fire
emergencies. This means that on average 90.25% of the residents intend to use
stairs, whereas 3.75% of the residents rely on fire brigade facilities and 6% of the
residents intend to use lifts and stairs. The option of using both lifts and stairs was
also given to the residents, which they may like to use depending upon the
circumstances. However, no one intends to use lifts solely or stay in their unit. From
these results, it is clear that majority of the residents intended to use stairs in the case
of fires. However, 9.75% of the residents (mainly include aged or disabled) were not
capable of evacuating the building using stairs and they have to rely on using a
combination lifts and stairs or fire brigade facility. Permanently disabled residents
80
were on the 11th and 21st floors, while the temporarily disabled residents were on the
3rd and 18th floors.
Table 4-3: Residents’ Preferred Exit Routes during Fire Emergencies
Storey By stairs By lifts By lifts and
stairs
By fire brigade
appliance
Remain
in unit
1 to 10 storeys 84% - 8% 8% -
11 to 20 storeys 93% - - 7% -
21 to 28 storeys 92% - 8% - -
29 to 38 storeys 92% - 8% - -
Overall average 90.25% - 6% 3.75 -
Awareness of Emergency Evacuation Procedure: Residents were surveyed for
their awareness about the emergency evacuation procedure (EEP). The awareness
level helps the residents in taking rational decisions during fire emergencies. The
survey results (see Figure 4-1) show that 69% of the residents admitted that they
were not trained in EEP on average. Only 42% of those living on the upper top
levels between 29th and 38
th storeys were trained. Majority of residents living in the
middle levels between 21st and 28
th storeys stated that they were not trained in EEP.
Untrained, 83%
Untrained, 57%
Untrained, 77%
Untrained, 58%
Untrained,
69.00%
Trained , 17%
Trained , 43%
Trained , 23%
Trained , 42%
Trained , 31%
0% 20% 40% 60% 80% 100%
1 to 10 storey
11 to 20 storey
21 to 28 storey
29 to 38 storey
Overall average
% of Residents
Untrained Trained
Figure 4-1 – Residents’ Awareness of Emergency Evacuation Procedure
81
Experience with Fire Drill: Residents were asked if they participated in fire drills in
the building. Experience with fire drill determines if the residents were trained in
evacuating the building confidently. The survey results showed that 69% of the
residents revealed that they did not experience any fire drill in the building (see
Figure 4-2). Ten percent of the residents felt that they faced difficulties during the
fire drills and expressed a sense of crowdedness, bottleneck effect in movement,
narrow width of stairwell, different walking speed, knee trouble and wheel chair
movement problem. Twenty one percent of the residents were confident and
experienced no difficulty during fire drills. The survey results also indicated that 8%
of the residents observed that stair doors were locked at all times or some times.
No difficulty,
21%
No fire drill
experience,
69%
Difficulty
encountered,
10%
Figure 4-2 – Residents’ Experience during Fire Drill
4.2.3 Discussion and Conclusion
The confidence intervals were determined for the use of lifts during normal
circumstances and intended use of stairs during fire emergencies. The confidence
intervals were calculated with the help of BETAINV (see Appendix C). A 95%
confidence interval for the use of lifts during normal circumstances is determined to
be in the range of 0.849 and 0.961, whereas 95% confidence interval for the stairs as
an emergency exit is determined to be in the range of 0.8069 and 0.9526. The use of
lifts during normal circumstances is slightly higher than the use of stairs during fire
emergencies, which indicates that residents were more inclined to use lifts as a mode
of vertical transportation during normal circumstances than the use of stairs during
emergencies. The occupants can be more confident in using lifts as an alternative
evacuation facility during fire emergencies.
82
The small percentage (9.75%) of residents intending to use lifts during an emergency
indicates that the burden of managing the lift evacuation during an emergency is
small, since most of the residents will be using stairs for evacuation. However, this
small percentage does not represent the willingness of the entire population for the
use of safe and reliable lifts. If lifts are permitted as an alternative evacuation facility,
the number of evacuees using lifts may increase as others may also join during
emergency evacuation. Therefore, if the existing lift infrastructure does not provide
adequate safety to the majority of population in the building, only aged and disabled
can be considered for lift evacuation (with the limited number of occupants, output
variables relating to RSET can be reduced to achieve adequate safety). Aged and
disabled revealed their concern of life safety while living on upper levels. Some of
them commented that they had to wait in the building for fire brigade help. The total
population of aged and disabled was approximately 16% based on this survey and the
result agrees with the data published in the literature (ABS, 2004 and Pauls, 1977).
However, the population of aged persons was only 7.75%, whereas the literature
review indicated 13%. The population of disabled persons was 8%, whereas the
literature review indicated only 3% are living in high-rise apartment buildings in
Australia. If aged and disabled representing 16% of the population are using lifts and
it is assumed that half of them are assisted or helped by others (say 9%) in using the
lift evacuation route, about 25% of the population may require an alternative
evacuation facility in the building.
The survey findings also showed that the majority of residents were not trained in the
EEP. A significant population of residents reported that they never experienced or
witnessed fire drill in buildings. In such circumstances, residents may not be
confident in building evacuation. This may add complexity to their evacuation
procedure and a state of confusion may arise. It is concluded that the residents may
need to use lifts for emergency evacuation in high-rise apartment buildings. Further,
there are also chances that a significant population may also join 25% of the
population for lift evacuation.
Above discussion shows that (a) the majority of residents are inclined to use lifts in
their routine life (b) aged and disabled are concerned about their emergency
evacuation as the majority of them are residing on mid or upper top levels (c) the
83
majority of residents are not trained in EEP (d) necessity to investigate an alternative
evacuation facility for the entire population. If it is not feasible, the use of lifts can be
investigated for at least 25% of the building population. Regular drills and practices
are required to be conducted to avoid ambiguity in evacuation procedures as there are
chances that a few occupants from 75% of the population may also join 25% of the
building population for lift evacuation.
The residents near the ground floor of the building did not show their intent for using
stairs as their prevalent ingress or egress route. Thus, efforts were also made to
conduct surveys in other buildings of similar nature and height, but building
corporate management did not give permission. Further, there were a few other
limitations of this pilot survey, which are given below:
- The sample size was limited. The results generated from a sample of 50
residents in one apartment building did not necessarily present adequate
information for the general population in apartment buildings.
- The data on the choice of emergency exit route reflected the building
occupants’ intentions expressed when filling in the questionnaire, which may
not be the true indication of what they will do in a real emergency.
4.3 Interviews
Some proponents indicated that the lifts in protected lift lobby can be used as an
evacuation facility (Sekizawa, et al., 1996) and that protected lift lobby space can be
used as staging area for fire department (Kuligowski and Bukowski, 2004).
Opponents argued that the lift lobby protection would not provide additional safety.
One of the papers presented by the regular fire brigade officers strongly disagreed
with the intent of proposing lift as an evacuation facility in the buildings
(O’Donoghue and O’Donnel, 2003). Although, the use of lifts has always been a
controversial issue, the occupants have used lifts as an escape route on many
occasions (Barnett et al., 1992, see also Appendix A). Therefore, interviews were
conducted with fire brigade personnel to determine their willingness toward the use
of lifts as an evacuation facility. Three interviews were conducted with the officers
of regular fire services in Australia.
84
The objectives of the interviews were to:
1. determine the interviewees’ perception toward the use of lifts in high-rise
buildings; and
2. know the fate of general public, disabled and aged persons during fire
emergencies.
The questions asked from the emergency officials and their replies are given in
Appendix D. The summary of the interview findings are given below:
1. All interviewees expressed their concern about the impeded movement of
occupants in stairs as a result of over-crowding, bottle neck, counter current
flow between fire brigade personnel and evacuees.
2. Interviewees stated that disabled, aged, children and physically weak persons
can go to refuge areas (such as lift lobbies) for temporary staging; then use
lifts and/or stairs with assistance as needed. Interviewees also recommended
that lifts can be explored as an option in high-rise buildings.
3. Interviewees mentioned that experienced people act rationally and occupants
use stairs for emergency evacuation. One of the interviewees stated that
about half of the population behave rationally depending upon the
circumstances.
4. All interviewees expressed that lifts can be considered as an option for
disabled and elderly people. However, lift systems should satisfy design,
commissioning and maintenance requirements, which can prove to support
safety of the evacuees.
5. Interviewees informed that fire emergency evacuation drills are not regularly
held in apartment buildings as fire drills are not mandatory as per the
provisions of Building Fire Safety Regulation.
6. Comfortable travelling via stairs depends upon the safety measures in the
stairs and individual’s strength. However, there are re-entries from stairs to
building floor levels at every 4th level in high-rise buildings.
7. In a working fire environment, fire fighters use lifts or stairs depending upon
the fire conditions in the building.
8. Electric power supply is generally reliable in Australia, although power
outage can not be ruled out.
85
9. Interviewees recommended that provisions of pressurisation and lift stack
arrangement can be considered for lift evacuation system.
It should be pointed out that the information obtained from the interviews is
qualitative. The accuracy of the figures given by the interviewees was not verified.
4.4 Analysis of Building Evacuation Periods
To determine the risks relating to human behavioural response and fire hazards, the
variables relating to building evacuation periods (or RSET) are analysed. The
objectives of the analysis are to:
1. develop a model for the building lift evacuation under uncertainties
associated with human social, behavioural and physical movement along with
a priori heuristics of the lift domain and determine the probable time for safe
evacuation; and
2. develop a model for the building stair evacuation under uncertainties and
determine the probable time for safe evacuation; and
3. compare lift evacuation and stair evacuation times for ascertaining human
behavioural response (decision uncertainty); and
4. examine the potential ways of reducing the building evacuation time periods.
4.4.1 Methodology
Two stochastic models of lift and stair evacuation systems are developed to
determine the probability of successful evacuation. The model analyses the passenger
optimum service level for lifts in comparison to stairs. The following variables are
determined:
• lift waiting time, lift transportation time, lift evacuation time and number
of evacuees in queue
• stair waiting (queuing) time, stair travelling time, stair evacuation time
and number of evacuees in queue
86
Lift evacuation time: Lift evacuation time is the total time period for evacuating the
building occupants using lifts. The lift evacuation time tLE for the number of
evacuees in a building can be expressed as:
xiLE tttt = t ∪∪∪∪∪ ..........21 4.1
where x is the number of evacuees
(Union symbol ∪ shows that lift evacuation time includes the evacuation time of all
individual evacuees)
The lift evacuation time of an individual ti can be expressed as:
LILTLWLMLPMFDi tttttt = t +++++ 4.2
where
tFD is the fire detection time (second)
tLPM is the lift pre-movement (coping and response) time for an evacuee (second)
tLM is the movement time for an evacuee (second)
tLW is the lift waiting time for an evacuee (second)
tLT is the lift transportation time for an evacuee (second)
tLI is the evacuee intermittent floor movement time via lift (second)
A combination of lift pre-movement time, movement time and lift waiting time is the
lift pre-evacuation time. The lift pre-evacuation time can be expressed as:
LWLMLPMLPE tttt ++= 4.3
Lift evacuation time of an individual can be considered as the lift evacuation time
from the building, if that individual takes the maximum time to evacuate the
building. It can be expressed as:
87
tt LE imax= 4.4
Stair evacuation time: Stair evacuation time is the total time period for evacuating
the building occupants using stairs. The stair evacuation time tSE for the number of
evacuees x in a building can be expressed as:
xiSE tttt = t ∪∪∪∪∪ .........21 4.5
while the evacuation time of an individual ti can be expressed as:
SISTSMSPMFDi ttttt = t ++++ 4.6
where
tSPM is the stair pre-movement (coping and response) time for an evacuee (second)
tSM is the movement time for an evacuee (second)
tST is the stair travelling time for an evacuee (second)
tSI is the evacuee intermittent floor movement time through stairs (second)
Further, stair evacuation time of an individual can be considered as the stair
evacuation time from the building, if that individual takes the maximum time to
evacuate the building. It can be expressed as:
tt SE imax= 4.7
Pre-evacuation activities are considered the same as the evacuees are likely to choose
lift or stair at a later stage. Therefore, stair pre-movement time and lift pre-movement
time are the same for both the evacuation routes. Similarly, stair movement time and
lift movement time are considered the same as both evacuation routes are considered
equidistance from the dwelling units.
The methodology used in this analysis is given in Figure 4-3. The lift simulation
model is based on the conventional lift group controller (see Section 4.5.2), which is
88
the most prevalent lift group controller system used in apartment buildings. The stair
simulation model is based on the most prevalent stair design in apartment buildings.
The lift simulation model generates the results of performance parameters (RSET
and queue length) during down peak traffic. The results from simulation are analyzed
in comparison with that from the stair simulation model. Output variables are plotted
with the help of a computer program @RISK (Palisade Corp, 1996) for obtaining the
mean and standard deviation (see Chapter 2, Section 2.10), which are used in
Chapter 5 in determining the fire hazards.
Figure 4-3 – Flow Diagram for Analysing the Output Variables of Models
If the results generated by the simulation models are not acceptable, the parameters
(lift performance parameters or lift logic controller) can be developed/ modified, for
which the advanced technologies such as genetic algorithm, reinforcement learning,
Lift supervisory
controller and variables
Lift simulation model
Determine variables
relating to stair evacuation
Hypothetical building
model and variables
No
Yes
Determine variables
relating to lift evacuation
Comparing lift and
stair evacuations
Acceptable level
Establish parameter for
simulation model
Stair simulation model
Develop performance parameters for future
research
Analysis of output
variables
Verification of
proposed models
Lifts for a limited number
of evacuees
Modify lift simulation
model
89
fuzzy logic or neural networks lift group controller need to be considered for
apartment buildings. These technologies are not currently used in apartment
buildings; however, they are used in office buildings for efficient transportation.
Researches have shown that these technologies have significant time saving of up to
25% to 40% in transportation (Siikonen, 1997 and Cortes et al., 2004). The
technologies include advanced adaptive systems with short-term and long-term
memories and ongoing calculation for how much time has elapsed between initial
call and arrival of the car in order to prioritize call. If advanced technologies are used
in the lift evacuation system, the performance parameters need to be modified for the
evacuation model. This research begins with the lift time periods and the number of
evacuees in queue determined within the acceptable level for a fraction of population
(25% of the population, which may include aged and disabled). The partial
evacuation of 25% of the population is estimated from the literature review (ABS,
2004 and Pauls, 1977) and pilot survey (see Section 4.2.3).
The simulation models can never be validated over the whole range of their
behaviour (Phillips, 1995). Hence, the lift evacuation model was verified using the
results from a building egress model, ELVAC (Klote et al., 1991) while the stair
evacuation model was verified with a mathematical expression. ELVAC is a
deterministic model and is not capable of incorporating random variables and
distributions. Hence the lift and stair evacuation models were verified without
incorporating random variables. The verification is limited to the output parameters
that are common to stochastic and deterministic models. The common parameters of
both the models are identified in Section 4.9.
4.4.2 Discrete Event Simulation
The building egress models such as ELVAC (Klote et al., 1991) and EVACNET
(Kisko et al., 1998) are based on deterministic approach and assume exact
knowledge of input parameters. This causes a discrepancy between the evaluated
results and real situations. An alternative approach is to model the uncertainty by
introducing random variables, in which many numerical observations are made as a
probability distribution for obtaining a condensed approach of final results. The
90
condensed approach is the average value of the results with minimum, maximum and
half width (or 95% confidence level) of multiple replications.
The non-deterministic stochastic modelling reflects inherent variability found in
physical system parameters that demonstrate random behaviour (Hoeksema and
Kitandis, 1985). This modelling approach is intended to incorporate the uncertainties
associated with physical system in predicting the system behaviour. In a
probabilistic risk assessment, He et al. (2003) adopted a simple approach where the
building occupant evacuation was treated as a Poisson process and no differentiation
was made between pre-movement and movement activities. Vistnes et al., (2005)
later employed a stochastic approach to estimate the time associated with the pre-
movement activities.
The modelling is intended to represent all the uncertainties associated with physical
system. Lift evacuation time and stair evacuation time (or RSET) are quantified by
defining random variables relating to pre-evacuation activities, human social and
physical movement. ARENA (Rockwell, 2000) is a commercially available package
for stochastic modelling. ARENA is the animation component of the SIMAN
language, which is a powerful general purpose simulation language for modeling
discrete, continuous and combined systems. SIMAN has a logical modeling
framework in which the simulation problem is divided in to model and experiment
components (see Figure 4-4). The model describes the physical and logical elements
of the system. The experiment specifies the experimental conditions (inputs) under
which simulation runs including the initial condition, attributes, running time and
replications. Once a model and an experiment have been defined, they are linked and
executed by SIMAN language to generate the simulated response of the system. The
SIMAN output analyzer is used to generate plots, tables and bar charts. ARENA has
been used successfully in many major projects including railway stations, airport
terminals, fire departments, manufacturing and processing industries and high rise
buildings (for evacuation purpose). This discrete model is used in the analysis of
output variables relating to the issues of human response and fire hazards where it
handles the behaviour of thousands of occupants, hundreds of queues and the
dynamics of tens of lifts.
91
Figure 4-4 – SIMAN Flow Diagram
4.5 Model Framework
In order to determine the maximum efficiency of lift and stair evacuation systems,
ARENA version 10.0 software was used to simulate the possible events. Passenger
traffic in the building can be described as down peak traffic during fire emergencies.
The occupants arrive at the lift lobby or stair lobby on upper floor levels and travel
down in lifts or stairs to the ground floor. This section presents the details of a
hypothetical building model, and experimental (variables) components used in
simulating the lift and stair evacuation systems.
4.5.1 Hypothetical Building and Parameters
A hypothetical building shown in Figure 4-5 with 38 floors was considered in the
analysis. The building was analyzed with the following assumptions:
• A symmetrical fire occurrence on three levels as different scenarios viz.
lower level (2nd
floor), middle level (19th floor) and top most level (38
th
floor) in the building. Fires on three floors are considered independently
Model components
Complier
Model object file Experiment object file
Complier
Experiment components
Linker
SIMAN program file
Execute
Results
92
to determine the effects on evacuation times. Evacuation times will be
used in Chapter 5.
• Emergency evacuation routes on the ground floor are suitably
compartmented to restrict the spread of smoke and hot gases. It complied
with the deemed to satisfy provisions of the building code.
• Pre-movement activities commence following a fire alarm.
• Evacuating population is evenly distributed at both the stairs.
• The service levels for both stairs and lifts are considered to be their
maximum capacity.
Figure 4-5 – Typical Floor of a Hypothetical Building (57m × 20m)
The test problems based on the generic apartment building, lift configurations and
traffic scenarios have been generated. Table 4-4 shows the details of building,
occupants, lifts and stairs. The generic design of the building and variables were
prepared with the intent of giving equal evacuation opportunities for lifts and stairs.
The main terminal (ground floor) is indexed 1. Occupants’ characteristics, such as
walking speed and behavioural response, are given in Sections 4.5.3 and 4.5.4.
93
Table 4-4: Model Parameters
Parameter Building data
Total floors 38
Floor population (each level) 32
Height between floors (m) 3
Number of lifts 4
Max lift capacity (persons) 16
Max velocity (m/s) 3.15
Acceleration (m/s2) 1.0
Number of stairs 2
Width of stairs (m) 1.2
4.5.2 Lift Supervisory Controller for Lift Simulation Model
The orders in which the lift cars respond to the landing calls play a vital role in the
performance of a vertical transportation system. Normally apartment buildings are
provided with a simplistic supervisory lift group controller. The proposed simulation
model is based on a conventional lift group control system. The flow diagram for lift
operation used in the model is shown in Figure 4-6.
The controller implements dispatch rules that make use of an IF-ELSE logical
command set (Cortes et al., 2004). The dispatch rules are simulated in the computer-
aided design suite LSD (Lift Simulation and Design), under the designation of the
THV algorithms (implemented under this designation at the University of
Manchester Institute of Science and Technology). This system collects the
information in most common rules in duplex or triplex algorithms. The duplex or
triplex algorithm assigns the hall call to the nearest lift in the travelling direction.
This method is bound to expert knowledge and a priori heuristics of the lift
application domain. Whenever an evacuee presses the hall call button for a lift,
micro-processor type lift group controller mounted in the lift machine room receives
the request and logs it for future reference and work on the principle of moving in
one direction at a time. Lift heading to higher floors is programmed to ignore calls
for lower floors until they have reached the top. Each hall call is attended by only
one lift car. The lift can stop at a floor only if there is a hall call from or a cabin call
94
to that floor. The cabin calls are served in accordance with the cabin’s current
travelling direction.
Figure 4-6 – Lift Controller Logic Diagram used for ARENA Simulation Model
There are two sets of doors, which work in tandem in order to safely allow evacuees
to enter and exit the cars. One set of doors remains shut until a lift car's presence is
detected. The other door is controlled by the lift's controller and opens sideways as a
powerful electric motor pulls the first panel. Once both doors are open, evacuees
95
leave the lift to allow new evacuees to board and more calls to be answered. The lift
car door possesses a sensor that detects if someone is obstructing the door and door
closing mechanism stops the closing of doors and remains in the open position until
the obstruction is removed. It is noted that the lift doors are open for 5 seconds. A
25% of additional delay is conservatively taken during emergencies as evacuees may
hold the lifts. Acceleration and deceleration functions as indicated in Table 4-4 are
added. If the lift car exceeds its capacity (weight), lift sends a signal to the lift group
controller not to pick up any more evacuees until the evacuees disembark the car.
4.5.3 Lift Simulation Variables
In order to consider a realistic situation, certain variables such as occupant arrival
rate, occupants’ characteristics (social), response and movement (physical – based on
age factor) are incorporated in the model. The times between occupants’ arrival to
the common evacuation route (such as lifts and stairs) are determined from fire
detection time, pre-movement (response and coping) time and movement time.
Occupant movement: Social groups within the building stay together throughout
the duration of the evacuation. Someone would like to ascend the lift to look for his
relatives or to travel inter-floor in the building. Building body corporate or
emergency crew would also like to ascend upper floor. For the current model, the
down-movement, inter-floor movement and up-movement are specified below:
� From the upper floors – all 1184 (32 × 37) occupants would be using lifts to
go to ground or intermittent floors
� Intermittent floors – 12 occupants would be using lifts for intermittent floor
movement (approximately 1% of total population – 1184 occupants)
� From the ground floor to the upper floors – 22 outsiders would like to go to
upper levels (approximately 2% of total population)
– Total number of occupants using lifts = 1206
Fire detection time (tFD): FDS model is used to determine the fire detection time,
which is 90 seconds for fire compartment (smoke alarm in SOU – individual alarm
96
only and not connected to public alarm) and 140 seconds for corridor (smoke
detector in public corridor and connected to public alarm) (see Chapter 5, Section
5.6).
Pre-movement (response and coping) time (tLPM): The pre-movement time is the
time spent by the occupant before starting the evacuation. It includes the time
required to perceive the event and the time required to respond to the event. The pre-
movement times can vary from seconds to many minutes (occupants are awake,
trained, familiar with the building, alarm systems and procedures). This is arguably
one of the poorest documented attributes in fire safety engineering. The existing
methods used to determine occupants’ response and coping are often inadequate and
insufficient to predict patterns of occupants’ spatial movement. Brennan and Thomas
(2001) argue that there is a difficulty with the traditional assumption that occupants
confronted with fire will react to fire. They proposed a paradigm shift from a reactive
to an interactive model of behaviour in fires. Occupants are likely to be intimately
involved in interactions with fire. However, occupants may have minimal interaction
with fire due to the speed it developed and these occupants are usually asleep or
otherwise unaware until escape is impossible.
Based on interactive model of behaviour of fires, the pre-movement (response and
coping) time of occupants in apartment buildings is selected from Fire Engineering
Guidelines (FCRC, 1996). The method considers the occupants’ physical, mental and
social conditions during the fire response and coping period and evaluates a
deterministic time period. The occupant response time based on primary weighting
factors such as alertness, commitment and familiarity and secondary weighting
factors such as role, focal point, mobility, social affiliation and position is determined
(see Appendix E). The coping time based on primary weighting factors such as role,
mobility and social affiliation and secondary weighting factors such as alertness,
position, commitment, focal point and familiarity is also determined. The calculated
response time is 186 seconds while the coping time is 255 seconds (see Appendix E).
Movement time (tLM): The horizontal walking speed of occupants is based on the
density. If the density is smaller than 0.54 persons/m2, the walking is free to a value
97
of 1.1 m/s. If the density is greater than 3.8 persons/m2, the occupants can not move
(Nelson and MacLennan, 1995). An average horizontal walking speed of 0.6 m/s for
the aged and disabled is conservatively considered (Nelson and MacLennan, 1995).
The travel distance varies between 9 m and 27 m. Therefore, the movement time of
occupants travelling to the lift lobbies is taken from 15 seconds (minimum) to 45
seconds (maximum). During this period, all the occupants are assumed to travel to
the lift lobby.
Time to arrive in lift lobby: Fire detection time, pre-movement time and movement
time are added to determine the time to arrive in the lift lobby. However, occupants
may also arrive without engaging themselves in response and coping activities.
Therefore, on the fire-affected floor:
• Minimum time to arrive in lift lobby = tFD + tLM
= 90 + 15 = 105 seconds
• Maximum time to arrive in lift lobby = tFD + tLPM + tLM
= 140 + 186 + 255 + 45 = 626 seconds
On other floors:
• Minimum time to arrive in lift lobby = tFD + tLM
= 140 + 15 = 155 seconds
• Maximum time to arrive in lift lobby = tFD + tLPM + tLM
= 140 + 186 + 45 = 371 seconds
(Coping time is not included here since the residents do not encounter fire).
The occupants arrive between the minimum and maximum times. These times are
considered as random variables in the stochastic modelling.
Occupant arrival rate: Occupant arrival rate is the average number of occupants
arriving in the lift lobby per unit time. Lift system provides discrete mode of vertical
transportation. The occupants arrive for lift evacuation in groups or one at a time in a
period t according to a Poisson process (Alexandris, 1977). Assuming that the
98
probability of x calls are registered for lifts in time t for an average rate of occupants’
arrivals is λ, then
( )!
)(
x
etxp
tx λλ −
= 4.7
where
t is the time period of the counting process (second)
x is the number of occupants who arrive at designated point
p(x) is the probability mass function (PMF) for x number of people arriving within
the time period t
λ is the average number of occupants arriving per unit time (s-1)
The PMF given by Eq. 4.7 is the Poisson (λ t) distribution. The values of occupant
arrival in the lift lobby are expressed in Poisson distribution with the help of ARENA
input analyser. ARENA input analyser is a component of ARENA model to
determine the quality of fit of probability distribution function to input data (number
of occupants and time variables). Figure 4-7 shows the occupant arrival for the entire
floor population and one-fourth of the floor population. The Poisson distribution
predicts occupants’ arrival to lift lobby as a peak load at any given time. The
probability of occupant arrival is uniformly distributed between two extreme time
intervals. The calculated values are between 105 and 626 seconds on fire-affected
and ground floors, and 155 and 371 seconds on other floors for 32 occupants (entire
population of the floor) and 8 occupants (one-fourth of the floor population). The
expressions are given as Poisson (16.6) on the fire-affected floor, Poisson (6.81) on
the rest of the floors and Poisson (23.5) on the ground floor for the entire floor
population {Poisson (16.6) – distribution function contains the value of mean}.
Similarly the expressions are given as Poisson (67.6) on the fire-affected floor,
Poisson (27.6) on the rest of the floors and Poisson (89.5) on the ground floor for
one-fourth of the floor population.
99
Figure 4-7 – Poisson Distribution for Occupant Arrival (Lifts)
A discrete function of 25% delay is added to the model, which may be due to
occupants holding the lift for their next of kin (the discrete function in ARENA
returns a sample from a user-defined discrete probability distribution). A 25% of
additional delay is conservatively considered keeping in view the aged and disabled
population and their assistance. However, a general population can also hold the lift.
The inter-floor movement is considered from occupants’ movement time from lift to
apartment unit in 20 seconds, staying in apartment for 20 seconds and return in 20
seconds on average and therefore an exponential distribution function with a mean of
20 seconds is used for all. Exponential distribution contains a value of mean and
provides a random variable for stochastic modelling. These times are added for
social reasons.
Lift movement speed: Lift movement speed varies with the hall call and cabin call.
Velocity and acceleration are incorporated in the stochastic modelling.
100
4.5.4 Stair Simulation Variables
The building indicated in the lift model was considered for stair analysis. On each
floor, there are two stairways. Arrival patterns of the occupants are the same at lift
lobbies and stair lobbies. However, the floor population is equally distributed on both
the stairs with half the floor population using one stair (whereas there is only one lift
lobby and the entire floor population is arriving at one lift lobby). Total time frame
for evacuees’ arrival at lift lobby or stair lobby is the same.
Occupant movement: The down-movement, upper floor movement and inter-floor
movement are defined below:
� From the upper floors – 50% of building population, 592 (16 × 37) occupants
would be using one stair to go to ground or intermittent floors
� Intermittent floors – 6 occupants would be going to intermittent floors
(approximately 1% of total population – from 592 occupants)
� From the ground floor to the upper floors – 6 outsiders would like to go to
upper levels (approximately 1% of total population – 592 occupants)
– Total number of occupants using one stair = 598
Movement from the ground floor to the upper floors in the stairs is considered for
building corporate management and general public, not the emergency crew
members as they are permitted to use lifts during fires. Disabled and aged persons
can go to stair lobbies for temporary staging and then can be assisted by others.
Pre-movement and movement times: The pattern of occupant arrival for stairs is
considered similar to lift pre-movement (response and coping) and movement time.
The calculated response time is 186 seconds and coping time is 255 seconds. The
movement time of occupants travelling to the stairs is taken as 15 seconds
(minimum) to 45 seconds (maximum). During this period, all the occupants would
have travelled to stairs.
Occupant arrival rate: Figure 4-8 shows the occupant arrival rate for the entire
floor population and three-fourths of the floor population. The expressions Poisson
101
(34) on the fire-affected floor, Poisson (14.5) on other floors and Poisson (33) on
ground floor are considered for the entire floor population (see Figure 4-8).
Similarly, the expressions Poisson (44.8) on the fire-affected floor, Poisson (18.5) on
other floors and Poisson (49.3) on ground floor are considered for the three-fourths
of the floor population. These distributions are given after equal distributions of
population between the two stairs in the building.
Figure 4-8 – Poisson Distribution for Occupant Arrival (Stairs)
Stair capacity: It is generally considered that a walker on stairs needs to perceive
two vacant treads ahead and occupies an area of approximately 0.7 m2. Thus free
flow design is possible at a density of 0.6 P/m2 and full flow design is possible at a
density of 2.0 P/m2. The walking speed is considered to be 1.25 P/m2, which will
permit passer-by, intermittent floors movement or reverse flow, although severely
restricted (Kisko et al., 1998). The stairway is 1.2 m wide and considering a 27
degree stairway with 3 m between floors and with an intermediate landing, the
102
horizontal distance walked in descending one floor is the ratio of 3 m to the tangent
of 27 degrees, which works out to be 11.4 m. Therefore the area is 13.68 sq.m.,
which would hold approximately 17 persons in the stair in one floor height.
Stair walking speed: Walking speed in stairs varies from 0.52 to 0.62 m/s (0.57) for
general public and 0.22 to 0.79 m/s (0.505) for occupants carrying children on
average in non-crowded stairs in apartment buildings (Proulx, 1995). The walking
speed of occupants carrying children varies to a large extent when compared to
general public. The speed was 0.45 m/s for small children (aged 2-5, not carried by
adults) and 0.43 m/s for over 65 years old (Proulx, 1995). The research data of stair
walking speed of aged people is over optimistic. It is not expected that aged and
physically weak people can maintain their travel speed all the way to the ground.
Aged and physically weak people may require rest in the stairs. This may cause
impedance to the traffic flow. There is a percentage of population who are physically
immobile and can not use stairs at all. If lifts are not permitted, they have to wait for
fire brigade intervention and therefore take much longer time to evacuate.
Keeping in view the emergency situation for flowing condition, a walking speed
varying from 0.43 to 0.79 m/s with an average of 0.54 m/s is considered at a density
1.25 P/m2 for stair analysis. The average of 0.54 m/s is obtained based on the 75%
of normal population having a walking speed of 0.43 m/s and the 25% population,
including aged and disabled persons, having a walking speed of 0.43 m/s. A
triangular distribution for stair travelling time of 32 travellers with a minimum of
13.4 seconds, mode (most likely value) of 19.6 seconds, and maximum of 24.6
seconds is obtained for one floor level (see Figure 4-9). Triangular distribution is
skewed with a mode of 19.6 seconds. This distribution is applied to all floor levels.
Figure 4-9 – Triangular Distribution for Occupants’ Stair Travelling Time
103
4.6 Simulation Models
4.6.1 Lift Simulation Model
The block diagram indicated in Figure 4-10 shows the lift simulation model for the
hypothetical building. The model is comprised of four major zones: controller and
lift zone, passenger zone, re-entry zone and exit zone. Animation Zone is supportive
for demo and is not shown here. The various input components relating to the lift
system used in these zones are such as entity – occupants; attributes – upper levels
and main terminal (ground level); activities – vertical traveling; events – arrival at
the lift lobbies and arrival at the main terminal; state variables – number of occupants
waiting at each floor level and number of evacuees in transit and resource – number
of lifts.
Elev ato r Ariv a lS tation 1As s ign 1
Arriv a l Floor 01Pas s enger
Arriv a l Floo r 02Pas s enger
Tr ue
False
Dec ide 1
S tation 2 Dis pos e 1
A partment 02 D elay 1 R oute 1
Floor 02. Q ueue
Queue
Floor 01. Q ueue
Queue
ReleasedAt Floor ==What Floor
Floor Num ber ==What Floor
( Elevat or _Load < Elevat or _Capacit y) * (What Floor > Floor Num ber )Else
Dec ide 2
As s ign 2
Can_Fi t_1_Pas s enger
Tr ue
False
R oute 2
As s ign 3
As s ign 4
As s ign 5
O r iginal
Member s
D ropoff 1
S tation 8
Passenger Zone
Controller and Lift Zone
Re-entry Zone
Route 5
As s ign 6
Elevat or _Rem ainingCapQ ueues_2(What Floor )
P ickup
As s ign 8
TWhat Floor ( 1) | | ( NQ (Q ueues_2( Floor Num ber ) ) < 1) * ( Dispose 1. Number out <1210) * EWhat Floor ( 2)
W hi le
E ndW hi le
FloorD ropP ickD elay
( Q E( 1) + Q E( 2) + Q E( 3) + Q E( 4) + Q E( 4) + Q E( 5) + Q E( 6) + Q E( 7) ) + Elevat or _Load >= 1
S can
Create 43 As s ign 9 Rec o rd 1H old 1 Dis pos e 2As s ign 10
As s ign 11
Exit Zone
0
0
0
0
0
0
0 0
Figure 4-10 – Various Zones in Lift Simulation Model (two floors only)
Assign modules have attributes of lift number, lift load, floor number, released at
floor, lift remaining capacity and acceleration. Station module contains the set of
floors. Decide modules have conditions of expression for lift load, lift capacity, floor
number and where to go. Delay module assigns the value for pick-drop delay during
the lift movement and occupants’ momentarily staying in an apartment during
intermittent floor movement. The passenger exit zone consists of the location of
104
evacuees with an arrival expression on various floor levels. The evacuees are in the
waiting queue for lifts in the lobby. The passenger re-entry zone consists of the
location for evacuees’ re-entry in the apartment unit and lift waiting station.
4.6.2 Stair Simulation Model
The block diagram indicated in Figure 4-11 shows the stair simulation model for the
hypothetical building.
S tation 2 Dispose 1
A partment 02 Delay 1 Route 1
Floor 37Passenger Arriva l
S tation 65
Floor 38Pas senger Arriva l
S tation 66
A partment 38 Delay 39 Route 39
Route 40
S tation 80
Route 41
S tation 81 R oute 42
Route 43
S tation 117
Route 114
S eize 1
S eize 2 Release 1
Decide 1
Want edFloor == 37
Floor _ent er ed == 37 && Want edFloor > 37
Floor _ent er ed == 37 && Want edFloor < 37
Want edFloor > 37Else
Ass ign 2
Release 39
Route 115S eize 78
Route 117S eize 79
Route 118S eize 80 Release 40
Release 41 Route 119
Ass ign 74
Ass ign 109
Decide 73
Want edFloor == 38
Else
Release 147 Route 295
Floor 01Passenger Arriva l
S tation 155 R oute 296Ass ign 110
Decide 74
Floor _ent er ed == 01 && Want edFloor > 01
Else
Route 297S eize 221
Stair Traveller Zone
Re-entry Zone
Exit Zone
0
0
0
Figure 4-11 – Various Zones in Stair Simulation Model (three floors only)
The model is comprised of three major zones: stair traveler zone, re-entry zone and
exit zone. These zones contain several modules representing various attributes and
variables. Assign modules have attributes of stair capacity and walking speed.
Station module contains the set of floors. Decide modules have conditions of up or
105
down movement. The traveller exit zone consists of the location of travellers with an
arrival expression on various floor levels. The travellers’ re-entry zone consists of the
location for travellers’ re-entry in the apartment unit. The delay module indicates the
travellers’ momentarily staying in apartment during intermittent floor movement.
Appendix F gives the details of SIMAN ARENA language and the overall model
used for lift and stair systems (Figures F1 and F2).
4.7 Simulation Results
For down peak traffic scenario, 100 replications of two hour traffic for the maximum
evacuation period were generated. The test results are based on full building
evacuation using lifts and stairs separately. Detailed results are given in Appendix G.
The animations of both simulation models are shown in Figure 4-12.
Figure 4-12 – Animation of Lift and Stair Simulation Models at 300 seconds
106
4.7.1 Lift Simulation Model
Lift for Total Evacuation: Issues relating to lift evacuation system were adjudged
for three floor levels (viz. bottom level, middle level and top level). Output variables
relating to RSET (lift waiting time, lift transportation time and lift evacuation time)
and the number of evacuees in queue were analyzed for 100 replications (see
Appendix G, Tables G1, G2 and G3). Tables 4-5 to 4.7 give the values of the lift
time periods and the number of evacuees in queue for fire occurring at three levels.
Table 4-5: Lift Time Periods and Number of Evacuees in Queue (2nd floor fire)
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Lift Waiting Time,
tLW (second) 721.24 8.11 601.69 827.20 0 2371.36
Lift Transportation
Time, tLT (second) 41.83 0.21 39.73 44.31 1.34 251.80
Lift Evacuation
Time, tLE (second) 2321.74 105.48 1158.68 3169.28 51.69 3516.24
Number of Evacuees in Queue
8.17 0.87 0.90 19.96 0 34
Table 4-6: Lift Time Periods and Number of Evacuees in Queue (19th floor fire)
Identifier Average
Half
Width
Minimum
Average Maximum
Average Minimum
value
Maximum
value
Lift Waiting Time,
tLW (second) 719.66 7.89 610.03 810.20 0 2402.26
Lift Transportation
Time, tLT (second) 41.92 0.23 39.49 44.37 1.50 297.59
Lift Evacuation
Time, tLE (second) 2269.30 111.36 1176.37 3291.86 4.03 3564.95
Number of Evacuees in Queue
8.14 0.86 1.02 19.48 0 34
Table 4-7: Lift Time Periods and Number of Evacuees in Queue (38th floor fire)
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Lift Waiting Time,
tLW (second) 721.95 8.50 605.19 839.67 0 2406.00
Lift Transportation
Time, tLT (second) 41.83 0.25 39.57 42.12 1.49 279..78
Lift Evacuation
Time, tLE (second) 2270.00 116.39 1180.71 3403.47 20.85 3403.47
Number of Evacuees in Queue
8.19 0.86 0.94 20.00 0 34
107
The average is the mean time period of 100 simulations. The "Half Width" column is
included to determine the reliability of the results from all the replications. A value is
interpreted by saying "in 95% of repeated trials, the sample mean would be reported
as within the interval sample mean ± half width". The maximum and minimum
averages are the averages of simulations. The maximum and minimum values
indicate the maximum and minimum of simulations. It may be inferred from the
similar results presented in the above three tables that the model did not consider the
impact of the location of the fire floor on evacuation.
Lift for 25% Population: Considering the use of lifts for a limited number of
evacuees, the lift time periods and the number of evacuees in queues were calculated
for three floor levels (see Appendix G, Tables G1, G2 and G3). The results are
shown in Tables 4-8 to 4-10.
Table 4-8: Lift Time Periods and Number of Evacuees in Queue (2nd
floor fire) –
25% population
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Lift Waiting Time,
tLW (second) 299.91 7.36 211.23 424.95 0 1336.34
Lift Transportation
Time, tLT (second) 40.09 0.30 36.61 44.16 1.96 237.43
Lift Evacuation
Time, tLE (second) 1613.20 74.88 785.97 2319.05 0.88 2349.90
Number of Evacuees in Queue
1.28 0.14 0.22 3.48 0 9
Table 4-9: Lift Time Periods and Number of Evacuees in Queue (19th floor fire) –
25% population
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Lift Waiting Time,
tLW (second) 294.96 7.34 208.79 387.43 0 1391.90
Lift Transportation
Time, tLT (second) 40.30 0.30 36.48 43.89 1.95 283.52
Lift Evacuation
Time, tLE (second) 1586.56 71.77 706.84 2401.98 8.52 2459.25
Number of Evacuees in Queue
1.28 0.14 0.22 3.52 0 9
108
Table 4-10: Lift Time Periods and Number of Evacuees in Queue (38th floor fire)–
25% population
Identifier Average
Half
Width
Minimum
Average Maximum
Average Minimum
value
Maximum
value
Lift Waiting Time,
tLW (second) 298.13 6.55 241.85 403.91 0 1261.44
Lift Transportation
Time, tLT (second) 40.36 0.34 36.91 45.43 1.49 254.90
Lift Evacuation
Time, tLE (second) 1560.94 80.42 874.57 2306.55 3.63 2306.55
Number of Evacuees in Queue
1.26 0.14 0.24 3.47 0 9
It can be seen from the lift simulation results that building population was reduced by
three quarters but the average lift waiting time, lift transportation time and lift
evacuation time were only reduced by just over half. However, the number of
evacuees in the queue has been reduced considerably.
4.7.2 Stair Simulation Model
Stair for Total Evacuation: Evacuees’ using stairs was analyzed for evacuation
periods for 100 replications (see Appendix G, Tables G4 and G5). Tables 4-11 to 4-
13 give the values of stair travelling time and stair evacuation time. The stair waiting
time varied from 0 to 9 seconds while the number of travellers in queue was less than
0.1. The stair waiting time and the number of travellers in queue were not significant
in relation to bottleneck or queuing, and are therefore not given.
Table 4-11: Stair Time Periods (2nd
floor fire)
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Stair Travelling
Time, tST (second) 373.10 0.45 367.77 377.90 14.62 1554.79
Stair Evacuation
Time, tSE (second) 1556.03 60.05 800.35 2114.19 17.58 2114.19
109
Table 4-12: Stair Time Periods (19th floor fire)
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Stair Travelling
Time, tST (second) 372.69 0.48 367.40 379.44 14.62 1534.87
Stair Evacuation
Time, tSE (second) 1544.13 62.35 917.38 2157.91 18.71 2157.91
Table 4-13: Stair Time Periods (38th floor fire)
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Stair Travelling
Time, tST (second) 372.82 0.74 366.30 378.38 14.62 1510.38
Stair Evacuation
Time, tSE (second) 1559.61 57.42 822.04 2002.17 10.85 2002.17
Stair for 75% Population: For strategic evaluation, occupants’ evacuation for 75%
of the population was also analyzed for evacuation periods for 100 replications (see
Appendix G, Tables G4 and G5). Tables 4-14 to 4-16 give the values of stair
travelling time and stair evacuation time.
Table 4-14: Stair Time Periods (2nd floor fire) – 75% population
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Stair Travelling
Time, tST (second) 366.98 0.35 363.40 372.06 14.52 1058.87
Stair Evacuation
Time, tSE (second) 1102.51 39.68 619.42 1476.53 11.44 1476.53
Table 4-15: Stair Time Periods (19th floor fire) – 75% population
Identifier Average
Half
Width Minimum
Average Maximum
Average Minimum
value
Maximum
value
Stair Travelling
Time, tST (second) 367.72 0.38 362.63 373.80 14.62 1192.78
Stair Evacuation
Time, tSE (second) 1153.90 42.06 505.61 1462.94 66.98 1631.48
Table 4-16: Stair Time Periods (38th floor fire) – 75% population
Identifier Average
Half
Width
Minimum
Average Maximum
Average Minimum
value
Maximum
value
Stair Travelling
Time, tST (second) 367.14 0.49 362.42 373.72 14.69 1382.95
Stair Evacuation
Time, tSE (second) 1312.93 50.78 672.45 1701.60 86.57 1701.60
110
It can be seen from the stair simulation results that even when the building
population was reduced by one quarter, the average stair travelling time and stair
evacuation time were only reduced marginally.
4.8 Analysis of Results
The parameters i.e. lift waiting time (tLW), lift transportation time (tLT), lift pre-
evacuation time (tLPE), lift evacuation time (tLE) and number of evacuees in queue,
were analysed for the hypothetical 38-storey building. The occupant profile, building
symmetry and evacuation strategy were the same for both lift and stair evacuation
systems. The results show that the lift and stair evacuation times are not varying
significantly irrespective of the origin of fire at floor levels in the building.
4.8.1 Lift Waiting Time
While evacuating the entire building using the lifts, a few evacuees are required to
wait for a considerable time near the lift lobby before the first lift arrived (see
Appendix G, Tables G8, G10 and G12). However, the minimum and maximum
waiting time values reaching from 0 to 2400 seconds can not be considered an
isolated lowest or highest values (see Tables 4-5 to 4-7). These are low or high
values and represents an approximation to the most efficient or worst case
eventualities for all extreme values on the floors. The maximum average lift waiting
time is approximately 1000 seconds and is considered justified (see Figure 4-13).
0
200
400
600
800
1000
1200
0 5 10 15 20 25 30 35 40
Floor Level
Time (second)
2nd floor fire
19th floor fire
38th floor fire
Figure 4-13 – Lift Waiting Times during the Fire Occurrences at Three Levels
111
During down peak traffic in the event of fires, a lift car may fill at two, three or four
floors at different levels and then makes a run to the ground floor. Although this
reduction in number of stops results in a shorter round trip time and higher handling
capacity, the result predicted that the evacuees on the lower floors may face a slightly
longer waiting period. Figure 4-13 also shows that the waiting time is lower at 30th
level (dip in the graph), which is due to the fact that the evacuees arrive at the early
stage than the upper levels and use the lifts for downward movement (a stochastic
input). Waiting time is also slightly lower at the fire-affected floor, which is caused
by longer time to evacuate from the unit due to occupants’ coping action.
Mean lift waiting time was approximately 720 seconds from 100 replications. The
lifts served within 297 seconds with the reduction in number of evacuees (see
Appendix G, Tables G6 and G7). The probability distribution functions (PDFs) were
given with the help of @RISK (Palisade, 1996) for determining the mean and
standard deviation. The PDF and a comparison are shown in Figure 4-14 (a), (b) and
(c).
Logistic(720.913, 16.582)
672.1
5.0%
769.7
95.0%
0
0.004
0.008
0.012
0.016
640 660 680 700 720 740 760 780 800
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(a) PDF for Lift Waiting Time (100% population)
112
Weibull(2.9525, 72.645) Shift=+232.880
259.4
5.0%
338.2
95.0%
0
0.005
0.01
0.015
0.02
240 260 280 300 320 340 360
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(b) PDF for Lift Waiting Time (25% population)
0
100
200
300
400
500
600
700
800
900
0 20 40 60 80 100
Replication
Time (second)
Lift Waiting Time (100% Population)
Lift Waiting Time (25% Population)
(c) Comparing the Average Lift Waiting Times
Figure 4-14 – Lift Waiting Times
4.8.2 Lift Transportation Time and Stair Travelling Time
Mean lift transportation time was approximately 42 seconds for 100% population and
approximately 40 seconds for 25% population. The mean stair travelling time was
approximately 372 seconds for 100% population and 366 seconds for 75%
population (see Appendix G, Tables G6 and G7). The lift transportation times and
stair travelling times are not varying significantly with the change in evacuation
proportion. This shows that there is unnecessarily withholding of lift and no
bottleneck in the stair. The PDFs and comparison are shown in Figure 4-15 (a) to (e).
113
InvGauss(13.901, 4826.739) Shift=+27.966
40.675
5.0%
43.127
95.0%
0
0.1
0.2
0.3
0.4
0.5
0.6
40 40.5 41 41.5 42 42.5 43 43.5 44
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(a) PDF for Lift Transportation Time (100% population)
Logistic(40.22769, 0.55930)38.581
5.0%41.875
95.0%
00.050.10.150.20.250.30.350.40.45
37.5 38 38.5 39 39.5 40 40.5 41 41.5 42 42.5
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(b) PDF for Lift Transportation Time (25% population)
LogLogistic(351.289, 21.534, 25.466)
370.471
5.0%
375.462
95.0%
00.050.10.150.20.250.3
0.350.4
369 370 371 372 373 374 375 376 377
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(c) PDF for Stair Travelling Time (100% population)
114
Logistic(367.24305, 0.76537)
364.989
5.0%
369.600
95.6%
0
0.1
0.2
0.3
0.4
363 364 365 366 367 368 369 370 371 372
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(d) PDF for Stair Travelling Time (75% population)
0
100
200
300
400
0 20 40 60 80 100
Replication
Time (second)
Lift Transportation Time (100% population)
Lift Trasnportation Time (25% population)
Stair Travelling Time (100% population)
Stair Travelling Time (75% population)
(e) Comparing the Average Lift Transportation and Average Stair Travelling
Times
Figure 4-15 – Lift Transportation and Stair Travelling Times
4.8.3 Lift Pre-Evacuation Time and Stair Pre-Evacuation Time
The lift pre-evacuation time was obtained from lift pre-movement time, lift
movement time and lift waiting time. The lift pre-evacuation time is basically the
RSET on the fire-affected floor (if the lift shafts are free from fire effluents), and is
used in the next chapter. The pre-movement time was calculated from simulation
timer on a fire-affected floor. Mean lift pre-evacuation time was approximately 1346
seconds for 100% population and 924 seconds for 25% population. The PDFs and
comparison are shown in Figure 4-16 (a) and (b).
115
Logistic(1346.913, 16.582)
1298.1
5.0%
1395.7
95.0%
0
0.004
0.008
0.012
0.016
1260 1280 1300 1320 1340 1360 1380 1400 1420
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(a) PDF for lift pre-evacuation time for 100% population
Weibull(2.9525, 72.645) Shift=+858.880
885.4
5.0%
964.2
95.0%
0
0.005
0.01
0.015
0.02
860 880 900 920 940 960 980 1000
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(b) PDF for lift pre-evacuation time for 25% population
Figure 4-16 – Lift Pre-Evacuation Times
As the stair waiting time was 0 to 9 seconds and did not reflect any significance of
bottleneck or queuing, stair pre-evacuation is confined to stair pre-movement time
(coping and response times) and stair movement only. Therefore, stair pre-movement
time is not given here.
116
4.8.4 Lift Evacuation Time and Stair Evacuation Time
The mean lift evacuation time was approximately 2288 seconds while the mean stair
evacuation time was approximately 1550 seconds, which is less, in comparison. The
mean lift evacuation time for 25% of the population was approximately 1624
seconds and close to the stair evacuation time of the entire population (see Appendix
G, Tables G6 and G7). The mean stair evacuation time for 75% population was
approximately 1193 seconds. The PDFs and comparison are shown in Figure 4-17 (a)
to (e).
BetaGeneral(1.4834, 1.2550, 1259.2, 3158.9)
1469
5.0%
3036
95.0%
0123456789
1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200
Time (second)
F(t)
Values x 10^-4
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(a) PDF for Lift Evacuation Time (100% population)
Triang(877.79, 1856.6, 1914.5)
1103
5.0%
1860
95.0%
0
0.0007
0.0014
0.0021
0.0028
900 1120 1340 1560 1780 2000
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(b) PDF for Lift Evacuation Time (25% population)
117
Triang(772.11, 1971.5, 2128.0)
1057
5.0%
2025
95.0%
00.20.40.60.81
1.21.41.61.8
800 1000 1200 1400 1600 1800 2000 2200
Time (second)
F(t)
Values x 10^-3
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(c) PDF for Stair Evacuation Time (100% population)
BetaGeneral(2.9555, 1.0058, 502.56, 1428.40)
837.6
5.0% 1412.2
95.0%
0
0.0021
0.0042
600 700 800 900 1000 1100 1200 1300 1400 1500
Time (second)
F(t)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(d) PDF for Stair Evacuation Time (75% population)
0
500
1000
1500
2000
2500
3000
3500
0 20 40 60 80 100Replication
Time (second)
Lift Evacuation Time (100% population)
Lift Evacuation Time (25% population)
Stair Evacuation Time (100% population)
Stair Evacuation Time (75% population)
(e) Comparing the Average Lift Evacuation and Average Stair Evacuation Times
Figure 4-17 – Lift and Stair Evacuation Times
118
4.8.5 Number of Evacuees in Queue
The number of evacuees in queue ranged from 8.14 to 8.19 during the use of lifts for
the entire population (see Appendix G, Tables G9, G11 and G13). No queue was
observed for 25% of the population as the number of evacuees was less than 2 (see
Figure 4-18 (a), (b) and (c)).
Lognorm(3.2721, 0.43962) Shift=+4.9022
7.505
5.0%
8.943
95.0%
00.10.20.30.40.50.60.70.80.91
7 7.5 8 8.5 9 9.5
Number of Evacuees in Queue
F(x)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(a) PDF for Number of Evacuees in Queue (100% population)
ExtValue(1.22448, 0.11107)1.1026
5.0%1.5544
95.0%
00.51
1.52
2.53
3.54
4.5
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7
Number of Evacuees in Queue
F(x)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
(b) PDF for Number of Evacuees in Queue (25% population)
119
0
1
2
3
4
5
6
7
8
9
10
0 20 40 60 80 100
Replication
Number in Queue (person)
Number in Queue (100% population)
Number in Queue (25% population)
(c) Number of Evacuees in Queue
Figure 4-18 – Number of Evacuees in Queue in Lift System
4.8.6 Findings
It is determined that lifts provide longer evacuation time periods in comparison to the
use of stairs only. However, using lifts for a limited number of persons (25%
population) gives approximately equal evacuation time periods. The use of lifts for
evacuation can be explored for a limited population because
• there is no queue (less than 2 persons in queue – nearly 1.2 persons only).
Evacuees will not have competitive behaviour for lifts; and
• a total of lift waiting and transportation times (298 second + 40 seconds) is
less than the total stair travelling time (373 seconds). Hence evacuees’
anxiousness or decision uncertainty would be minimal.
The results also showed that for apartment building having 38 storeys with 32
persons on each floor, the stairs are the fastest way of evacuation for the entire
population. Figure 4-17 gives the means and standard deviations of output variables
derived from the PDFs with the help of @RISK. These variables will be used in
Chapter 5. The results presented in the table did not demonstrate much benefit with
regard to timings in using lifts. In a residential building, the evacuation times by lifts
become more favorable for buildings with more than 100 persons per floor (50
persons per staircase) and with more than 50 floors, or for 200 persons per floor and
with more than 25 floors (Siikonen and Hakonen, 2003). The generic residential
120
building is lower than 50 floors and the population per floor is less than 200. As per
observations made by Klote et al., (1993a) (see Chapter 1, Section 1.1) the use of
lifts for emergency evacuation is also for mega high-rise buildings and are not
applicable to normal high-rise buildings.
Table 4-17: Means and Standard Deviations of Output Variables
Output Variables Evacuation Strategy Mean µL (second)
Standard
Deviation σL (second)
100% population 720.91 30.07 Lift Waiting Time (tLW)
25% population 297.70 23.90
100% population 41.86 0.74 Lift Transportation Time (tLT)
25% population 40.22 1.01
100% population 1346.91 30.07 Lift Pre-Evacuation Time (tLPM)
25% population 923.70 23.90
100% population 2288.30 489.56 Lift Evacuation Time (tLE)
25% population 1623.9 302.83
100% population 8.17 0.43 Number of Evacuees in Queue (x)
25% population 1.28 0.14
100% population 372.87 1.54 Stair Travelling Time (tST)
75% population 367.24 1.38
100% population 1549.6 237.83 Stair Evacuation Time (tSE)
75% population 1187.81 181.8
4.9 Model Verification
Computer model ELVAC (Klote et al., 1991) was used to verify the results of lift
evacuation model. One hundred replications were made in order to analyze the
performance of ARENA model. During 100 replications, a deterministic approach
was used without incorporating variables relating to human movement, inter-floor
and upward movements. A pilot run was made in order to perform the system
verification from ELVAC. The parameters for ELVAC model include number of
persons, number of floors, floor height, number of lifts, lift capacity, velocity,
acceleration, trip inefficiency and type of doors. The parameters of human movement
such as pre-movement time and movement time are not included in the model
verification. The details of ELVAC model and results are shown in Appendix H.
121
ELVAC determined the time taken to evacuate the entire building as 1590.5 seconds
while ARENA determined it to be 1618.8 seconds.
A mathematical derivation was used to verify the results of stair evacuation model.
The travel time via stairs was calculated for the evacuees in the building. The
formula established by Melinek and Booth (1975) for calculating the evacuation time
(Eq. 2.8, Chapter 2, Section 2.8.2) was used for the evacuees travelling down the
stairs (see Table 4-18). Hand calculations are given in Appendix H. Hand
calculations of the time taken to evacuate the entire building gave 1116 seconds
while ARENA determined it to be 1111 seconds (Table 4-18). The variations in
simulation results for both the systems (lift simulation model and stair simulation
model) are less than 2%.
Table 4-18: Verification of ARENA model for Lifts and Stairs
Building Evacuation Time (second) Mode of
Evacuation ARENA Model ELVAC/ Hand Calculation % Variation
By lifts 1618.8 1590.5 (ELVAC) + 1.78
By stairs 1111 1116 (Hand calculation) - 0.44
4.10 Conclusion
The pilot survey demonstrated that residents were inclined to use lifts as an
alternative evacuation facility in buildings. The confidence interval of lift use during
normal circumstances was slightly more than the use of stairs during emergencies.
The occupants can be more confident in using lifts as an alternative evacuation
facility during fire emergencies. Aged and disabled revealed their concern for life
safety while living on upper levels and can be considered for lift evacuation. At least,
25% of the population can be considered using an alternative evacuation facility in
the building. The majority of residents were not trained in the Emergency
Evacuation Procedure. The use of lifts for limited population may add complexity in
the evacuation procedure. Fire brigade personnel also expressed their concern about
the current situation in high-rise building evacuations.
122
With the help of stochastic evacuation models, the lift time periods and the number
of evacuees in queue were determined for the lift evacuation system. Random
variables relating to human social, physical characteristics and a priori heuristics of
the lift domain were considered. The results did not demonstrate much benefit with
regard to total times in using lifts. The time for which occupants would wait at a lift
station was uncertain. However, if lifts are permitted for use by 25% of the
population, the lifts serve the individuals within the acceptable time frame in
comparison with stairs. The number of evacuees in queue was less than 2 persons.
This signifies that the human behavioural response such as decision uncertainty
would be minimal for one-fourth of the population using the lift system. This aspect
will be further analysed as an integral part of other risks in Chapter 7. Lift evacuation
time is related to the total building evacuation that needs to be compared with time to
exceed the tenability limits of life threatening conditions in the evaluation of safety
indices. With the help of field model (FDS), available safe evacuation times (ASETs)
will be determined in the next chapter.
123
5. FIRE HAZARD MODELS OF LIFE THREATENING CONDITIONS
5.1 Introduction
Available safe evacuation time (ASET) is an important parameter in fire safety
engineering (ABCB, 2005b). ASET is defined as the time period between fire
initiation and onset of life threatening conditions. ASET depends on a variety of
variables associated with the fire scenario, including the intensity of the fire and the
geometry of the fire (distribution of fire load), fire protection measures and building
spatial environment. ASET should be reasonably greater than the required safe
evacuation time (RSET) for a successful building evacuation during fire
emergencies. In the absence of adequate fire safety measures, smoke and hot gases
may spread to evacuation routes (for example, lift lobbies, lift shafts and stair shafts)
and occupants at the fire-affected floor and other floors may expose to hazardous
conditions leading to psychological and physiological effects. The parameters such
as smoke visibility, CO, CO2 and O2, temperature and radiant heat exposure are used
in evaluating hazardous conditions. Such parameters can be obtained from fire tests
or computational models. The time to untenable conditions that may lead to
incapacitation or death is predicted from the combination of smoke toxic products
and heat exposure.
This chapter is focused on the development of fire hazard models under variable
conditions to determine ASET. Given a typical configuration and fuel load in an
apartment building (BCA 2005 Specification Figure 2.1 specifies 5 MW fire for
residential buildings), the time to attain incapacitating dosage is obtained. With
known intake rate of intoxicants by human, times to exceed tenability limits are
computed using the field model ‘FDS’ (McGrattan et al., 2004). The parameters
relating to fire and smoke are probabilistic in nature. By specifying uncertainty
(stochastic) in the input variables, the variable outcomes are obtained for defining a
probability distribution. Wind speed and stack effect influence the smoke spread in
high-rise buildings (Klote, 2003). Therefore, wind speed and varying vertical
location of fire are considered as input variables. The piston effect by lift car
movement is not significant in multiple shafts in high-rise buildings (Klote and
124
Tamura, 1986) and is therefore not considered. The concept of safety index is used in
the evaluation of evacuees’ safety.
5.2 Analysis of Fire Hazards
With the help of FDS, a fire hazard model was prepared using stochastic uncertainty
that is capable of predicting hazards in the evacuation routes. Various phases
occurring sequentially in space and time during the fire growth period and occupant
evacuation were specified together with the associated probability of occurrence.
The objectives of this analysis are to:
1. develop a model for the fire and toxic hazards under variable conditions and
determine the probable time when evacuees are predicted to be exposed to the
incapacitation dose of fire effluents (ASET); and
2. determine the safety of evacuees in evacuation routes under variable
conditions; and
3. analyse the effect of wind speed and stack effect on the spread of smoke and
toxic gases in the lift system; and
4. analyse potential ways of reducing the risk associated with the use of lifts.
5.2.1 Effects of Fire Effluents and Evaluation Criteria
Psychological and Physiological Effects: There are psychological and physiological
effects of fire effluents. The evacuees may be exposed to incapacitated dose of fire
effluents in evacuation routes. Smoke reduces the visibility for evacuees and there
may be needless psychological unrest or panic (Jin, 2002). Evacuees may be trapped
in the early stage of fire. However, the loss of visibility is not a direct cause of death.
Continuous exposure of smoke, heat and toxic products affect the evacuees’
capability to escape and can lead to physiological effects. The physiological effects
may be varying degrees of impaired judgment, disorientation, reduced capability of
performing work, loss of motor coordination, unconsciousness and deaths. The
physiological effects of fire effluents causing deaths are divided into the following
categories (Miller, 2005) and are used in this research:
125
– inhalation of toxic products of combustion {smoke, CO2, CO, and other
poisonous gases, hypoxia (lack of oxygen) and asphyxia}; and
– consequences of exposure to fire (burns, thermal injuries to airways and
incineration); and
– shock from injuries that precipitate death from pre-existing health conditions
(cardiac failure and respiratory diseases).
Evaluation Criteria: Evaluation criteria are based on the concept of fractional
effective dose (FED) of human incapacitation model (see Chapter 2, Section 2.6.3).
It considers concentration and time and assumes that incapacitation occurs after
adding the effects of exposure to a toxic concentration at each time period. The
following FEDs are used in calculating the effects of fire effluents:
• FED smoke: FED for smoke incapacitation is the summation of the acquired
doses of smoke. When the FED smoke reaches unity, visual obscuration is
assumed to occur. At an FED smoke value of 1.0, many evacuees are likely to
be visually obscured or impaired by smoke. The extinction coefficient should
not be higher than 0.5 m-1. Visibility is determined at 4 m horizontal
distance at a focal point in small compartments (see Appendix L).
• FED asphyxiant: FED for asphyxiant incapacitation is the summation of the
acquired doses of asphyxiant toxicants. When the FED asphyxiant reaches
unity, incapacitation is assumed to occur. At an FED asphyxiant value of 1.0,
many evacuees are likely to be overcome by the combined effects of
asphyxiant toxicants. The combined effects of asphyxiant toxicants should
not exceed one-tenth of the dose.
• FED heat: FED for heat incapacitation is the summation of the acquired doses
of convective heat and radiant heat. When the FED reaches unity,
incapacitation is assumed to occur. At an FED heat value of 1.0, many
evacuees are likely to be overcome by the combined effects of convective
heat and radiant heat. Temperature more than 60°C and radiant heat flux
more than 2.5 kW/m2 are required consideration for the effects of convective
heat and radiant heat (Purser, 2002). The combined effects of convective heat
126
and radiant heat should not exceed one-tenth of the exposure (similar to
asphyxiant).
The FED safe criterion of one-tenth of the exposure is used in the evaluation of
ASET. The criterion of FED incapacitation is not used to allow uncertainty relating
to parameters in fire models (for example, variable fire load and intensity and
building spatial environment) and the effect of irritants.
Parameters: To determine the time for occupants to receive incapacitating doses of
smoke, asphyxiant toxic gases and heat exposure, the following parameters are
determined at each discrete increment of time:
• smoke in the lift lobby, lift shaft and stair shaft
• asphyxiant toxic gases in the lift lobby, lift shaft and stair shaft
• temperature and radiant heat flux in the lift lobby, lift shaft and stair shaft
In addition, temperature in LMR is also determined. Temperature may cause
malfunctioning of lifts due to which evacuees may be trapped in lift cabins and be
exposed to hot and toxic gases.
Smoke
Fractional effective dose of smoke visibility is measured from extinction coefficient.
The results from FDS modelling give the values of extinction coefficient Cs for
direct measurement. The FED is calculated from the following expression:
5.0
ssmoke
CFED = 5.1
Asphyxiant Toxic Gases
Asphyxiant toxic gas carbon monoxide is present to some extent in all fires.
Fractional incapacitating dose of asphyxiant gas carbon monoxide (CO) is
determined. Fractional incapacitating dose (FID) refers to the state of physical
127
inability to accomplish a specific task. FIDCO is calculated using the following
expression (Purser, 2002):
30
102925.8 036.14 ppmCOFIDCO
××=
−
5.2
Carbon dioxide (CO2) is itself asphyxiant and increases the rate of uptake of other
toxic gases (hyperventilation). Where the concentration exceeds 2%, the total
fractional effective dose (FED) for asphyxiant at each time increment is multiplied
by a factor, to allow for the increased rate of asphyxiant uptake due to
hyperventilation (Purser, 2002). Therefore, concentration of CO2 is expressed in
terms of multiplication factor VCO2 using the following expression (Purser, 2002):
=5
%exp 2
2
COVCO 5.3
Oxygen (O2) is a required consideration for the effects of oxygen vitiation lower than
13% (Purser, 2002). The fractional incapacitating dose of low-oxygen hypoxia
FIDO2 is calculated using the following expression (Purser, 2002):
( )[ ]2%9.2054.013.8exp
12 O
FIDo −−= 5.4
A combined FED is calculated from the associated risks of asphyxiant gas CO, CO2
and O2. The concentrations are determined at an interval of one minute and the
cumulative effects are determined to cause incapacitation. The fractional effective
dose of incapacitation FEDasphyxiant is calculated using the following expression
(Purser, 2002):
( )22 OCOasphyxiant FIDVCOFIDFED +×= 5.5
128
The effects of other asphyxiant toxicants and irritant gases are not considered.
Heat
The evacuees may be exposed to hot environment. Purser (2002) proposes the
following two expressions for the relationship between the time to exceed tenability
limits for radiant heat and convective heat:
33.1
80
qt Irad = 5.6
where q is the radiant heat flux (kW/m2)
4.37105 −××= Tt Iconv 5.7
where T is the air temperature (°C).
The temperature and radiant heat flux are determined at a time interval and
cumulative effects are determined to cause incapacitation. The fractional effective
dose for heat is calculated by the following expression (Purser, 2002):
ttt
FEDt
t IconvIrad
heat ∆
+=∑
2
1
11 5.8
where ∆t is the time increment between t1 and t2 (minute)
129
5.2.2 Safety Index
If the probability of occupant evacuation is higher for a place of safety, the
probability of the hazardous exposure would be lower. The risk is expressed in terms
of probabilities for unsuccessful evacuation. A First Order Second Moment (FOSM)
method is used to determine the lift evacuation safety for evacuees in apartment
buildings. In the FOSM method, safety is calculated using the first two statistic
moments of the parameters viz. mean and standard deviation and by a first order
linearization of the limit state equation (hence termed as FOSM). The safety index
β quantifies the safety associated with the evacuation system during emergencies.
The safety index β is a measure of the uncertainty in the output variables and can be
used to estimate the probability that the escape time will exceed the available time.
This approach has been used by many researchers (Frantzich, 1997a, Frantzich,
1997b, Frantzich et al., 1997, Magnusson, 1997, Hasofer and Beck, 2000).
The evacuation is described by the escape time margin for the last person leaving the
threatened area. The escape time margin, expressed by the limit state function, is the
difference between the available safe evacuation time (ASET) and the required safe
evacuation time (RSET). By subjecting some variables, in the limit state function, to
uncertainty the safety inherent in the function can be determined. The escape time
margin is calculated as M = S – L where S and L are independent stochastic
parameters. The parameter S is interpreted as a strength variable (ASET) and L as a
load variable (RSET). The escape time margin is described by means and standard
deviations. The system is functional if the margin is positive i.e. the strength is
higher than the load. The mean and standard deviation of the margin can be
described as:
LSM µµµ −= and 22
LSM σσσ += 5.9
The safety index (Cornell, 1969) is calculated by:
130
M
M
σµ
β = 5.10
If the parameters S and L are normally distributed, the margin M will also be
normally distributed. The values of safety indices 0, 1, 2 and 3 are roughly equivalent
to 50, 15, 2, and 0.1% probability of failure respectively (Frantzich, 1997c). The
discrete event simulation model (ARENA) and the field model (FDS) are used in
calculating the safety index (see Figure 5-1).
Figure 5-1 – Flow Diagram for Calculating Safety Index
ARENA has already been discussed and the relevant results and analysis of load
variables (RSET) are given in Chapter 4. The strength variables (ASET) are
determined using fire hazard models (FDS models). The calculations are performed
on three concept design options viz. unprotected lift lobby, protected lift lobby (with
additional evacuation strategy for one-fourth population) and double protected lift
lobby. Safety indices are determined for both lift and stair systems.
Determine strength variables (ASET) for lift
and stair systems
Determine load variables (RSET) for
stair system
Safety Index
‘β’
Establish parameters
for FDS simulation
Determine load variables (RSET) for
lift system
Concept designs for
FDS model
Analysis of Safety Indices
Analysis of FDS output variables
Analysis of load variables (lift system)
Analysis of load variables (stair system)
131
5.2.3 Safety Index for Three Locations
Lift car follows the discrete mode of transporting the evacuees, serves the floors and
leaves a few evacuees behind. These evacuees may be incapacitated due to exposure
to fire effluents in the lift lobby. The evacuees may also encounter toxic smoke
products in the lift and stair shafts. It is important to consider the effects of fire at
three locations:
• Lift lobby – when the evacuees are waiting for lifts on the floor. Strength
variable (ASET) is determined for the location of lift lobby. Lift pre-
evacuation time tLPE is considered for load variable L1 (RSET) i.e.
t L LPE=1 5.11
and tLPE is given by Eq. 4.3
• Lift shaft – when the evacuees are in the lift shaft. Strength variable (ASET)
is determined for the location of lift shaft. Lift evacuation time tLE is
considered for load variable L2 (RSET) i.e.
tLLE
=2 5.12
and tLE is given by Eqs. 4.1 and 4.4
• Stair shaft – when the evacuees are in the stair shaft. Strength variable
(ASET) is determined for the location of stairs. Stair evacuation time tSE is
considered for load variable L3 (RSET) i.e.
tLSE
=3 5.13
and tSE is given by Eqs. 4.5 and 4.7
132
Figure 5-2 illustrates the difference between the load variables of lift lobby and lift
shaft.
Figure 5-2 – Load Variables for Safety Index for the locations of Lift Lobby, Lift
Shaft and Stair Shaft
Load variable of stair shaft is similar to that of lift shaft. Lift pre-evacuation time and
lift evacuation time are used to calculate the load variables for lift lobby and lift
shaft, respectively. Stair evacuation time is used to calculate the load variables in the
stair system. With this approach, safety index β is determined for these three
133
locations. The average safety index for the lift system is used in the comparison with
that for the stair system. The corridor is common for the lift and stair systems and
the load variables have equal effects on both of them. The load variables in the
corridor have little significance when compared with the lift and stair systems.
Hence, load variables in the corridor are not determined.
5.2.4 Field Model ‘FDS’
In order to evaluate the fire and smoke spread in the buildings, the field model ‘Fire
Dynamics Simulator’ Version 4.07 (McGrattan et al., 2004) was used. This model
solves numerically a form of the Navier-Stokes equation appropriate for low-speed,
thermally driven smoke and heat transfer from fire. Radiant heat transfer is included
via the solution of the radiation transport equation for a non-scattering gray gas. All
solid surfaces are assigned thermal boundary conditions and information about the
burning behaviour of the material. Material properties are stored in the database and
used in the simulation. Heat and mass transfer to and from solid surfaces is usually
handled with empirical correlations. This model is used to determine the flow of hot
gases in the stair shafts, lift shafts and lift machine room; thus it predicts the
conditions in the buildings.
5.3 Model Framework and Variables
By specifying the variability in the location of fire, wind speed and
compartmentation, the fire hazards in the lift system are determined in terms of
probability of the output variables for the safety indexes. The parameters describing
the variables are chosen according to judgment based on experiments and statistics.
Each distribution is then described with its mean and standard deviation.
5.3.1 Hypothetical Building Model
The hypothetical building shown in Chapter 4 is considered for detailed analysis.
Figure 5-3 shows the typical floor of the hypothetical building used in the FDS
model. Dark lines in the model show the framework of the FDS model to avoid long
computational time. All the fire safety features are assumed to be compliant with the
134
building code. The horizontal and vertical separations are provided between fire-
affected unit and other dwelling units with fire rated construction, so that fire size is
confined to fire-affected unit only. The distance between the fire-affected unit door
and the lift door is equal to the distance between the fire-affected unit door and the
stair door. Therefore the effects of fire effluents can be equally distributed.
Figure 5-3 – Typical Floor of a Hypothetical Building for a Generalised Fire
Scenario (dimensions not to scale)
Due to faulty sprinklers, sprinkler controlled fire is not considered in the analysis. A
battery operated smoke alarm (AS 3786) is provided in each apartment. Smoke
detectors (AS 1670) are provided on each level in the public areas such as public
corridor, staircase and utility ducts and services (e.g. electric shaft, communication,
air handling equipment enclosure). The sounders of the smoke detectors are located
in the public corridors in the building. The building is analyzed with the following
assumptions and justifications in the FDS model:
• Fires have equal probability of occurrence on all floors. The generic
building studied was a residential building and have identical floor layout
throughout. In order to investigate the influences of stack effect and wind
speed on the spread of smoke and hot gases in the building, three typical
vertical fire locations viz. the lower level (2nd floor), the middle level (19th
floor) and the top most level (38th floor) in the building were selected. Fire
characteristics are given in Appendix J. The location of fire was not varied
135
horizontally as this will not significantly reveal the impact of stack effect
and wind effect.
• Inadvertent opening of the door of the fire-affected unit to assume a
reasonably worse fire scenario. The apartment doors are normally of self-
closing type, however, in determining the risks, the door is assumed opened,
which may occur under pre-flashover condition, flashover condition or door
closure mechanism failure.
• The compartmentation failure of the fire-affected unit does not occur during
the fire simulation except inadvertent opening of the door as stated above.
All the units on the fire-affected floor are in compliant to BCA for the fire
resistance rating of wall, floor and ceiling.
• The doors leading to the protected lift lobby (or stairs) are assumed to be
closed. The doors are maintained through an effective and graded program
of fire prevention inspections. They will be opened only during the occupant
movement, letting smoke into the lift lobby (or stairs). Protected lift lobby
doors are modelled for the movement of entire floor population (door
opening/closing by 32 occupants), whereas stair doors are modeled for half
the floor population (door opening/closing by 16 occupants as there are two
stairs).
• No automatic or manual fire control or suppression system is considered, or
these systems are assumed to be non-operational. This assumption leads to a
worst credible scenario that prompted the need to evacuate the building.
Other fire protection measures such as smoke alarm in the fire-affected unit
and smoke detector in the public corridor are considered in the risk analysis.
The occupants may not be subjected to high risk in sprinkler protected apartment
buildings, but the possibility of building evacuation can not be ruled out. Sprinklers
do not eliminate the possibility of a fire producing large volume of smoke.
Sprinklers are also not entirely 100% reliable at all times. Therefore, sprinklers are
not considered. The evacuees in the lift cabins may be exposed to smoke, toxic gases
and heat. The FDS model does not consider the wall of lift cabin in the lift shafts, as
the lift cabin would be in motion (it can be modelled only in a stationary lift). This
model does not also consider stairs in order to avoid the specifications of
136
complicated three dimensional obstructions inside the stair and to reduce
computation time. Such a simplification is believed to err on the conservative side in
the risk assessment and the impact on the outcomes of the results is small.
5.3.2 Concept Designs
Methods to estimate the probability of failure and associated risks of the unprotected
lift lobby, protected lift lobby and double protected lift lobby are examined for the
hypothetical building (see Chapter 3, Figure 3-9 and Chapter 4, Section 4.5.1). The
pressurisation of the lift lobby and stairs is not considered in FDS modelling. Either
the smoke lobby or the pressurisation (without smoke lobby) is acceptable under the
provision of codes and regulations (see Chapter 3, Section 3.1.4). The concept design
C incorporates the smoke lobby. Stairs in all the concept designs incorporate the
smoke lobby.
5.3.3 FDS Model Boundary Conditions
The Building Code of Australia has estimated that peak Heat Release Rate (HRR) for
apartment building is 5 MW for un-sprinklered and 1.5 MW for sprinklered building
(ABCB, 2005). Therefore a peak HRR of 5 MW is used in the analysis for un-
sprinklered fire. The fire specified in the FDS model follows medium t-squared curve
fire Q= αt2, where the fire growth rate α is the fire growth coefficient (0. 01172
kW/s2)} to a constant peak value (depending on the ventilation conditions – with and
without wind). The fire was assumed to be wood, typically found in kitchen
(cupboards) and drawing rooms of apartment buildings. The size of the fire is 2.5 m
× 2.0 m × 0.6 m. The outside environment temperature and building temperature
were 20°C. The combustion yields, HRR output and temperature output are given in
Appendix J.
The limited number of random variables (varying location of fire floor, wind speed
and compartmentation) affecting the spread of smoke and fire in lift shafts are
subjected to uncertainty in fire scenarios. Therefore the building is analyzed with the
following effects:
137
• Stack effect – influence the spread of smoke and toxic gases due to
vertically varying location of fire
• Wind effect – influence the spread of smoke and toxic gases in the lift
evacuation route
Three heights were considered for varying stack effect and wind effect analysis viz.
fire at 2nd floor, wind at 19th floor and wind at 38th floor. Other layout parameters of
the building are the same. The fire floor is considered with a view to create
reasonably worse case fire scenarios in both the evacuation routes (lifts and stairs).
Reasonably worse case fire scenario is essentially a deteriorating fire condition
arising from inadvertent opening of unit doors and failure of sprinklers. The building
model is prepared for 38 storeys and therefore wind velocities are determined at
various building heights using the following formula (Dalgliesh and Boyd, 1962):
k
r
rhh
hVV
= 5.14
where
h is the height (m)
hr is the reference height (m)
Vh is the mean wind speed at height h above the ground (m/s)
Vr is the mean speed at the reference height hr above the ground (m/s)
k is the exponent for the best-fitting curve
A reference height of 10 m is internationally recommended as the standard, and
exponent for mean wind speed is taken as 0.5 for the urban areas. For an arbitrarily
selected gradient wind of 2.22 m/s (8 kmph), the mean wind speed for the selected
building is calculated from 1.4 m/s (5.05 kmph) to 7.43 m/s (26.77 kmph) depending
upon the height of the building (see Figure 5-4).
138
Figure 5-4 – Wind Speed Profile
The wind speed is given as a ramp function in the FDS model and the wind effect is
observed after breaking of window glass. The wind direction is perpendicular to the
building window. The wind speed could not increase beyond 7.43 m/s as the model
crashes frequently due to the use of different grid sizes (see Chapter 9, Section 9.3).
5.3.4 Fire Simulation Scenarios
Table 5-1 gives the concept designs and the values of variables (vertical floor
location and wind speed) used in FDS models for 24 fire simulation scenarios. In
each concept design, there are 6 fire scenarios under the variable conditions of
vertical location and wind speed. Limited number of fire scenarios are selected for
exploring the issues of fire hazard (stack effect and wind effect) on the lift
evacuation system (see Chapter 3, Section 3.1.1). However, the fire scenarios can be
extended to variable horizontal locations and functionality of fire protection systems.
4 m
55 m
112 m 5.21 m/s H
eight (Z
-axis
)
Wind Speed
1.4 m/s
7.43 m/s
Y-axis
Z-axis
Lift shaft
139
Table 5-1: Description of Fire Simulation Scenarios
Concept Design Fire Floor Level Fire Floor
Height (m)
Wind Speed
(m/s)
Fire Simulation
Scenario No.
0 1 Lower Level (2nd Floor)
3
1.4 2
0 3 Middle Level (19th Floor)
54
5.21 4
0 5
Concept Design ‘A’
Top level (38th Floor)
111
7.43 6
0 7 Lower Level (2nd Floor)
3
1.4 8
0 9 Middle Level (19th Floor)
54
5.21 10
0 11
Concept Design ‘B’
Top level (38th Floor)
111
7.43 12
0 13 Lower Level (2nd Floor)
3
1.4 14
0 15 Middle Level (19th Floor)
54
5.21 16
0 17
Concept Design ‘B’
(Evacuation for 25% population) Top level
(38th Floor) 111
7.43 18
0 19 Lower Level (2nd Floor)
3
1.4 20
0 21 Middle Level (19th Floor)
54
5.21 22
0 23
Concept Design ‘C’
Top level (38th Floor)
111
7.43 24
5.4 FDS Model Set Up
5.4.1 Conventional Domain and Grid System
A simplistic 38 storey structure incorporating relevant building features such as fire
compartment, corridor, lift shaft and stair has been considered since it is virtually
impossible to run a FDS model with a finer grid for the entire building. The
computational domain of the FDS for the building was set with main components of
the bounding walls and the roof top of the building. The computational domain of the
main building framework has a volumetric space of 57 m × 20 m × 114 m (see
Figure 5-5). The computational domain was extended beyond the physical boundary
of the building by 0.9 m (front and rear window sides). The space allows wind to
flow around the building. The front and rear boundaries were open to the outside.
140
Figure 5-5 – Computational Domain for FDS Model
FDS uses a Large Eddy Simulation (LES) model for computation and the grid size
allows the sub grid scale stress model to accurately compute the viscous stress of the
flow field. The characteristic length scale near the fire is mainly the characteristic
fire diameter D* (m) (Baum and McCaffery, 1989), as given below:
5/2*
*
=
∞∞ gTc
QD
pρ 5.15
where
Q* is the heat release rate (kW)
ρ∞ is the density at ambient temperature (kg/m3)
cp is the specific heat of gas (kJ/kg.K)
T∞ is the ambient temperature (K)
g is the acceleration due to gravity (m/s2)
141
Baum and McCaffery (1989) and Bounagui et al. (2004) recommended that grid
independence could be achieved at grid size of 0.1D* near the fire. For the fire size
of 5 MW, D* was calculated to be 1.11 m and so 0.1D* was about 0.11 m.
Therefore, the grid size was 0.1 m for the fire compartment. The grid size was 0.3 m
for the corridor, lift lobby and stair lobby, and 0.6 m for the lift shaft and stair shaft
(see Figure 5-6). The grid sizes of 0.3 m and 0.6 m were selected to reduce the
computational time.
Figure 5-6 – Three Grid Sizes used in the FDS Model (38th floor view)
5.4.2 Smoke Leakages/ Openings
Door opening and closing for lift lobbies and stair lobbies are incorporated in FDS
model for occupant’s movement as shown in Appendix K. The following openings
are considered in modelling:
• a permanent gap between the lift landing doors and the frames
• a temporary gap between the lift landing door and the frame at the time of lift
service (temporary gap operates at the time of lift service only)
• a temporary gap for the entire door width for the protected lift lobby at the
time of occupant movement
• a temporary gap for the entire door width for the protected stair lobby at the
time of occupant movement (first door for stair protection)
Stair shaft (0.6 m)
Fire compartment (0.1 m)
Lift shaft (0.6 m)
Public corridor (0.3 m)
Lift lobby
(0.3 m)
142
• a temporary gap for the entire door width for the protected stair shaft at the
time of occupant movement (second door for stair protection)
A maximum gap of 6.5 mm between lift landing door and frame is permitted (AS
1735.1, 2003) and the area calculated is 0.045 m2 per door. For two doors, the area
calculated is 0.09 m2. Therefore, permanent gap between the lift landing door and lift
frame is considered to be equivalent to a square opening of size 0.3 m × 0.3 m for
two lifts. Additional temporary gap between the lift landing door and the frame is
considered as a rectangular opening of size 0.3 m × 2.0 m since the lift cabin
provides an obstruction to smoke propagation from lobby to shaft. Hence the gap
between landing door and frame is considered to be 0.3 m of 2.0 m height (see
Figure 5-7).
Figure 5-7 – Smoke Leakage Openings in the Lift Shaft Wall
The doors protecting the lift and stair lobbies are considered to be 0.9 m. The timings
for door openings and closings were determined from ARENA simulation model,
which were further added with 3 seconds. Generally, it is observed that the duration
of door opening/closing (or door swing) is 5 seconds, whereas effective opening of
full door width is considered for 3 seconds.
5.5 FDS Output
The output points were placed in the lift lobby, lift shaft, LMR and stair shaft to
record the predicted extinction coefficient, species concentrations of CO, CO2, O2,
temperature and radiant heat flux (see Figure 5-8).
143
Figure 5-8 – Output points in the Fire Compartment, the Lift Lobby, the Lift Shaft
and the Stair
The simulation results from the output points located 1.8 m above the floor level in
the lift lobby and 3.0 m above the floor level in the lift/stair shafts are presented to
highlight the visibility, smoke and toxic gases and temperature hazards on the
evacuees. The output points are placed 3.0 m above the floor level in the stair shafts
as travellers from upper level would be coming down in the stair shaft. The output
was located 1.0 m above the floor level in LMR to highlight the temperature rise to
electronic components. Two detectors (one in fire compartment and another in
public corridor) were placed near the ceiling.
Snap shots of FDS output were taken for the 2nd
, 19th and 38
th floors. The snapshots
of smoke view and PLOT3D temperature contour for the 2nd
floor are shown in
Figure 5-9 (a) and (b). The smoke view indicates the conditions of lift shaft at 600
seconds, which is quickly filled up with smoke and hot gases.
144
(a) Smoke in the Building at 600 seconds (top view)
(b) Temperature Contour in the Corridor at 720 seconds (side view)
Figure 5-9 – Snapshots of Smoke View and Temperature Contour (Fire Scenario 1)
145
The snapshots of PLOT3D temperature contour for the 19th floor are shown in Figure
5-10 (a) and (b). Temperature in the corridor is closer to 60°C at 720 seconds at 1.8
m above the floor level while it is closer to 350°C at 1200 seconds in the lift shaft,
which arises due to the consumption of oxygen contents (yellowish-orange plume).
(a) Temperature Contour in the Corridor at 720 seconds
(b) Temperature Vector Slice in the Lift Lobby and Shaft at 1200 seconds
Figure 5-10 – Snapshots of Temperature Contour and Vector slice (Fire Scenario 3)
146
The snapshot of smoke view for the 38th floor of Fire Scenario 5 is shown in Figure
5-11. The snapshot indicates smoke spread in the unprotected lift lobby. Tenability
limits exceed quickly in the unprotected lift lobby. The visibility diminishes to 4 m in
the lift lobby at 241 seconds.
Figure 5-11 – Slice Snapshot of Visibility in the Lift Lobby (Fire Scenario 5)
147
The snapshot of PLOT3D temperature contour for the 38th floor is shown in Figure
5-12. Temperature is closer to 60°C at 1.8 m above the floor level in the corridor.
Figure 5-12 – Snapshot of Temperature Contour at 720 seconds (Fire Scenario 5)
The snapshots of temperature, radiant heat flux, CO2, CO, visibility, extinction
coefficient through lift landing door frame gap (square opening) for the 38th floor of
Fire Scenario 5 are shown in Figure 5-13 (a) to (h). Temperature and CO2 are closer
to the tenability limit criteria. Extinction coefficient is also more than the tenability
limit.
148
(a) Temperature Vector Slice through Lift Door (b) Temperature
(c) Vector Slice of Radiant Heat Flux (d) Radiant Heat Flux
(e) CO2 Concentration (f) CO Concentration
(g) Visibility (h) Extinction Coefficient
Figure 5-13 – Slice Snapshots in a Vertical Plane in the Lift Lobby at 600 seconds
(Fire Scenario 5)
149
5.6 FDS Results
The FDS model was used to calculate the times to exceed the tenability limits in the
lift and stair systems. The corridor may become untenable due to which evacuees
may not be in a position to use lifts or stairs. In such circumstances, the strategy ‘stay
in place’ may be adopted, which is another area for research. This research was
confined to explore the feasibility of using lifts on a comparative basis. It determined
that the smoke alarm operates at 90 seconds in the fire compartment while the smoke
detector operates at 140 seconds in the public corridor. FDS results are given for Fire
Scenarios 1 to 6 in Section 5.6.1 followed by analysis. The results for Fire Scenarios
7 to 24 are given in Appendix M.
5.6.1 Concept Design A (Unprotected Lift Lobby)
Figure 5-14 gives the results of time vs. extinction coefficient, concentrations of CO,
CO2 and O2, temperature and radiant heat flux in lift lobby and lift shaft for Fire
Scenarios 1 to 6. Temperature in the lift machine room (LMR) is also shown.
Fire Scenario 1 (Smoke and Gases in Lift Lobby)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m)
and CO (hundred ppm)
0
5
10
15
20
25CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
150
Fire Scenario 1 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
0
1
2
3
4
5
6
7
8
9
10
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
Fire Scenario 1 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
151
Fire Scenario 1 (Temperature and Radiant Heat Flux in Lift Shaft)
0
50
100
150
200
250
300
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
0
5
10
15
20
25
30
Radiant heat flux (kW/m2)
Temperature Radiant heat flux
Fire Scenario 1 (Temperature in LMR)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
Temperature
152
Fire Scenario 2 (Smoke and Gases in Lift Lobby)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 2 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
0
1
2
3
4
5
6
7
8
9
10
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
153
Fire Scenario 2 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scneario 2 (Temperature and Radiant Heat Flux in Lift Shaft)
0
50
100
150
200
250
300
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
0
5
10
15
20
25
30
Radiant heat flux (kW/m2)
Temperature Radiant heat flux
154
Fire Scenario 2 (Temperature in LMR)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
Temperature
Fire Scenario 3 (Smoke and Gases in Lift Lobby)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
155
Fire Scenario 3 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (c)
0
1
2
3
4
5
6
7
8
9
10
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
Fire Scenario 3 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
156
Fire Scenario 3 (Temperature and Radiant Heat Flux in Lift
Shaft)
0
50
100
150
200
250
300
0360
720
1080
1440
1800
2160
2520
2880
3240
3600
Time (second)
Temperature (C)
0
5
10
15
20
25
30
Radiant heat flux
(kW/m
2)
Temperature Radiant heat flux
Fire Scenario 3 (Temperature in LMR)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
Temperature in LMR
157
Fire Scenario 4 (Smoke and Gases in Lift Lobby)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 4 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
0
1
2
3
4
5
6
7
8
9
10
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
158
Fire Scenario 4 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 4 (Temperature and Radiant Heat Flux in Lift
Shaft)
0
50
100
150
200
250
300
0360
720
1080
1440
1800
2160
2520
2880
3240
3600
Time (second)
Temperature (C)
0
5
10
15
20
25
30
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
159
Fire Scenario 4 (Temperture in LMR)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
Temperature
Fire Scenario 5 (Smoke and Gases in Lift Lobby)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
160
Fire Scenario 5 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
0
1
2
3
4
5
6
7
8
9
10
Radiant heat flux
(kW/m
2)
Temperature Radiant heat flux
Fire Scenario 5 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m)
and CO (hundred ppm)
0
5
10
15
20
25CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
161
Fire Scenario 5 (Temperature and Radiant Heat Flux in Lift Shaft)
0
50
100
150
200
250
300
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
0
5
10
15
20
25
30
Radiant heat flux
Temperature Radiant heat flux
Fire Scenario 5 (Temperature in LMR)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
Temperature
162
Fire Scenario 6 (Smoke and Gases in Lift Lobby)
0
10
20
30
40
50
0 360 720 1080 1440 1800
Time (second)
Extinction coefficient (1/m
)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 6 (Temperature and Radiant Heat Flux in Lift Lobby)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800
Time (second)
Temperature (C)
0
1
2
3
4
5
6
7
8
9
10
Radiant heat flux (kW/m2)
Temperature Radiant heat flux
163
Fire Scenario 6 (Smoke and Gases in Lift Shaft)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800
Time (second)
Extinction coefficient (1/m)
and CO (hundred ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 6 (Temperature and Radiant Heat Flux in Lift
Shaft)
0
50
100
150
200
250
300
0 360 720 1080 1440 1800
Time (second)
Temperature (C)
0
5
10
15
20
25
30
Radiant heat flux
(kW/m
2)
Temperature Radiant heat flux
164
Fire Scenario 6 (Temperature in LMR)
0
20
40
60
80
100
120
140
160
0 360 720 1080 1440 1800
Time (second)
Temperature (C)
Temperature
(Simulation crashed after 1800 seconds under the influence of wind)
Figure 5-14 – Smoke, Gases and Heat in the Lift Lobby, the Lift Shaft and the LMR
(Fire Scenarios 1 to 6)
165
5.6.2 FDS Results Analysis
Concept Design A: Figure 5-14 highlighted the danger of smoke and toxic gases in
the unprotected lift lobby under variable conditions of wind and vertical location.
The concentration of asphyxiant gas CO has increased to about 18 500 ppm while
that of CO2 has increased from small traces to about 16%. The concentration of
oxygen reduced from 20.72% to zero. The extinction coefficient was nearly 40 m-1
(tenability limit is 0.5 m-1 for a visibility distance of 4 m).
Unprotected Lift Lobby: During the fire occurrence on the 38th floor, wind increases
the temperature in the lift lobby, but dilutes the concentration of fire effluents (see
Figure 5-15 (a) and (b)). Time to exceed tenability limit for temperature is decreased
while time to exceed tenability limit for asphyxiant is increased.
38th Floor Fire (temperature with and without wind)
0
20
40
60
80
100
120
140
160
0 500 1000 1500
Time (second)
Temperature (C)
Fire Scenario 5 Fire Scenario 6
(a) Temperature on the 38th floor lift lobby
{Fire Scenario 5 (without wind) and Fire Scenario 6 (with wind)}
166
38th Floor Fire (CO with and without wind)
0
2
4
6
8
10
12
14
16
18
20
0 500 1000 1500
Time (second)
CO (hundred ppm)
Fire Scenario 5 Fire Scenario 6
6
6
(b) CO on the 38th floor lift lobby
{Fire Scenario 5 (without wind) and Fire Scenario 6 (with wind)}
Figure 5-15 – Temperature and CO on the 38th Floor for Unprotected Lift Lobby
Temperature in the lift lobby is the function of opening and closing of lift landing
door. The opening and closing of lift landing door caused a flow of hot gases to lift
shaft. Minor dips are seen in the temperature curve at the times of lift landing door
opening (see temperature and heat flux curves in Figure 5-14). Temperature
increases if lifts are not served.
Unprotected Lift Shaft: The maximum temperature in the lift shaft during the
simulation period is 280°C for the fire occurrence on the 2nd floor, 263°C for the fire
occurrence on the 19th floor and 167°C for the fire occurrence on the 38
th floor (see
Figure 5-16 a). Temperature and radiant heat flux curves show spikes in the lift shaft
due to turbulence (whereas temperature and radiant heat flux curves are smooth in
the lift lobby). The temperature is slightly diluted under wind conditions on the
upper levels (due to ventilation from window in LMR). The maximum temperature
in the lift shaft during the simulation period is 280°C during the fire occurrence on
the 2nd floor, 209°C during the fire occurrence on the 19th floor and 163°C during the
fire occurrence on the 38th floor (see Figure 5-16 b). The concentration of oxygen is
comparatively improved with wind on upper levels.
167
Lift Shaft Temperature (without wind)
0
50
100
150
200
250
300
0 1000 2000 3000 4000
Time (second)
Temperature (C)
2nd floor fire 19th floor fire 38th floor fire
(a) Temperature in the Lift Shaft (without wind)
Lift Shaft Temperature (wind)
0
50
100
150
200
250
300
0 1000 2000 3000 4000
Time (second)
Temperature (C)
2nd floor fire 19th floor fire 38th floor fire
(b) Temperature in the Lift Shaft (with wind)
Figure 5-16 – Temperature in the Lift Shaft (with and without wind)
Unprotected LMR: The maximum limit of temperature is reached quickly in the
LMR. The temperature in the long shafts is also diluted with height. The maximum
temperature is 59°C for the fire occurring on the 2nd floor, 140°C for the fire
occurring on the 19th floor and 140°C for the fire occurring on the 38
th floor in the
early stages (see Figure 5-17). The safe limit is 43°C. There is significant increase
168
in the LMR temperature due to the fire source being closer to LMR. Major dips (ups
and downs) in Fire Scenario 5 (38th floor fire) are caused by the lift landing door
opening near LMR (see Figure 5-17).
LMR Temperature (without wind)
0
20
40
60
80
100
120
140
160
0 1000 2000 3000 4000
Time (second)
Temperature (C)
2nd floor fire 19th floor fire 38th floor fire
Figure 5-17 – Temperature in the LMR (without wind)
Concept Design B – Protected Lift Lobby, Shaft and Stairs: Figure M1
(Appendix M) shows the concentrations of smoke, toxic gases and temperature in
protected lift lobby, lift shaft and stairs under variable conditions of wind and
vertical location (Fire Scenarios 7 to 12). The concentration of asphyxiant gas CO is
about 2000 ppm. The concentration of CO2 is less than 2%. The concentration of
oxygen is not much changed. However, extinction coefficient is nearly 4 m-1. Stack
effect contributes to the spread of smoke, toxic gases, temperature and radiant heat
flux to the shaft. Wind speed also contributes to the spread of smoke, toxic gases,
temperature and radiant heat flux to the lift lobby. However, the spread of fire
effluents is reduced considerably with the provision of protected lift lobby.
Concept Design B with 25% population – Protected Lift Lobby, Shaft and
Stairs: Figure M2 (Appendix M) shows the concentrations of fire effluents under
variable conditions (Fire Scenarios 13 to 18). The concentration of asphyxiant gas
CO is less than 100 ppm. The concentration of CO2 is less than 1%. The
concentration of oxygen is also not affected. Only the extinction coefficient for
169
visibility is more than the tenability limit. Temperature is not increased much in the
lift lobby and thus no appreciable temperature rise in the lift shaft.
Concept Design C – Double Protected Lift Lobby and Shaft: Figure M3
(Appendix M) shows that the concentrations of smoke, toxic gases and temperature
in the double protected lift lobby and lift shaft (Fire Scenarios 19 to 24). The
concentration of asphyxiant gas CO is less than 100 ppm. Extinction coefficient is
increased under wind speed. Due to double protection, there is a very little change in
the lift lobby temperature.
Summary
• Time to exceed tenability limit for visibility is influenced by wind speed.
• Smoke and toxic gases stay longer in the protected lift lobby (confined
location) and are dissipated only at the time of lift service.
• There is a significant temperature increase in the unprotected lift shaft due to
‘stack effect’.
• There is a significant temperature increase in the LMR due to ‘stack effect’.
• Temperature in the lift lobby is a function of lift landing door opening and
closing. Radiant heat flux is less than 2.5 kW/m2 in the protected lift lobby.
170
5.7 FED of Smoke, Asphyxiant and Heat
Fractional Effective Dose (FED) for smoke, asphyxiant and heat are determined from
Fire Scenarios 1 to 24. Equations 5.1 to 5.5 are used for calculating the FED in lift
lobby, lift shaft and stair shaft. The calculation is given for 40 minutes, during which
the entire building can be evacuated by lifts or stairs. FED results are given for Fire
Scenarios 1 to 6 in the next section followed by analysis. Appendix N gives the
calculation for Fire Scenario 1. The results for Fire Scenarios 7 to 24 are given in
Appendix M.
5.7.1 Concept Design A (Unprotected Lift Lobby)
Figure 5-18 shows the representations for FED smoke, asphyxiant gases and heat in
Fire Scenarios 1 to 6.
Fire Scenario 1 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant
and FED heat
Smoke Asphyxiant Heat
(The representation shows FED 5; incapacitation occurs at 1)
171
Fire Scenario 1 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant
and FED heat
Smoke Asphyxiant Heat
Fire Scenario 2 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant and
FED heat
Smoke Asphyxiant Heat
172
Fire Scenario 2 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant and
FED heat
Smoke Asphyxiant Heat
Fire Scenario 3 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant and
FED heat
Smoke Asphyxiant Heat
173
Fire Scenario 3 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant and
FED heat
Smoke Asphyxiant Heat
Fire Scenario 4 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant
and FED heat
Smoke Asphyxiant Heat
174
Fire Scenario 4 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant
and FED heat
Smoke Asphyxiant Heat
Fire Scenario 5 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant
and FED heat
Smoke Asphyxiant Heat
175
Fire Scenario 5 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke, FED asphyxiant
and FED heat
Smoke Asphyxiant Heat
Fire Scenario 6 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30
Time (minute)
FED smoke, FED asphyxiant and
FED heat
Smoke Asphyxiant Heat
176
Fire Scenario 6 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30
Time (minute)
FED smoke, FED asphyxiant
and FED heat
Smoke Asphyxiant Heat
Figure 5-18 – FED in Fire Scenarios 1 to 6
177
5.7.2 Summary of FED Results and Analysis
Tables 5-2 and 5-3 give the times to exceed tenability limits in the lift lobby and lift
shaft for smoke obscuration, asphyxiant toxic gases and heat. These times based on
FED results are given in seconds. Table 5-3 also gives the times to exceed the
tenability limit for temperature in the LMR.
Table 5-2: Time to Exceed Tenability Limits in Lift Lobby
Time to Exceed Tenability Limits (second)
Asphyxiant Gases Heat
Building Fire
Scenario Smoke
Obscuration
Limit Tenability
Limit
Lethal
Dose
Tenability
Limit
Lethal
Exposure
1 241 600 900 720 1260
2 241 600 840 720 1320
3 241 600 900 720 1080
4 241 600 840 480 840
5 241 600 900 720 1140
6
Conce
pt D
esig
n A
241 660 900 480 720
7 403 2040 N.E. N.E. N.E.
8 400 2040 N.E. N.E. N.E.
9 403 2040 N.E. N.E. N.E.
10 353 1740 N.E. N.E. N.E.
11 404 1980 N.E. N.E. N.E.
12
Conce
pt D
esig
n B
367 2100 N.E. N.E. N.E.
13 508 N.E. N.E. N.E. N.E.
14 508 N.E. N.E. N.E. N.E.
15 511 N.E. N.E. N.E. N.E.
16 446 N.E. N.E. N.E. N.E.
17 529 N.E. N.E. N.E. N.E.
18 Conce
pt D
esig
n B
(25%
)
501 N.E. N.E. N.E. N.E.
19 N.E. N.E. N.E. N.E. N.E.
20 N.E. N.E. N.E. N.E. N.E.
21 N.E. N.E. N.E. N.E. N.E.
22 554 N.E. N.E. N.E. N.E.
23 N.E. N.E. N.E. N.E. N.E.
24
Conce
pt D
esig
n C
518 N.E. N.E. N.E. N.E.
178
‘N.E.’ denotes ‘Not Exceeded’ during FED calculation period. Table 5-1 may be
referred for the definitions of fire scenarios.
Table 5-3: Time to Exceed Tenability Limits in Lift Shaft and LMR
Time to Exceed (second)
Asphyxiant Gases Heat
Building
Fire
Scenario Smoke
Obscuration
Limit Tenability
Limit
Lethal
Dose
Tenability
Limit
Lethal
Exposure
Temperature
Limit in
LMR
1 317 1140 1500 1140 1260 2473
2 310 1020 1500 1080 1200 2485
3 327 1020 1560 1140 1320 1290
4 306 900 1440 900 1140 1022
5 303 840 1140 1080 1680 854
6
Conce
pt D
esig
n A
306 780 1080 780 1380 522
7 N.E. N.E. N.E. N.E. N.E. N.E.
8 N.E. N.E. N.E. N.E. N.E. N.E.
9 N.E. N.E. N.E. N.E. N.E. N.E.
10 475 N.E. N.E. N.E. N.E. N.E.
11 N.E. N.E. N.E. N.E. N.E. N.E.
12
Conce
pt D
esig
n B
493 N.E. N.E. N.E. N.E. N.E.
13 N.E. N.E. N.E. N.E. N.E. N.E.
14 N.E. N.E. N.E. N.E. N.E. N.E.
15 N.E. N.E. N.E. N.E. N.E. N.E.
16 N.E. N.E. N.E. N.E. N.E. N.E.
17 N.E. N.E. N.E. N.E. N.E. N.E.
18 Conce
pt D
esig
n B
(25%
)
N.E. N.E. N.E. N.E. N.E. N.E.
19 N.E. N.E. N.E. N.E. N.E. N.E.
20 N.E. N.E. N.E. N.E. N.E. N.E.
21 N.E. N.E. N.E. N.E. N.E. N.E.
22 N.E. N.E. N.E. N.E. N.E. N.E.
23 N.E. N.E. N.E. N.E. N.E. N.E.
24
Conce
pt D
esig
n C
N.E. N.E. N.E. N.E. N.E. N.E.
Time relating to LMR is used for determining the reliability of lift operational system
in the next chapter.
179
Table 5-4 gives the times to unsafe conditions in the stair shaft.
Table 5-4: Time to Exceed Tenability Limits in Stair Shaft
Time to Exceed Tenability Limits (second) Building Fire
Scenario Smoke Asphyxiant Gases Heat
1 or 7 554 N.E. N.E.
2 or 8 554 N.E. N.E.
3 or 9 532 N.E. N.E.
4 or 10 428 N.E. N.E.
5 or 11 530 N.E. N.E.
6 or 12
100%
popula
tion
497 N.E. N.E.
13 N.E. N.E. N.E.
14 558 N.E. N.E.
15 N.E. N.E. N.E.
16 508 N.E. N.E.
17 558 N.E. N.E.
18
75%
popula
tion
461 N.E. N.E.
Note: Fire Scenarios 1 and 7 are the same as they contain the protected stair lobby
and are considered for the entire population. (‘N.E.’ denotes Not Exceeded)
Summary of Results
• Tenability limits for smoke visibility, asphyxiant gases and heat are exceeded
in the unprotected lift lobby.
• Tenability limits for smoke visibility and asphyxiant gases are exceeded in
the protected lift lobby.
• Tenability limit for smoke visibility are exceeded in the protected lift lobby
with partial evacuation.
• Tenability limit for smoke visibility are exceeded in the double protected lift
lobby under the influence of wind.
• Tenability limit for smoke visibility are exceeded in stairs. However, smoke
visibility was improved during the limited number of occupants using stairs
(75% of the building population).
180
• Lift door is used by the entire population whereas stair door is used by half
the number of evacuees on the floor (as evacuees were evenly distributed at
two stairs). Therefore smoke spreads quickly in the lift lobby.
• Smoke logged in the protected lift lobby takes considerable time to dilute
(being in an isolated place with minor gap opening in lift shaft).
• Two level compartmentalization strategies in the double protected lift lobby
provide a buffer for smoke and hot gases leakage to lift lobby from the fire-
affected unit.
Analysis of Untenable Conditions: In the unprotected lift lobby, extinction
coefficient exceeds the tenability limit for visual obscuration and sensory irritation
during the fifth minute. The FED limit exceeds the combined effects of temperature
and heat during the period of eighth to twelfth minute. The FED limit exceeds the
combined effects of asphyxiant gases during the tenth minute. In protected lift
lobby, extinction coefficient exceeds the tenability limit for visual obscuration and
sensory irritation during the period of sixth and seventh minutes. The FED limit
exceeds the combined effects of temperature and heat under the influence of wind
during the period of twelfth to fourteen minutes. In the protected lift lobby for 25%
of the population, extinction coefficient exceeds the tenability limit for visual
obscuration and sensory irritation during the period of eighth and ninth minutes. In
the double protected lift lobby, extinction coefficient exceeds the tenability limit for
visual obscuration and sensory irritation under the influence of wind during the ninth
minute. In stairs, the extinction coefficient exceeds the tenability limit for visual
obscuration and sensory irritation during the period of eighth to tenth minutes.
Aged and disabled persons have a ‘high risk’ level than others (Miller, 2005), and
thus consideration needs to be taken into account. Therefore, the effects of FED will
be more in the ‘high risk’ group. Aged and disabled persons (16% of the population)
are exposed to fire effluents in the unprotected lift lobby (asphyxiant gases and
temperature/ heat) and protected lift lobby (asphyxiant gases). Aged and disabled
persons are not exposed to fire effluents in the protected lift lobby for 25% of the
population, double protected lift lobby and stairs.
181
5.8 Determination of Safety Index
The probability of time period for occupants’ evacuation must be greater than the
probability of time to exceed the tenability limits. Random variables were
determined from the stochastic evacuation modelling and fire hazard modelling. The
safety index β is used to determine the safety margin between the lift evacuation time
and onset of hazardous condition.
5.8.1 Strength Variables ASET
The mean and standard deviation of the strength variables (ASET) are obtained from
Tables 5-2 to 5-4. The period for which the tenability limits are not exceeded is the
safe period for evacuation, and the entire population will evacuate the building. The
means and standard deviations of the ASET for the locations of lift lobby, lift shaft
and stairs are given in Tables 5-5 to 5-7. These values are determined from the time
to exceed tenability limits for asphyxiant toxic gases or heat, whichever is reached
first. The ASET, for which tenable limits are not exceeded, is considered for the
FED calculation period (2400 seconds).
Table 5-5: Means and Standard Deviations of ASET (lift lobby)
Concept Design/
Evacuation strategy
Evacuation
Strategy Mean
(µs) Standard
Deviation (σs)
Concept Design A 100% population 610 24.5
100% population 1990 128 Concept Design B
25% population > 2400 -
Concept Design C 100% population > 2400 -
Table 5-6: Means and Standard Deviations of ASET (lift shaft)
Concept Design/
Evacuation strategy
Evacuation
Strategy
Mean
(µs) Standard
Deviation (σs)
Concept Design A 100% population 950 134
100% population > 2400 - Concept Design B
25% population > 2400 -
Concept Design C 100% population > 2400 -
182
Table 5-7: Means and Standard Deviations of ASET (stairs)
Concept Design/
Evacuation strategy
Evacuation
Strategy
Mean
(µs) Standard
Deviation (σs)
100% population > 2400 - Stairs
75% population > 2400 -
5.8.2 Load Variables RSET
The load variables RSET were obtained from the stochastic evacuation modelling
and results are reproduced from Chapter 4 (see Table 5-8). The time values are given
to the nearest second.
Table 5-8: Means and Standard Deviation of RSET
Output Variables Evacuation
Strategy Mean
(µL)
Standard
Deviation (σL)
100% population 1347 30 Lift Pre-Evacuation Time
(tLPM) – RSET in lift lobby
25% population 924 24
100% population 2288 490 Lift Evacuation Time
(tLE) – RSET in lift shaft 25% population 1624 303
100% population 1550 238 Stair Evacuation Time
(tSE)
75% population 1188 181
183
5.8.3 Safety Index
The margin of mean and standard deviation are calculated from Eq. 5.9 and the
safety indices are determined from Eq. 5.10 (see Table 5-9). The system is functional
if the margin is positive i.e. the strength is higher than the load. Where tenable limits
are not exceeded, safety indices are based on the maximum FED calculation period
(2400 seconds). The difference between ASET and RSET is more in the lift lobby
than in the lift shaft. Therefore, average safety indices are calculated to avoid
extreme values.
Table 5-9: Safety Indices for Lift and Stair Evacuation
Conceptual design
and evacuation
strategy
Location Mean
(µM) Standard
Deviation (σM)
Safety
Index
(M
M
σµ
β = )
Average
Safety
Index
(β)
Lift lobby -737 38.73 -19.02 Concept Design A
Lift shaft -1338 508 -2.63
-10.82
Lift lobby 643 131.46 4.89 Concept Design B
Lift shaft > 112 490 > 0.22
> 2.55
Lift lobby > 1476 24 > 61.5 Concept Design B (25% population)
Lift shaft > 776 303 > 2.5
> 32.0
Lift lobby > 1053 30 > 35 Concept Design C
Lift shaft > 112 490 > 0.22
> 17.61
Stair (100% population) > 850 238 > 3.57 > 3.57
Stair (75% population) > 1212 181.8 > 6.66 > 6.66
5.8.4 Safety Index Analysis
The safety index β was determined for the concept designs and evacuation strategies.
The safety index is positive in Concept Designs B and C whereas negative in
Concept Design A (the system is functional – the strength is higher than the load).
Protected lift lobby for one-fourth of the building population, double protected lift
lobby and stairs provided a life safety index of more than 3 to the evacuees. The
safety index is the maximum for partial building lift evacuation. Protected lift lobby
for the entire population had lower safety indices when compared to that of stairs.
Unsafe conditions arrive quickly in the unprotected lift lobby.
184
5.9 Conclusion
The safety index concept is based on probabilistic theory. On the other hand, the
computational fluid dynamics simulation using FDS is a deterministic approach. The
FDS predicted the time when the lift lobbies and stairs became unsafe under variable
conditions. The unsafe conditions within a hypothetical building were determined
with regard to temperature of the hot layer, concentration of toxic gases and
visibility. Two variables (vertical location and wind speed) were considered to
determine the impact on the three concept design options with additional evacuation
strategies (unprotected lift lobby, protected lift lobby for the entire population and
one-fourth population and double protected lift lobby)
Stack effect is significant in the unprotected lift lobby, but not in the protected lift
lobby for the spread of smoke and hot gases. An exponential increase in wind speed
with height demonstrated that the time to exceed tenability limits at upper levels
comparatively influenced more (compared to lower floors). Wind caused visibility
obscuration in protected and double protected lift lobbies. However, visibility could
be maintained for longer periods if the occupants use the doors less frequently (for
example, by 25% of the population). The temperature was significantly increased in
the unprotected LMR in the unprotected lobby scenarios due to stack effect and
wind, but was not much affected in protected lift lobby scenarios.
The load variables (RSET) and the strength variables (ASET) were used to calculate
the safety index. Unprotected lift lobby provided a negative safety index and exposed
to the evacuees to the maximum risks. Protected lift lobby, double protected lift
lobby and stairs provided a positive safety index to the evacuees. Time to exceed
tenability limits with regard to temperature and toxic gases did not arrive in the
protected lift lobby for one-fourth of the building population and in the double
protected lift lobby. The FED calculations showed that the times to exceed tenability
limits for visibility in lift lobbies were not significantly influenced by the protected
lift lobbies.
185
6. RELIABILITY OF LIFT OPERATIONAL MECHANISM
6.1 Introduction
During fire emergencies, many components may affect the reliability of lift
operational mechanism. While using lifts for emergency evacuation, the occupants
may inadvertently trapped in the lifts and be exposed to heat and smoke on the fire-
affected floor. The main issues that affect the reliability of lift systems are
temperature sensitive electronic components, electric power supply failure and close
proximity of water based fire fighting systems.
Lift operational mechanism makes use of temperature sensitive electronic devices.
The lift system can be affected by the malfunctioning of electronic devices due to
excessive temperature rise in fires. The electric power supply in the developed
countries is generally characterized as stable and reliable. However, this may not be
true for electric power supply in other countries of the world. The challenge of
providing a reliable electric power supply is made even more difficult in the presence
of fires, which makes electric systems more susceptible to power failures and shut
down the lift operation (Klote, 1982). Available statistics have shown that electrical
fault is one of the prevalent causes of fires in the buildings (NFPA, 2006). In fire
incidences that are initiated by electrical appliances, the electric power supply to lifts
will be a concern. The lift system may also be affected by water spread from fire
fighting system as it can damage the electrical and/or mechanical components of the
lift operating system (Klote, 1982).
This chapter addresses the issues related to reliability of lift operational mechanism
based on probabilistic risk assessment techniques. The techniques used include fault
tree and event tree analyses. The objectives of this chapter are to:
1. To determine the reliability of lift operational mechanism.
2. To determine the feasibility of reliability improvement.
186
6.2 Analysis of Lift Operational Mechanism
To determine the reliability of lift operational mechanism, the issues relating to
malfunctioning of lifts, electric power failure and water damage are analysed. The
objectives of analysis are to:
1. determine the probabilities of malfunctioning of lifts, electric power failure
and water damage; and
2. analyse the reliability of lift operational mechanism that can raise concerns of
human behavioural response (panic).
The reliability of lift operational mechanism is analyzed based on the following
assumptions:
• Water based fire extinguishment systems are not used on fire that occurred in
electrical systems.
• Apartment buildings are provided with secondary (alternate) sources of
power supply in the form of electric generators.
• Water based fire protective and fire fighting measures (for example, fire hose
reel, automatic sprinkler and hydrant system) are present. Fire extinguisher
constitutes only a tiny fraction of water spread and hence fire extinguishment
using fire extinguishers is not included in the water spread analysis. Manual
extinguishing facility is interpreted as fire hose reel.
For the high probability of operating fire protective and fire fighting measures, the
probability of impact from high risk is low (see Chapter 3, Section 3.1.5). However,
water based fire protection systems such as fire hose reel and sprinkler system were
not considered in Chapter 5 for evaluating the impact of high risk.
187
6.3 Methodology
Lift Malfunctioning due to Excessive Temperature Rise: @RISK software is used
to determine the probability of lift malfunctioning due to excessive temperature (see
Chapter 2, Section 2.10). The probability of temperature attainment in lift machine
room (LMR) can be given as:
( ) { }TTProbTF LMR <= 6.1
and TLMR is the temperature in LMR. F(T) is the probability distribution function and
it is the probability that the lift motor room temperature will not exceed any given
value T. Temperatures in LMR were determined from FDS modelling for Fire
Scenarios 1 to 6. Temperatures did not exceed the safe limit (43°C) in LMR for Fire
Scenarios 7 to 24. Therefore Fire Scenarios 7 to 24 are not considered for lift
malfunctioning (see Chapter 5, Table 5-3). The averages of the temperatures during
the simulation period for Fire Scenarios 1 to 6 are used to determine the probability
of lift malfunctioning.
Electric Power Failure: Fault tree analysis is used to determine the probability of
electric power failure. The impact of fire is considered on electrical system. A
typical electrical system of essential and non-essential electric supplies is illustrated
with its interrelationship with primary and secondary sources of power supply.
Boundary conditions are established during a fire scenario. Boundary conditions are
the physical boundaries of the system (i.e. which parts of the system are included in
the analysis and which parts are not?), the initial conditions (i.e. what is the
operational state of the system when the top event is occurring?) and external stresses
(impact from external events). Fault tree is constructed and outcomes are analysed
qualitatively and quantitatively. The effect of temperature in LMR is included in the
analysis.
Water Damage: Probabilistic analysis is conducted to determine the quantity of
water spread from various fire fighting measures. However, building evacuation may
or may not be required at the stage of fire extinguishment. The water spread is time
188
variant and can be described as a complex parallel and series system of water based
fire protective and fire fighting measures. This approach is based on a hybrid
combination of parallel and series system (Modarres and Billoch, 2002). The
probability of water spread can be derived from the individual probability of
occurrence POE of water spread from each fire protection system FPS. The
probability of water spread Pw can be derived as:
{ }∏=
=n
i
iiW POEFPSP1
)( 6.2
The water based fire protection systems are assigned a number from 1 to n. Each fire
protection system is assigned a probability of occurrence in order to arrive at a
probability of water spread. The estimated total quantity of water generated Qt from
the water based fire protection measures can be derived as follows:
∑=
=n
i
it tfQQ1
)( 6.3
>
<=
ttforQ
ttfortf
ii
i0)( 6.4
where
Qi is the water discharge of the individual fire protection system
t is the time of the fire protection system i to activate
ti is the total time of the fire protection system i to activate
The fire is extinguished with the fire protection and fire fighting measures and one of
the fire fighting system may be missing (or non-operational) during actual
emergencies. The scenario is considered for minimum and maximum water spread
cases. The complex parallel-series system is used in the analysis. The system is
divided into basic parallel and series modules and then the probability function for
each module is determined separately. The analysis determines the probability of
water spread and quantity of water spread on the floor.
189
6.4 Lift Malfunctioning due to Excessive Temperature Rise
Solid state electronic devices are used in the lift control systems and operate under
variable loads and demands. These devices are subjected to a regular stress during
the normal operating conditions. Internal heat is released from the losses in
machinery such as lift motor, control cabinet transformer, converter, invertors and
power supplies. External heat from ambient conditions and other electrical devices
such as lights is also added to the stress level. These stresses may reduce the
reliability of lift systems. Therefore, lift manufacturers have recommended a
temperature limit of 32°C for LMR (this value may vary from manufacturer to
manufacturer). This value is chosen by the lift manufacturers to ensure that the actual
temperature in the controller cabinets, which is typically 10°C to 15°C higher than
the ambient room temperature, is not above the design operating limits of the solid
state devices (Marchitto, 1991). Under the recommended limit, there is no effect on
the functionality of the lift system.
The Australian Standard AS 1735.1 (2003) specifies an ambient temperature not
more than 43°C in LMR. This is the maximum temperature under which electronic
devices can operate without a safety margin. Degraded performance and long term
reduced reliability may result between the recommended and specified temperatures
(see Table 6-1). Beyond the specified rating of electronic devices, the performance
is not assured. Electronic devices may malfunction under high stresses and may
recover during the reduced load conditions. Electronic devices for lift systems can be
commercial, industrial or military grades. The design temperature for commercial
grade is 70°C, for industrial grade is 85°C and for military grade is 125°C
(Robibero, 1991). Above the design temperature, electronic devices may not recover
and the lift system may go to a permanent failure mode (see Table 6-1).
190
Table 6-1: Temperatures and their Impact on Lift Systems
Temperature Impact
Less than 32°C No impact
Between 32°C and 43°C Long term reduced reliability
Between 43°C and 70°C Lift malfunctioning
More than 70°C Failure
The reliability of lift systems can be adjudged from the thermal environment that lift
devices may be subjected to. The field model (FDS) results indicated that lift
machine rooms via unprotected lift lobbies are susceptible to high temperatures.
However, protected lift lobbies provide adequate safety from high temperatures
occurring in LMR due to fires. The probability of lift malfunctioning depends on the
temperature rise beyond the temperature threshold of the electronic components (i.e.
43°C).
The computer package @RISK (Palisade Corp, 1996) is used to determine the
probability of lift malfunctioning. The cumulative graphs show the temperature to be
more than 43°C or 70°C in Fire Scenarios 1 to 6 (see Figure 6-1). They show the
probability (percentage) of malfunctioning due to excessive temperature. By
dragging the delimiters displayed on the cumulative graph, the probability
(percentage) of exceeding a temperature is calculated (for example, it is 100 – 88.4 =
11.6 for 43°C in Fire Scenario 1). The cumulative graphs indicate that the
temperature beyond the threshold value is a function of the vertical location of a fire.
In the cumulative graph, blue line indicates the actual values of temperature, whereas
red line shows the imaginary best fitting curve for determining the percentage of
malfunctioning due to excessive temperature.
191
Fire Scenario 1
43.00
88.4%20.00
5.0%
0
0.2
0.4
0.6
0.8
1
15 20 25 30 35 40 45 50 55 60
Temperature (C)
F(T)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
Fire Scenario 2
20.00
5.0%
43.00
88.2%
0
0.2
0.4
0.6
0.8
1
15 20 25 30 35 40 45 50 55 60
Temperature (C)
F(T)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
Fire Scenario 3
20.0
5.0%
70.0
84.2%
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100 120 140
Temperature (C)
F(T)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
192
Fire Scenario 420.00
5.0%
43.00
92.3%
0
0.2
0.4
0.6
0.8
1
15 20 25 30 35 40 45 50 55 60
Temperature (C)
F(T)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
Fire Scenario 5
20.6
5.0%
70.0
94.2%
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100 120 140 160
Temperature (C)
F(T)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
Fire Scenario 6
20.6
5.0%70.0
93.7%
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100 120 140
Temperature (C)
F(T)
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
@RISK Student Version
For Academic Use Only
Figure 6-1 – Probability of Excess Temperature Rise in LMR (Fire Scenarios 1 to 6)
193
The probability (percentage) of exceeding a temperature is used to calculate the
probability of reduced reliability or malfunctioning or failure (see Table 6-2). The
average probability (percentage) of exceeding the recommended temperature of 32°C
or more is 36.26% (0.36) for Fire Scenarios 1 to 6 whereas that for the specified
temperature of 43°C or more is 19.96% (0.2), and the designed temperature of 70°C
or more is 4.65% (0.04), for the lift machine room located on the top of unprotected
lift shaft.
Table 6-2: Probability of Excess Temperature Occurrence in LMR in Fire Scenarios
Probability (%) of excess temperature Fire Scenario
More than recommended
temperature 32°C
More than specified
temperature 43°C
More than designed
temperature 70°C
Fire Scenario 1 24.9 11.6 -
Fire Scenario 2 25.3 11.8 -
Fire Scenario 3 40.8 33.1 15.8
Fire Scenario 4 19.8 7.7 -
Fire Scenario 5 40 20.6 5.8
Fire Scenario 6 66.8 35 6.3
Average 36.26 19.96 4.6
The probability of lift malfunctioning (including failure) is conservatively considered
to be for temperatures more than 43°C. The average probability of reduced
reliability between temperature 32°C and 43°C is calculated as 0.16 (36% – 20% =
16%). Similarly, probability of malfunctioning between temperature 43°C and 70°C
is calculated as 0.16 (20% – 4% = 16%). The probability of complete failure, above
temperature 70°C, is 0.04 (4%). The probability of lift malfunctioning, above
temperature 43°C, is therefore equal to 0.2 (see Figure 6-2).
194
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 20 40 60 80 100 120 140
Temperature (C)
Probability
32 7043
Complete
failure
(0.04)
Reduced
reliability
(0.16) Malfunction
(0.16)
No effect
(0.64)
Figure 6-2 – Probability Distribution and Consequences of Excess Temperature in
LMR
With the failure probability of 0.2, on average 41% of the building population may
not be able to use the alternative evacuation facility i.e. lifts (see Table 6-3). A
portion of this population may also be trapped in the lift cabins in the unprotected lift
shaft.
Table 6-3: Unavailability of Lifts to Building Population
Fire scenario Probability of excess temperature rise
(43°C)
Time to exceed limit for specified
temperature (second)
% of population remaining in the
building
Fire Scenario 1 0.11 2473 0
Fire Scenario 2 0.11 2485 0
Fire Scenario 3 0.33 1290 36
Fire Scenario 4 0.07 1022 54
Fire Scenario 5 0.2 854 67
Fire Scenario 6 0.35 522 88
Average 0.2 1441 40.8
Probability of lift malfunctioning increases as the fire is on the upper portion of the
building. A significant number of persons may face the problem of unavailability of
lifts (although remaining persons would have the option of using stairs).
195
6.5 Electric Power Failure
There are no general statistics on power failures in high rise building fires. With the
help of fault tree analysis, the probability of electrical power failure is determined
during the occurrence of a fire.
6.5.1 System Descriptions
The common provision is made for two sources of electric power supplies to a
common bus bar as shown in Figure 6-3.
Figure 6-3 – Typical Electrical System for Essential and Non-Essential Supplies
Hot gases
Non-essential electric
supply
Essential electric
supply
To LMR
Local fuse
To SOU
Circuit breaker
Circuit breaker
Circuit breaker
Transformer-2 (Secondary supply –
generator)
Transformer-1
Bush bar
Manual trip
by fire fighter
(Primary supply
– city mains)
Electric Fire
196
Electrical supply to the essential services is independent and uninterrupted at all time
even if the entire electrical supply in the premises is switched off. Essential services
include fire fighting pumps, smoke control system, emergency lighting and lifts
whereas non-essential services include apartment units. Secondary source of power
supply (generator set) is required to ensure the continued operation of essential
equipment during building evacuation.
6.5.2 System Boundary Conditions
Physical boundary of the system is defined by incorporating the primary and
secondary sources of power supplies, distribution to essential and non-essential
power supplies during the occurrence of a fire (see Figure 6-3). The initial condition
is assumed to be the occurrence of a fire in a sole occupancy unit (SOU) and then
spread to common area. The fire can result from an electrical short-circuit. Table 6-4
indicates the initial conditions and top event. Boundary conditions with respect to
external stresses (fire and hot gases) are included in the analysis.
Table 6-4: System Boundary Conditions
Condition Initial Conditions External Stresses
Condition 1 Electrical short circuit within the compartment
of fire origin
Nil
Condition 2 Fire spread beyond the origin of fire – Impact
on unprotected lift lobby
Lift malfunctioning due
to excessive temperature
Condition 3 Fire spread beyond the origin of fire – Impact
on protected lift lobby
Nil
6.5.3 Data and Statistics
The following data/ statistics were used in the fault tree analysis:
• NPFA recorded the causes of the fires in apartment buildings, which shows
that 10% of them are due to electrical fires. Hence 0.1 is considered as the
probability of electrical fire in apartment buildings.
• The failure rate of electrical surge is assigned 0.04 /D (WASH 1400, 1974).
This failure rate is an occurrence in which the circuit breaker ceases to
perform the required function on demand (D). Electrical surges on the mains
197
supply can come from a lightning or heavy current load through inductive
effects. Lightning may also induce currents into signal cables entering a
building. Lift motors can generate a significant number of surges.
• The failure rate of circuit breaker is assigned 1 × 10-3 /D (WASH 1400,
1974). This failure rate is an occurrence in which the circuit breaker ceases
to perform the required function on demand.
• The failure rate of electric fuse is assigned 1 × 10-5 /D (WASH 1400, 1974).
This failure rate is an occurrence in which the electric fuse ceases to perform
the required function on demand.
• The probability of fire spread beyond the origin of fire (fire compartment) is
0.06 (see Chapter 2, Table 2-2).
• For electric power failure, urban Australian locations are subjected to
approximately three outrages of 10 minutes duration per annum, i.e., a failure
rate of 5.70 × 10-5 per annum (Lacey, 2000). Failure rate is the frequency of
power outrage expressed in failure per annum and calculated as 0.5 hour / (24
× 365). The failure is based on day-to-day power supply.
• The failure rate of secondary power supply (generator set) is assigned 0.03 /D
(WASH 1400, 1974). This failure rate is an occurrence in which the
generator ceases to perform the required function on demand.
• The error rate of manual tripping by operator or fire fighter is assigned 0.003.
Operator or fire fighter may shut down the entire electrical system (including
essential supply) by selecting the wrong control in a group of identical and/or
labeled controls (NUREG/CR-1278, 1980).
• Teo (2001) highlighted that the lift breakdown rate is once in three months in
apartment buildings over a period of 10 years. Taking an average breakdown
for a period of 4 hours, a probability of failure is calculated as 1.82 × 10-3 per
annum. The breakdown rate is the frequency of lift maintenance during which
lifts are not available, expressed as breakdowns per annum and calculated as
4.0 hour × 4 breakdown services / (24 × 365).
The probability of lift malfunctioning due to excessive temperature is 0.2 for lifts
with unprotected lift lobby (see Section 6.4). The probability of lift malfunctioning is
considered nil for lifts with protected lobby.
198
6.5.4 Fault Tree Analysis
A fault tree is a “top-down” method of analysing the susceptibility of a product to
failure. Fault tree analysis contains the following modules i.e. Basic Event,
Undeveloped Event, ‘AND’ Gate and ‘OR’ Gate and Fault Event. Basic Event
contains a failure at the lowest level which has the capability of causing a fault to
occur. Undeveloped Event contains a failure at the lowest level which is not fully
developed due to lack of information. The ‘OR’ gate indicates that the output event
occurs if any of the input events occur. The ‘AND’ gate indicates that the output
event occurs only if all the input events occur at the same time. Fault Event contains
description of a lower level fault. External Event is the occurrence of fault externally.
Condition 1: Fault tree analysis to estimate the probability of electric failure due to
fire in electrical origin in an apartment is shown in Figure 6-4. The data used in the
fault tree analysis is obtained from section 6.6.3. The power is fed from non-essential
services, which is not connected to the lift system. This has little effect on lift electric
power supply.
Figure 6-4 – Fault Tree Analysis for Electric Fire in SOU
Excessive current in main circuit
“OR” Gate (add)
1.0E-6
Legend
“AND” Gate (m) Excessive current in local circuit
Local fuse fails to open
Primary wiring failure (short circuit)
0.104
Primary power supply failure (surge)
0.104
1.0E-05
1.0E-6
Circuit breaker fails to open
Main electric supply failure
1.0E-09
Basic Event
0.1 0.04
Fault Event/
Top Event
1.0E-03
Electric power failure in non-essential system
(though primary power supply is available)
1.0E-09 1.0 × 10-9
199
Condition 2: Fault tree analysis to estimate the probability of electric power failure
in unprotected lift lobby is shown in Figure 6-5. The probability of lift electric power
supply failure is determined from the impact of fire in a common area (beyond the
origin of fire – SOU), primary and secondary sources of electric power failure,
malfunctioning of lift due to excessive temperature, lift maintenance breakdown and
manual tripping. The probability of lift electric power failure in unprotected lift
lobby is 0.2648.
Figure 6-5 – Fault Tree Analysis for Electric Power Failure in Unprotected Lift
Lobby
Malfunctioning of lifts due to hot
gases
Secondary power supply failure (Generator)
Primary power
supply failure
Lift power supply failure
0.2
0.03
0.003
Essential power supply
failure Manual
tripping
5.7E-5
1.71E-6
1.71E-6
0.2648
Lift maintenance breakdown
Electric wire short circuit due to fire
or hot gases
1.82E-3
“OR” Gate
Legend
“AND” Gate
Undeveloped Event
Basic Event
Fault Event/
Top Event
External Event
0.06
200
Condition 3: Fault tree analysis to estimate the probability of electric power failure
in protected lift lobby is shown in Figure 6-6. The probability of lift electric power
supply failure is determined from the primary and secondary sources of electric
power failure, lift maintenance breakdown and manual tripping. The probability of
lift system failure due to electricity is 0.0048, which is during the reasonably worst
circumstances (manual breakdown and manual tripping).
Figure 6-6 – Fault Tree Analysis for Electric Power Failure in Protected Lift Lobby
Secondary power supply failure (Generator)
Primary power
supply failure
Lift power
supply failure
0.03
0.003
Essential power supply
failure Manual
tripping
5.7E-5
1.71E-6
1.71E-6
4.82E-03
Lift maintenance breakdown
1.82E-3
“OR” Gate
Legend
“AND” Gate
Undeveloped Event
Basic Event
Fault Event/
Top Event
201
6.5.5 Analysis of Results
Table 6-5 indicates the probability of lift failure in an event of fire.
Table 6-5: Impact on Lift System
Condition Initial Scenario Probability Impact on Lift System/
Reasons
Condition 1 Electrical short circuit within the compartment of fire origin
1.0 × 10-9 No impact on lift electric system; The probability is related to the failure of non-essential services.
Condition 2 Fire spread beyond the origin of fire – Impact on unprotected lift lobby
0.2648 High impact is on essential services; Hot gases in LMR are the main reasons for high probability of failure.
Condition 3 Fire spread beyond the origin of fire – Impact on protected lift lobby
0.0048 A little impact on lift electric system; The probability is related to the failure of essential services.
For condition 1, fire of electric origin in SOU does not contribute to the failure of lift
electric power system. Figure 6-3 illustrates that electric fire in SOU is confined to
the non-essential electric services only (not connected to the essential services). The
derived probability (1.0 × 10-9) is related to the failure of non-essential services.
For condition 2, fire spread from SOU to a common area (including unprotected lift
lobby) contributes significantly to lift failure. External stress from lift malfunctioning
due to excessive temperature in the building is the main reason for electrical failure.
The condition is assumed under which (a) fire spread beyond the origin of fire in
common area, (b) smoke and hot gases spread in unprotected lift lobbies, (c) manual
tripping of electrical essential services, and (d) failure of electric supply system. If
the wiring is laid in the common area, the damage from fire or hot gases may occur
due to direct exposure of wire. However, FDS results have demonstrated that the
maximum temperature during the period of fire simulation is less than 300°C in
unprotected lift shaft, which may not be capable of causing electric short circuit
(exposing/ melting of PVC wire) during a short time.
202
For condition 3, protected lift lobby provides a higher protection against the lift
failure. The probability of electrical failure is only 0.0048.
The probability of a fire outbreak and incoming electrical power failure is relatively
small, provided that the fire and electrical power failure do not have the same cause
in the building. However, the probability of electrical power failure during a fire is
relatively high in unprotected lift lobbies since the fire frequently damages electrical
systems in the building.
6.6 Probabilistic Analysis of Water Damage
The water requirement for fire fighting mainly depends on the size of fire. The water
based fire protective system includes fire hose reels (FPS1), sprinklers (FPS2) and
fire hydrants (FPS3) in the apartment buildings. Sprinklers cover the risk area
whereas fire hose reels and fire hydrants are generally positioned near the exit routes
such as locations near stairs or lift lobbies. Professional fire fighting operation is
generally conducted at a later stage, when fire brigade is called automatically or
verbally.
Fire suppression technology employs the life cycle of fire progress in different forms.
The incipient stage where heat is low to moderate and there is not much visible
smoke or flame but there is the threat of fire propagation. At this stage, the occupants
may undergo coping stage and they would like to use some handy form of fire
extinguishment such as fire hose reel (FHR). The water spread is minimal. If the fire
is not controlled and visible smoke or heat threatens the life, occupants have no
option but to leave the premises. At this stage or later, sprinkler heads will actuate to
control or extinguish the flame. The water spread is moderate. As the fire further
progresses and is not extinguished with sprinklers, fire-fighters intervene to
extinguish the flame. The water spread is the maximum since the quantity of heat, at
this stage, is at its peak and therefore fire fighters are using nozzles of bigger sizes. A
large quantity of water is accumulated by this time on the floor. With the application
of water, fire may be uncontrolled, controlled or suppressed, which may or may not
require building evacuation. Figure 6-7 shows the heat release rate (HRR) with time
203
during the three stages of water application. This graph is a modified form of graph
indicated in FCRC (1996) for HRR during sprinkler operation.
Figure 6-7 – HRR during Three Stages of Water Application
6.6.1 Water Spread from Fire Protection and Fire Fighting Measures
The probabilities of water spread from fire protective and fire fighting measures
determined from statistics are given below:
FPS 1 Fire Hose Reel – Nearly 10% of fires are extinguished with the help of
building manual extinguishing facility (NZFS incident statistics, 1993-1997).
Uncontrolled fire may include non-operational (non-functional) of manual fire
fighting facility and failure to extinguish the fire with the help of manual fire fighting
facility. The data on non-operational/ non-functional and failure to extinguish with
the help of manual fire fighting facility such as fire hose reel are not available. It is
assumed that the percentage of non-functionality of manual fire fighting facility is
12.4. This is considered at par with sprinkler system (see FPS2 sprinkler). Out of the
remaining 88.6%, 10% is considered as fire extinguishment with the help of fire hose
reel and 77.6% is considered as the failure to extinguish the fire. Therefore, the
probability of water spread due to fire hose reel for extinguishment is assigned as
Manual fire fighting
(FPS1)
Automatic fire
fighting (FPS2) H
RR
Time
FHR uncontrolled
Sprinkler extinguishment
Sprinkler uncontrolled
Sprinkler control
FHR extinguishment
Fire brigade extinguishment
Fire brigade extinguishment
Internal fire fighting
Fire brigade intervention +
others (FPS3)
FHR failure
204
0.1, the probability of water spread for uncontrolled fire is 0.776 and the probability
of non-water spread due to non-functionality is 0.124. Technically, a fire hose reel is
capable of producing 0.55 l/s (AS1221, 1997). However, when applied to fire, some
amount of water is used for extinguishing, some goes to vaporisation and the rest
goes on spreading. Building evacuation is required under uncontrolled fire conditions
arising in the preliminary stage.
The use of FDS model showed that the fire detector in the SOU operated at 90
seconds (see Chapter 5, Section 5.6). So, it is assumed that the occupants would
commence fire fighting operation after travelling to the corridor and fetch fire hose
reel for extinguishment. Fire hose reel is considered as first aid fire fighting
equipment and meant for the use of general public. The fire fighting operation starts
after 120 seconds and the tenability limits are exceeded at 187 seconds in the SOU.
Therefore the occupants presumably continue fire fighting for 67 seconds and are
forced to leave the premises due to untenable environmental conditions.
FPS2 Sprinklers – The NFPA statistics (Rohr, 2001) for the ten year reporting period
from 1989 to 1998 indicates that the operational reliability of automatic sprinkler
systems for apartment buildings is 87.6%. However, most of the sprinklers are
designed to control the fire but not necessarily to extinguish the fire. Statistics for the
same period showed that the percentage of fires, where sprinklers are present and
that are reported as being extinguished by an automatic suppression system, is 20%
in apartment buildings. Automatic suppression system is primarily the sprinkler
system in apartment buildings. Therefore the probability of fire extinguishment using
sprinklers is 0.2 and controlled fire is 0.676 while that of failure is 0.124.
Marryatt (1988) reports an average of 1.22 sprinklers in operation for the 33 recorded
fires in apartment buildings. As a result the fire is kept from spreading with a
minimal amount of water. Generally sprinklers are installed in apartment buildings
according to extra light hazards. Water spread from the sprinkler system depends on
sprinkler water spray rate, water distribution pattern, inter-spacing among the
sprinklers and design of the building. Heat is absorbed by the discharged water from
the sprinklers and a portion of discharged water vaporizes into steam. The excess
water flows and causes the water spread on the floor and subsequently may cause
205
damage to the lift system. The water application rate for apartment building is 2.25
mm/min, the area of sprinkler operation is 84 m2 and water flow rate is 225 l/m for
the duration of 30 min. Hence, the water discharge rate is calculated as 4.57 l/s in
the sprinkler operated area.
FPS3 Fire Hydrant – An analysis of residential fire incidents reveals that 65% of the
fires are extinguished by the fire service. Remaining 35% of the fires were
extinguished by passerby or other methods (Davis, 2000). About 90% of residential
fires are controlled in less than an hour (Beever and Davy, 1999). The emergency
response time of fire brigade is established to a maximum of 10 minutes in order to
arrive at the fire incidents. The Fire Brigade Intervention Model (FBIM) estimates
the time of arrival of the fire service in the enclosure of fire origin (AFAC, 1997).
Residential properties require less than 10 l/s of water to extinguish a fire (Davis,
2000) and typical hydrant flow rate is 30 l/s.
6.6.2 Complex Parallel and Series System for Water Spread
The probabilities arising from the water based fire protective measures and fire
brigade intervention, due to which water can spread, are shown in Figure 6-8. The
quantity of water spread is shown in Figure 6-9. The values are represented
individually against the outcomes.
Three levels of water spread are assumed for causing damage. Low level of water
spread is assumed from fire hose reel. Medium level of water spread is assumed from
sprinkler system alone (or including fire hose reel). High level of water spread is
assumed from fire brigade intervention alone (or including sprinkler system and fire
hose reel). The levels and their combined probabilities of occurrence are given in
Table 6-6. The level of water spread is virtually absent if none of the water based fire
protective or fire fighting measures is functioning.
206
Table 6-6: Probability of Water Spread at Three Levels
Level of Water
Spread
Fire Protective or Fire
Fighting
Combined Probability of Occurrence
Low Fire hose reel 0.1 + 0.034 + 0.005 = 0.139
Medium Sprinkler system alone (or including fire hose reel)
0.025 + 0.155 + 0.029 + 0.184 = 0.393
High Fire brigade intervention alone (or including sprinkler and fire hose reel)
0.01 + 0.34 + 0.055 + 0.063 = 0.468
The combined probability of occurrence for water spread at the low level is 0.139,
the medium level is 0.393 and the high level is 0.468. Water gets moving at a slope
of approximately 1 cm per metre. Water may flow toward the lift shaft during the
medium or high level of water spread (sprinkler or fire brigade intervention) and
damage the lift system. Water spread from fire hose reel may not cause lift damage.
Further, unprotected lift lobbies are more likely to be subjected to water damage as
lift lobbies are directly connected to the public corridor or risk area. Protected lift
lobbies provided with suitable form of compartmentation from risk area and adhering
with norms for water control measures can restrict water spread. However, this is a
design aspect (see Chapter 2, Section 2.7.4).
207
Figure 6-8 – Complex Parallel-Series System for Probability of Water Spread
208
Figure 6-9 – Complex Parallel-Series System for Quantity of Water Spread
209
6.6.3 Water Spread Result Analysis
The probability of water spread due to sprinkler and fire fighting operation is
considerably high. The combined probability of water spread (0.468) includes fire
brigade intervention, during which fire is controlled and extinguished. The maximum
quantity of water spread occurs from fire brigade intervention.
The probability of maximum water spread as a single outcome was 0.34 for the
quantity of water accumulated from fire hose reel, sprinklers and fire brigade
intervention (see Figure 6-8). The maximum quantity of water spread was 28 000 ltrs
approximately (see Figure 6-9). The maximum quantity of water spread is plotted
against the population remaining in the building (see Figure 6-10). The remaining
population was determined from the stochastic evacuation model (see Chapter 4).
0
5000
10000
15000
20000
25000
0100
150
187
200
250
300
350
370
600
900
1200
1500
1800
2100
2400
Time (second)
Water (litre)
0
10
20
30
40
50
60
70
80
90
100
Population (%)
Water Quantity
Population remaining in building
FHR
Phase
Sprinkler Phase Fire Brigade
Intervention
Figure 6-10 – Quantity of Water Spread and Building Evacuation
The water damage may occur if the building is not provided with adequate drainage
facilities. About 80% of the population remains in the building at the time of fire
brigade intervention (see Figure 6-10). By this time, about 1800 litres of water has
already accumulated from the fire hose reel and sprinkler system. In such
circumstances, fire brigade at the scene can envisage if their fire fighting operation is
causing enough water damage to the lift system and assist lift evacuation.
210
6.7 Outcomes
Following outcomes are obtained from the analysis:
• The protected lift shafts provide adequate safety to LMR and do not require
additional provisions of redundancy measures. The unprotected lift shafts do
not provide adequate safety to LMR.
• The protected lift shafts containing protected electric supply do not require
additional provisions of redundancy measures. Dual electrical supply in
protected lift shaft ensures continuity of lift operations. The electrical
installation needs to conform to the relevant codes and standards (see Chapter
2, Section 2.7.4). The unprotected lift shafts do not provide adequate safety to
electric power supply.
• Water spread is a design aspect, which must follow the provisions of relevant
codes and standards (see Chapter 2, Section 2.7.4).
6.8 Influence on Human Behavioural Response
Lifts with unprotected lobby do not provide adequate safety against lift
malfunctioning, electrical power failure and water spread. The probability of lift
malfunctioning is 0.2. A combined probability of lift malfunctioning, electric power
failure and maintenance breakdown is 0.26. During the probability of maximum
water spread (0.43), about 80% of the population are remaining in the building.
Irrational human behavioural response would be the maximum due to unavailability
of evacuation route.
Lifts with protected lobby provide adequate safety against electrical malfunctioning,
electrical failure and water spread. The probability of lift malfunctioning due to
excessive temperature is little as there is no rise in LMR temperature due to
compartmentation. The probability of electrical power failure is also reduced to a
great extent (0.0048). Water spread can be restricted by adhering the norms such as
compartmentation, floor slope and drainage facilities. The reliability of lift
operational mechanism has increased considerably and irrational human behavioural
response would be the minimal.
211
6.9 Conclusion
This chapter has described an investigation into the reliability of lift operational
mechanism. The results demonstrated that the lifts protected with lobby are
significantly more reliable in comparison to the lifts without lobby. The spread of
hot gases to the LMR via unprotected lift lobby increased the probability of lift
malfunctioning. In unprotected lift lobbies, temperature rise in LMR depends upon
the location of fire. Protected lift lobbies provided adequate safety against the
temperature rise in LMR. The probability of electric power failure due to fire in a
residential unit was found to be small. Electrical system adhering to relevant codes
and standards in protected lift lobbies provides adequate safety for electrical
installation.
Another cause of lift malfunctioning is the water used in fire fighting. A complex
event tree analysis was conducted for determining the quantity and probability of
water spread from fire protective and fire fighting measures. It was found that a
copious amount of water generated from fire fighting measures could damage the lift
components. The maximum quantity of water spread is predicted to occur from fire
brigade intervention. The probability of maximum water spread was higher with all
fire fighting measures. Unprotected lift lobbies are more susceptible to water damage
in the absence of any protection measures (as getting water flow in the absence of
barrier). Protected lift lobbies with protection measures (for example floor slope and
water drainage facility) could provide a suitable barrier to water spread.
The reliability of lift operational mechanism contributes to the satisfactory use of lift
evacuation system. Evacuees’ irrational behavioural can not be ruled out in
unprotected lift lobbies. The probability of evacuees’ irrational behavioural have
considerably reduced in protected lift lobbies. The reliability is used in the
calculation of overall risks of using lift evacuation systems in the next chapter.
212
7. RISK ASSESSMENT OF EVACUATION ROUTES
7.1 Introduction
Fire safety strategies are often developed based on uncertain conditions and sparse
data (Watts, 1995). Modelling fire risk is extremely complex process and involves a
network of interacting components. Analysing the risks in lift evacuation routes
require systematic and practicable approaches. The risks in lift evacuation system
range from asphyxiant toxic gases, movement of smoke and hot gases in the
evacuation routes, behaviour of people in fire condition to reliability of lift
operations. The consequences can vary from psychological to physiological impact
with known (or unknown) probabilities. Therefore, an integrated risk assessment
methodology is required for a consequence based analysis. The consequence based
analysis gives the overall probability of psychological and physiological impacts on
evacuees arising in the evacuation routes.
The ‘Multi-Objectives Decision Analysis’ (MODA) is a consequence based analysis
method and provides an evaluation of influencing, dependent and interacting issues.
The MODA does not provide solutions, but is rather an information source,
providing insight into the situation, uncertainty, objectives and trade offs (Pfeiffer,
1997). The MODA is defined as an approach to decision making under conditions of
complexity, with inherent uncertainty, multiple objectives and different perspectives
towards the decision problem (Clemen, 1996). It is a systematic procedure for
transforming opaque decision problems into transparent decision problems on the
basis of a sequence of transparent steps (Howard, 1988). Opaque means ‘hard to
understand, solve or explain’ and transparent means ‘readily understood, clear or
obvious’. This approach offers a plausible solution and is used for the first time in
the field of fire safety engineering in this research project. It has been successfully
used earlier in other fields such as economics, nuclear energy and resources, policy
analysis, scientific research management, industrial management, manpower
planning and medical diagnosis and defence.
213
The MODA approach demonstrates the impact on evacuees on the same scale for all
the concept design options considered. The parameters (variables) are given
appropriate weights based on data generated from statistics and survey reports. The
priorities of conflicting key issues are assigned with the help of Analytical Hierarchy
Process (see Chapter 3, Section 3.1.2). The analysis is prescriptive and the approach
can assist decision makers to understand all aspects of the building evacuation
process and can reveal insights into design options.
7.2 Risk Analysis of Building Evacuation System
Important aspects of lift evacuation system include evacuation time periods and
tenability limits of fire, smoke and toxic gases in evacuation routes. Researchers
have addressed these areas in isolation (Kuligowski, 2003 and Klote, 2003). A
comprehensive approach for addressing the risks in relation to these aspects is
discussed here. An integrated risk evaluation model is proposed. The objectives of
this chapter are to:
1. develop a model for analyzing the risks involving ‘uncertainty’, ‘panic’ and
‘injuries (nonfatal and fatal)’ in building evacuation system; and
2. develop a decision model applicable to risk assessment of building
evacuation system; and
3. analyse the risks associated with the lift evacuation system.
7.2.1 Assumptions
The risks are compared between lift and stair systems based on the following
assumptions:
• low risk (or decision uncertainty) occurs in the lift and stair systems. This
may be caused by longer lift waiting time or longer travelling time in the
stairs.
• medium risk (or panic) occurs in the lift and stair systems. This may be
caused by visual threat (pre-life threatening condition) or unavailability of
evacuation route (lifts or stairs).
214
• high risk (or injuries) occurs in the lift and stair systems. This may be caused
if the evacuees are exposed to life threatening conditions in the lift system or
stairs. Life threatening conditions can arise from smoke, toxic gases, fire/
temperature and precipitated risk from pre-existing health conditions.
• evacuees are not considered safe inside the building, although they may be in
the protected stairs or protected lift lobbies or lift cabins. Hence evacuation is
considered complete, if they exit the building.
• lift evacuation strategy for one-fourth of the building population includes
mainly aged and disabled persons.
7.2.2 Methodology – Multi-Objectives Decision Analysis
The Multi-Objectives Decision Analysis (MODA) technique used here is based on
the Simple Multi-Attribute Rating Technique (SMART) of Edwards (1977) and
further illustrated by Donegan (2002) as explained in Figure 7-1.
Figure 7-1 – Multi-Objectives Decision Analysis Methodology
wn …… w2 w1
p2 p1 …… pn
Parameters
(Step 2)
Concept Design Option 1 Concept Design Option 2 …
Concept Design Option…
Values
(Step 4)
Parameters (Step 2)
Model Options (Step 1)
v1 v2 ….... vn
p2 p1 …… pn
Weights (Step 3)
w1 v1 w2 v2 …… wn vn + + +
Sensitivity Analysis (Step 5)
Result Evaluation (Step 4)
( )∑=
=n
i
iii pvwR1
215
The MODA involves the following steps:
Step 1: Identify concept design options
Step 2: Specify evaluation considerations and evaluation measures
Step 3: Specify weights to each parameter
Step 4: Determine value functions
Step 5: Analyse the results (sensitivity analysis)
The first step involves identification of potential concept design options for risk
assessment. It takes into account relevant events and risks in a system. This is carried
out with the help of influence diagram (see Figure 7-2). Influence diagram is a
structure display of decisions, uncertain events and outcomes, and provides a
snapshot of the decision environment at a single point in time. Evaluation measures
are the quantitative weights (importance) assigned to the evaluation considerations as
the contributing parameters may not have equal weights. A formal process is adapted
to award weights using AHP risk priorities and data from various sources. Values are
measures of the intensity, level or degree of hazard afforded by the parameters in a
particular design. Building evacuation models, fire hazard models and reliability
assessment models are used to determine the values. A scaling technique is used to
capture the essential meaning of quantitative values on a ratio scale. The ratio scale is
an interval scale with absolute zero on one end so that the values on it are absolute
rather than relative. The ratio scale identifies each individual parameter so that
reliable difference among the design models can be represented. Thus models can be
rated on a quantitative basis for results. The model is constructed in terms of n
evaluation parametric measures p1, p2, . . . , pn, and the overall value of the model is
given by:
( )∑=
=n
i
iii pvwR1
7.1
where the wi are weights and the vi(pi) are non-dimensional value functions that
normalise dimensional parametric measures pi (i=1, 2, …, n) into non-dimensional
values between 0 and 1 (or between 0 and 100). When the MODA approach is used
for risk assessment, the risk, as defined by Eq. 7.1, is interpreted as a total of
216
weighted multiple non-dimensional risk attributes. The impact of varying the
relative weight for the evaluation measures can be studied by sensitivity analysis.
7.3 Risk Assessment
7.3.1 Identify Concept Design Options and Evacuation Strategies
The use of lifts can be associated with evacuation strategies. Two evacuation
strategies, namely use of lifts by the entire building occupant population or a 75-25%
stair-lift split, were considered. In total five concept design options along with
alternative evacuation strategies are considered as listed and labelled in Table 7-1.
Table 7-1: Concept Design Options and their illustrations
Concept Design Illustration
A Lifts with unprotected lobby for use by 100% population
B Lifts with protected lobby for 100% population
C Lifts with double protected lobby for 100% population
D Stairs for 100% population (stairs only)
E Stairs used by 75% population (E-75) and protected lifts by 25% (E-25)
The concept design option D is an option for which the associated risk is deemed
acceptable. This option was included as a reference for the comparative study.
7.3.2 Evaluation Considerations and Evaluation Measures
Figure 7-2 indicates three risk levels in the influence diagram of building evacuation
risk model. Based on perceptual information, evacuees’ behavioral response
determines the actions he or she would perform. Perception is defined as the
awareness of the human being of environment through physical sensation. Evacuee
may simply respond in his/her environment or may cope (or interact) with this
environment or even communicate with other evacuees. Based on the level of
perception, two kinds of behavioural autonomy may be generated: either to evacuate
by stairs or lifts. The autonomous behaviour concerns the capability of acting
independently exhibiting control over their internal state.
217
Figure 7-2 – Influence Diagram of Building Evacuation Risk Model
Main issues of human behavioural response and life safety can cause decision
uncertainty, panic and injuries (nonfatal or fatal) in the lift and stair systems. The
evaluation considerations are kept common for risks in both systems (lifts and stairs).
Evaluation criteria (or parameters), units, symbols are specified in Table 7-2. The
parameters of similar nature are selected for both systems. The parameters may or
may not be directly related to the risk involved in the lift and stair systems. The
parameters which are not directly related to the risk can be referred as proxy-
parameters. The parameters of each issue are identified for assigning a corresponding
parametric weight.
Three main risks have seven parameters, which need to be given weights and values.
These parameters may be argued. The parameters p1, p2 and p4 are considered as
proxy-parameters as these parameters are not directly related to the risk. The
parameters are considered after a deliberation of thoughts, judgments and availability
of statistical data. For example, three parameters are considered for decision
uncertainty i.e. the time for building evacuation, the number of evacuees in queue
and the percentage of aged and disabled persons. Decision uncertainty may also
depend on building features, namely, the number of lifts, the fire safety and fire
protection systems and the level of the fire-affected floor. These parameters are
DECISION FOR ROUTE CHOICE HIGH RISK MEDIUM RISK LOW RISK
OR
Overcrowd
Decision
uncertainty
Untenable condition in
lift
Long
waiting
Use of
lifts
Use of
stairs
Perceptual
information
Action
Autonomy
decision
Untenable condition in
stair
Long
travelling
Lift un-
availability
Panic
Lift stuck during
transport
Injuries (nonfatal
or fatal)
218
lacking their importance and statistical resources. Evacuees may not be concerned
about the number of lifts (or stairs) and the types of fire safety and fire protection
measures in the building. The level of fire-affected floor is also involved in multiple
floor analyses (keeping in view of fire at each floor level). Hence these parameters
are not considered. In MODA, large number of parameters can be reduced to smaller
number of parameters or appropriate sub-sets (Watt, 1991).
Table 7-2: Risk related Parameters
Risk Category Parameter Symbol Unit
Lift waiting time (tLW) and transportation time (tLT ) or stair travelling time (tST)
p1 Second
Number of evacuees in queue in lift or stair lobbies
p2 Person
Decision uncertainty
Proportion of aged and disabled evacuees p3 %
Non-availability of evacuation route p4 % Panic
Time to exceed tenability limit for visual threat
p5 Second
Safety index p6 − Injuries (nonfatal and fatal) Presence of fire effluents in evacuation route
(e.g., temperature, concentrations of smoke and asphyxiant toxic gases.)
p7 −
7.3.3 Specify Weights
Each parameter is given a degree of importance (weight). The weights are estimated
from the survey reports and statistics keeping in view the maximum risk in the
evacuation routes. Using these weights, all the concept designs are analysed.
Decision Uncertainty: The parameters related to the decision uncertainty are lift
waiting time tLW and lift transportation time tLT (or stair travelling time tST), number
of evacuees in queue and percentage of aged and disabled persons in evacuation
route. The aged and disabled persons are more prone to the risk of decision
uncertainty (see Chapter 1, Section 1.2). A survey report by Sekizawa et al. (1996)
indicated that 47% residents used lifts and 42% residents used stairs and 7%
residents used both during a fire. A small 4% were shown as others. Splitting this
proportion and the proportion that used both lift and stair, an estimate of 52.5% was
given to the proportion of lift use and the rest to stair use (see Chapter 2, Section
219
2.4). Considering the average number of evacuees for lift evacuation, the required
timings and number of evacuees in queue are determined from ARENA stochastic
evacuation model (see Chapter 4). Literature review also indicated that the
percentage of aged and disabled was 16% (ABS, 2004 and Pauls, 1977). The lift time
period, number of evacuees in queue and percentage of aged and disabled persons
are given as:
� Average tLW and tLT (p1) = 449 seconds
� Average number of evacuees in queue (p2) = 3.51 persons/ floor
� Percentage of aged and disabled persons (p3) = 16% population
Panic: Panic may arise due to unavailability of the evacuation route (lift or stair) and
pre-life threatening condition (visual threat). Unavailability of lift depends upon the
lift malfunctioning due to excessive temperature or electric power failure or lift
maintenance breakdown. However, the evacuees are mainly concerned of their
evacuation safety and are not concerned about the reasons of unavailability of
evacuation routes. A survey was conducted in high-rise apartments with the
objectives of determining comforts of high-rise living (Mori and UHK, 2002) (see
Chapter 4, Section 4.1). In response to a question relating to the disadvantages of
high-rise living, 36% population reported fire escape, 20% reported lift breakdown,
2% reported strong wind, 2% reported heat and 4% reported lack of play areas (see
Figure 7-3). The figure shows the importance of parameters p4 and p5. Other
parameters such as lack of play areas, heat and strong wind are not significant and
are therefore not considered. The 20% population was concerned with the non-
availability of lifts in the apartment building. The 36% population was concerned
with the fire escape. The concern of fire escape occurs due to the presence of visual
threat in the evacuation route (which is ultimately time to exceed the tenable limit for
visual smoke). The evacuees’ concern of non-availability of lift and fire escape is
given in the form of population as:
� Evacuees’ concern for unavailability of lifts (p4) = 20%
� Evacuees’ concern for fire escape (p5) = 36%
220
Disadvantages of High Rise Living
Figure 7-3 – Disadvantages of High Rise Living (Mori and UHK, 2002)
Injuries: The injuries are related directly to three general areas – consequences of
inhalation of toxic products of combustion (smoke, CO, CO2, and other poisonous
gases, hypoxia and asphyxia), exposure to fire (burns, thermal injuries to airways and
incineration), shocks from injuries that precipitate deaths from pre-existing health
conditions (cardiac failure and respiratory diseases) during the exposure of toxic
gases and/or fire (see Chapter 2, Section 2.3, Table 2.6). The fatal injuries are
caused as the strength variables are lower than the load variables. The system is non-
functional and can be determined from safety index (see Chapter 5, Section 5.2.2).
The causes of fatal injuries in residential fires are interpolated as the fatalities caused
by hazardous exposure in evacuation routes. Miller (2005) showed that 28.5% of
evacuees were found dead while attempting to evacuate the buildings, but did not
give the locations of victims. The findings of residential fire deaths are also assumed
for evacuation routes although there may be fewer victims of burns/ incinerations
and more victims of toxic gases in evacuation routes. The following data is taken
from Table 2.6 (Miller, 2005):
� Causes of deaths due to smoke asphyxiant toxic gases, fire and temperature/
incineration (p6) = 174 cases
� Causes of deaths from injuries that precipitate deaths from pre-existing health
conditions (p7) = 11 cases
221
The safety index is used for determining the fatal injuries caused by smoke
asphyxiant toxic gases, fire and temperature/ incineration (the probability of time
period for occupants’ evacuation is less than the probability of time to exceed the
tenability limits).
Parametric Global Weights: Table 7-3 gives the global weights of parameters. The
parametric global weights were obtained from the individual values of the parameters
{for example p1, 449 seconds is 58.84% of 763 seconds – the total waiting and
transportation time for the entire population, similarly for p2, 3.51 persons is 10.96%
of 32 persons – the floor population}. However, the values of parameters from p4 to
p7 have different basis (units) to derive the global weights. The risk priorities along
with group weights for p1 to p7 are indicated in the value tree (see Figure 7-4). A
value tree represents the structural hierarchical position of all the parameters. The
risk priorities were obtained in Chapter 3 from the Analytical Hierarchical Process
(AHP). The weights for parameters p1 to p7 did not share the same basis as the values
are based on evacuation periods, percentage of evacuees, evacuees’ response and
number of injuries (nonfatal and fatal) obtained under different conditions. Under
such conditions, multi-criteria decision approach is the most appropriate method to
determine the group weights and global weights.
Table 7-3: Parametric Values and Weights relating to Building Evacuation
Risk
Category
Parameter
(p)
Individual
Value
Individual
Percent
(%)
Group
Percent
(%)
Group
Weight
Global
Weight
(w)
p1 449 seconds 58.84 68.58 0.6858 0.0051
p2 3.51 persons 10.96 12.77 0.1277 0.0009
Decision uncertainty
p3 16 percent 16 18.65 0.1865 0.0014
p4 20 percent 20 35.7 0.3570 0.0370 Panic
p5 36 percent 36 64.3 0.6430 0.0668
p6 174 cases 94.05 94.05 0.9405 0.8359 Injuries
p7 11 cases 5.95 5.95 0.0595 0.0529
Total 1.0000
.
222
Figure 7-4 – A Value Tree for the Parametric Global Weights
7.3.4 Value Functions
The purpose of introducing value functions is to normalize dimensional risk
parameters with given lower and upper bounds into non-dimensional parameters with
0 to 1 or 0 to 100% scale. The values of the value functions are obtained from
various models and analysis for all the concept design options. The parameters are
multi-dimensional and require deriving a value function. The value functions for p1
to p7 are given below:
Lift Waiting and Transportation Time, and Stair Travelling Time (p1): The
building evacuation times {(tLW + tLT) or tST} were determined from ARENA
stochastic models (see Chapter 4). The parameter strength varies from 338 seconds
Group Weights from statistics, survey, stochastic modelling,
FDS modelling
Priorities from AHP
Matrix
Decision uncertainty
0.0073 Panic
0.1039
Building Evacuation Time
p1 = 0.6858
Evacuees in queue
p2 = 0.1277
Unavailability of the evacuation route
p4 = 0.3570
Injuries (nonfatal/ fatal)
0.8888
Exposure to asphyxiant toxic gases, fire and temperature
p6 = 0.9405
Building evacuation
Visual threat (pre-life threatening condition)
p5 = 0.6430
Aged / disabled persons
p3 = 0.1865
Parametric Global Weights:
p1: 0.0073 × 0.6858 = 0.0051
p2: 0.0073 × 0.1277 = 0.0009
p3: 0.0073 × 0.1865 = 0.0014
p4: 0.1039 × 0.3570 = 0.0370
p5: 0.1039 × 0.6430 = 0.0668
p6: 0.8888 × 0.9405 = 0.8359
p7: 0.8888 × 0.0595 = 0.0529
Total = 1.0000
Pre-existing health condition
p7 = 0.0595
223
to 763 seconds (see Table 7-5). The strength needs to be translated into a value. The
longest time is given a value function of 100 while the minimum time is given a
value function of 0. In mathematical notation, it can be written as:
• v (763) = 100 (A, B and C)
• v (338) = 0 (E-25)
where v(763) means the ‘value function of 763 seconds’, which is given as 100.
Similarly, the shortest time is 338 seconds and is given a value function of 0. This
enables the estimation of the value of (tLW + tLT) or tST of any model between the
most and the least times (100 and 0) on a ratio scale. However, the values are not
necessarily corresponding to arithmetically derived values between the two extremes.
This is further explained with the help of bisection curve (bisection divides the
parametric value equally according to its strength) (see Figure 7-5). There is no
significant decrease in stair travelling time with the reduction of population size. If
the population via stair is reduced to 50% or 25%, the stair travelling time will not be
significantly changed although the population size is varying considerably. The lift
evacuation time for 50% of the population is determined as about 450 seconds
[v(450)=50]. Similarly, the lift evacuation time for 75% of the population is
determined as about 619 seconds [v(619)=75]. The time gives the value function in
Y-axis. Accordingly the graph is skewed upward. The extreme values along with the
midpoint are plotted to determine the value for 367 seconds and 372 seconds. After
earmarking two value functions at the extremes, two value functions are obtained
from the graph. The value functions for 367 seconds and 372 seconds are calculated
as:
• v (367) = 15 (E-75)
• v (372) = 20 (D)
224
0
10
20
30
40
50
60
70
80
90
100
338 363 388 413 438 463 488 513 538 563 588 613 638 663 688 713 738 763
Building Evacuation Time (second)
Value
Bisection value curve
Arthmetic value curve
Figure 7-5 – Value Functions for Building Evacuation Times
Number of Evacuees in Queue (p2): The number of evacuees in queues was
determined from ARENA stochastic evacuation models (see Chapter 4). The
parameter strength varies from 0 to 8.16 persons/ floor (see Table 7-5). The longest
queue containing 8.16 persons/ floor is given a value function of 100. The stairs for
75% of the population has no queue and a value function of 0 is given. In
mathematical notation, it can be written as:
• v (8.16) = 100 (A, B and C)
• v (0) = 0 (E-75)
Lifts for 25% of the population has 1.2 persons/ floor. Stairs for the entire population
has only 0.05 person/ floor. The following value functions for these values are
determined proportionately:
• v (1.2) = 14.7 (E-25)
• v (0.05) = 0.6 (D)
225
Percentage of Aged and Disabled Persons in the Evacuation Route (p3): The
percentage of aged and disabled persons in the apartment building was determined
from the pilot survey (see Chapter 4). The parameter strength varies from 0 to
15.75% (see Table 7-5). Decision uncertainty may vary with the type of evacuation
routes. For example, aged and disabled persons require travelling several steps in
stairs, whereas there are no physical efforts in lift evacuation. Decision uncertainty
would be more in stairs than in lifts. It varies with the types of evacuation routes
such as lifts for aged and disabled persons, lifts for all persons and stairs for all
persons. If lifts are permitted for 25% of the population, lifts may contain most or all
of the aged and disabled persons and almost none in the stairs. Likewise, most or all
of the aged and disabled persons will be using stairs, if lifts are not permitted. The
parameter of aged and disabled persons in evacuation routes is given a risk factor
(see Table 7-4). The correlations among the sub-parameters are based on the Saaty’ 9
point scale (Saaty, 1980), as discussed in Chapter 3. The point scores are based on
judgment.
Table 7-4: Matrix (3 × 3) for Priorities Risk Factor (p3)
Category p3a p3b p3c Priority Risk Factor
Aged and disabled persons in
dedicated lifts (p3a)
1 1/3 1/9 0.077
Aged and disabled persons in lifts
– along with general public (p3b)
3 1 1/3 0.231
Aged and disabled persons in stairs
(p3c)
9 3 1 0.692
The aged and disabled persons can promptly evacuate the building using lifts and
therefore a risk factor of 0.077 (low risk for p3a) is given. The lift system for all
evacuees is given a risk factor of 0.231 (medium risk for p3b) and the stair system is
given a risk factor of 0.692 (maximum risk for p3c). Considering 15.75% as the
percentage of aged and disabled population, the values are assigned 1.21 (15.75 ×
0.077) in dedicated lifts, 3.63 (15.75 × 0.231) in lifts (with general public) and 10.9
(15.75 × 0.692) in stairs. A value of 10.9 for the stair route is given a value function
of 100. No one in the stair (75% population evacuation) is given a value function of
0. Therefore, the following value functions are proportionately given:
226
• v (10.9) = 100 (D)
• v (0) = 0 (E-75)
• v (3.63) = 33.3 (A, B and C)
• v (1.21) = 11.1 (E-25)
Unavailability of Evacuation Route (p4): The lifts may not be available due to
malfunctioning of lift components due to excessive temperature, electric power
failure or maintenance breakdown or manual tripping. The unavailability of lifts was
found to be 26.48% for unprotected lift lobby and 0.48% for single protected lift
lobby and double protected lift lobby (see Chapter 6). The unavailability of stairs
depends upon the crowd density (if the evacuees’ density is more than 3.5 P/m2, the
evacuees can not move) or locking of stair door. Data are not available for the
unavailability of stairs and it is conservatively assumed that the stairs would always
be available in the buildings. The parameter strength varies from 0 to 26.48 (see
Table 7-5). The following value functions are given:
• v (26.48) = 100 (A)
• v (0) = 0 (D and E-75)
• v (0.48) = 1 (B, E-25 and C)
Smoke Visual Threat (p5): The time to exceed the visual smoke obscuration in the
evacuation route was determined by fire hazard modelling (see Chapter 5). The
parameter strengths vary from 241 to 1778 seconds (see Table 7-5). The following
value functions are given:
• v (241) = 100 (A)
• v (1778) = 0 (C)
The maximum time is given a value function of 0 as there is the minimum risk of
visual smoke threat. The minimum time is given a value function of 100 as there is
the maximum risk of visual smoke threat. The following value functions are derived
proportionately:
227
• v (1147) = 42 (E-75)
• v (515) = 82 (D)
• v (500) = 83 (E-25)
• v (388) = 90 (B)
Safety Index (p6): The safety indices were determined from stochastic evacuation
and FDS models (see Chapter 5). The parameter strengths are between -10.82 and 32
(see Table 7-5). The following value functions are given:
• v (-10.82) = 100 (A)
• v (32) = 0 (E-25)
The following value functions are derived proportionately from the extreme values:
• v (3.57) = 66 (D)
• v (6.66) = 59 (E-75)
• v (2.55) = 68.77 (B)
• v (17.61) = 33.6 (C)
Number of Fire Effluents Causing Deaths from Pre-existing Health Conditions
(p7): The deaths from the pre-existing health conditions (mainly among the aged and
disabled persons) were influenced by the presence of fire effluents such as visual
smoke, asphyxiant toxic gases and fire/ temperature. The parameter strengths were
obtained from Chapter 5 (see Tables 5-2 to 5-4) and are given in Table 7-5. The
parameter strengths vary from 0 to 3. The following value functions are given based
on the extreme values:
• v (3) = 100 (A)
• v (0) = 0 (C and E-75)
The following value functions are derived proportionately based on the extreme
values:
228
• v (2) = 66 (B)
• v (1) = 33 (E-25 and D)
Parameter Values: The parameter strengths of p1 to p7 are given in Table 7-5 while
Table 7-6 gives the summary of assigned values.
Table 7-5: Parameter Strengths for Concept Design Options
Concept Design Option
E
Parameter
A B C D
E-25 E-75
p1 (second) 763 763 763 372 338 367
p2 (person) 8.16 8.16 8.16 0.05 1.2 0
p3 (%) 15.75 15.75 15.75 15.75 15.75 0
p4 (%) 26.48 0.48 0.48 0 0.48 0
p5 (second) 241 388 1778 515 500 1147
p6 (number) -10.82 2.55 17.61 3.57 32 6.66
p7 (number) 3 2 0 1 1 0
Table 7-6: Summary of Assigned Values
Concept Design Option
E
Parameter
A B C D
E-25 E-75
p1 100.0 100.0 100.0 20.0 0.0 15.0
p2 100.0 100.0 100.0 0.6 14.7 0.0
p3 33.3 33.3 33.3 100.0 11.1 0.0
p4 100.0 1.0 1.0 0.0 1.0 0.0
p5 100.0 90.0 0.0 82.0 83.0 42.0
p6 100.0 68.8 33.6 66.0 0.0 59.0
p7 100.0 66.0 0.0 33.0 33.0 0.0
Results: The value functions are multiplied by the weights and all the weighted
grades are added to give a final grade of risk. Table 7-7 shows the total risk obtained
for the five models (Options A, B, C, D and E). The risk grade reflects the risk
associated with a particular model.
229
Table 7-7: Total Risk Values for Concept Design Options
Concept Design Option
E = (E-25) + (E-75)
Parameter Global
Weight
(w) A B C D
E-25 E-75 E
p1 0.0051 0.510 0.510 0.510 0.102 0.000 0.077 0.077
p2 0.0009 0.090 0.090 0.090 0.001 0.013 0.000 0.013
p3 0.0014 0.047 0.047 0.047 0.140 0.016 0.000 0.016
p4 0.0370 3.700 0.037 0.037 0.000 0.037 0.000 0.037
p5 0.0668 6.680 6.012 0.000 5.478 5.544 2.806 8.350
p6 0.8359 83.590 57.510 28.086 55.169 0.000 49.318 49.318
p7 0.0529 5.290 3.491 0.000 1.746 1.746 0.000 1.746
Total Risk 1.0000 99.907 67.697 28.770 62.635 7.356 52.200 59.556
7.3.5 Sensitivity Analysis
The results of total risk values given in Table 7-7 show that the proposed use of
protected lift system to evacuate 25% of the population and the remaining population
by stairs (Option E) and lifts with double protected lift lobby (C) involve less risk
when compared with the use of stairs (D). On the basis of the overall risk values
reported in Table 7-7, suitable evacuation schemes can be recommended. To
determine the robustness of results, a sensitivity analysis is conducted against the
weightings used in the risk analysis for panic (w4 and w5). Table 7-8 and Table 7-9
show the value of total risks for the different design options as a function of changes
to the weights used in the risk analysis for panic {w (P4 and P5) = 0.0370, 0.0668}. If
panic is given zero weighting, two parameters (p4 and p5) would have zero
weighting. On the other hand if they are given 100% weighting, the remaining five
parameters (p1, p2, p3, p6, and p7) would have zero weighting.
230
Table 7-8: Total Risk Values from Analyses based on a Zero Weight for Panic
Concept Design Option Parameter Global Weight
(w) A B C D E
p1 0.0051 0.510 0.510 0.510 0.102 0.077
p2 0.0009 0.090 0.090 0.090 0.001 0.013
p3 0.0014 0.047 0.047 0.047 0.140 0.016
p4 0 0.000 0.000 0.000 0.000 0.000
p5 0 0.000 0.000 0.000 0.000 0.000
p6 0.8359 83.590 57.510 28.086 55.169 49.318
p7 0.0529 5.290 3.491 0.000 1.746 1.746
Total Risk 0.8962 89.527 61.648 28.733 57.158 51.170
Table 7-9: Total Risk Values from Analyses based on 100% Weight for Panic
Concept Design Option Parameter Global Weight
(w) A B C D E
p1 0 0.000 0.000 0.000 0.000 0.000
p2 0 0.000 0.000 0.000 0.000 0.000
p3 0 0.000 0.000 0.000 0.000 0.000
p4 0.037 3.7 0.037 0.037 0 0.037
p5 0.0668 6.68 6.012 0.000 5.478 8.350
p6 0 0.000 0.000 0.000 0.000 0.000
p7 0 0.000 0.000 0.000 0.000 0.000
Total Risk 0.1038 10.380 6.049 0.037 5.478 8.387
Figure 7-6 shows how the total risk values for the different design options vary with
changes in the weighting placed on panic. The use of protected lift system to
evacuate 25% of the population and the remaining population by stairs (Option E)
would have total risk values of 51.17 and 8.38 when the value of weight given for
panic was varied from 0 to 100%. Similarly, the concept design option B would have
risk values of 61.64 and 6.04 and the concept design option D would have risk values
of 57.15 and 5.47 at the two extremes. Comparing these risk values, it can be seen
that the safe design option E has changed to option D at 62 (equivalent risk) for the
case of the highest weight given to panic (100%). The safe design option E is
sensitive to the slight variation in the proportion of building population. The most
safe evacuation concept design option (Option C) is not affected by the change. The
risk value for unprotected lift lobby (Option A) remains relatively higher than those
for stairs (Option D) despite the change of weight values for panic.
231
Concept Design Option Weight Given
for Panic Resultant Weight w
(%) A B C D E
0 89.62 89.527 61.648 28.733 57.158 51.170
100 10.38 10.380 6.049 0.037 5.478 8.387
Figure 7-6 – Sensitivity Analysis for Different Weights Placed on Panic
7.4 Analysis of Results
The overall risk values calculated for the five design options are shown in Figure 7-7.
If the stair alone evacuation (Option D) is considered as an acceptable evacuation
design, then the lift with simply protected lobby for one-fourth of the building
population (Option E) and double protected lift lobby (Option C) provide acceptable
alternatives since the latter two have lower associated risk than the former. The
acceptable risk RA is shown in Figure 7.7 (RA = 62.635). Lifts with unprotected
lobby (Option A) has the highest risk, followed by lifts with simply protected lobby
232
for entire population evacuation (Option B). The results also show that the risks are
considerably reduced by providing protected lift lobby. A combination of lifts and
stairs provide a lower level of risks in comparison to stair alone evacuation (Option
D) or lift alone evacuation with the same lobby protection (Option B).
0 20 40 60 80 100
A
B
D
E
C
Risk Value
Concept Design Option
RA
Figure 7-7 – Risk Values for Concept Design Options
7.5 Conclusion
This chapter has demonstrated that the MODA method can provide a rational basis for
determining the risks. Multiple risk parameters in relation to human behavioural
response, fire hazard exposure and reliability of lift operational mechanism were
included in the assessment. This method was applied to all design options considered
and the out come of risk values were meaningful only on relative terms. The
imperfection in the selection of parameters might influence the absolute risk values
but would not significantly alter the relative scales of risk associated with alternative
design options. The results of the assessment are summarized as follows:
• Lifts with unprotected lobby contribute to the maximum for risks.
233
• Lift with protected lobby does not provide adequate safety in comparison to
stair for the entire building population. Overall risks of lifts with protected
lobby were slightly higher than the risks of stair system.
• A combined system of lifts with protected lobby for one-fourth of population
and stair for three-fourth of population can provide adequate safety in
comparison to individual system alone (lifts or stairs).
• Lifts with double protected lobby provides a better performance.
The MODA method provides a versatile means for risk assessments. It is based on a
comparative study of multiple options and incorporates multiple risk attributes into
the evaluation. The MODA method involves the ranking of level of importance for
multiple risk attributes. This ranking is still by and large empirical and/or subjective.
In its application to fire risk assessment, the method is linked to other means of
evaluations such as the ASET/RSET analysis, stochastic modeling and safety index
analysis. This chapter did not include the cost-effectiveness analysis in the risk
assessment, although in theory such an inclusion is achievable.
234
8. FEASIBILITY AND DESIGN CONSIDERATIONS
8.1 Introduction
The traditional prescriptive approach to building design embodied in the building
codes is based on various norms and guidelines. Experience based traditional fire
regulations may not be always adequate and therefore alternative solutions are put
forward. Such alternative solutions must be adequately investigated if uncertainties
exist. The performance-based fire safety designs are used in many parts of the world,
which are also acceptable under the provisions of the Building Code of Australia
(BCA). The objective of BCA is to safeguard occupants from illness, injury or death
during building evacuations. A building is to be provided with safeguards so that the
occupants have sufficient time to safely evacuate before the environment in any
evacuation route becomes hazardous due to fire. At one of the fire safety
conferences, a NSW Fire Brigade officer stated that the NSW Fire Brigade is
supportive of innovative designs that provide equal evacuation opportunities for all
the occupants (Honeybrook, 2002). The use of lifts for building evacuation during
fire emergencies goes against the “norm”, and therefore requires consideration of a
range of factors, and confirmation and demonstration that all the concerns and safety
factors are addressed.
The design of lift systems plays a vital role in the safety of lift evacuation system.
This research project was undertaken by considering the uncertainties of parameters
relating to evacuation procedure (human movement and behavioural response) and
variable conditions in fire development (wind and stack effects) in order to determine
feasible, safe and effective design solutions. The design alternatives were
investigated for a limited number of hypothetical fire scenarios. The results were
evaluated and the following two options were determined for lift evacuation as
discussed in Chapters 4 to 7:
• Lifts with protected lobby to evacuate one-fourth (25%) of the building
population and stairs for the rest of the population (75%)
• Lifts with double protected lobby to evacuate the entire population
235
This chapter shows the feasibility of these options as alternative evacuation systems
and proposed various redundancy measures to enhance the occupant safety during
evacuation.
8.2 Feasibility Options
8.2.1 Lifts with Protected Lobby to evacuate 25% of the Building Population
Protected lift lobbies can be managed for emergency evacuation efficiently and
effectively for one-fourth of the building population. Able-bodied people can use the
stairs for evacuation. The statistics showed that the percentage of aged and disabled
persons in apartment buildings is 16%, which can be managed using the lift system
(see Figure 8-1).
Figure 8-1 – Evacuation Option 1: One-Fourth of the Building Population using
Protected Lifts and the rest using Stairs
236
The Occupant Emergency Evacuation Plans are prepared for high-rise apartment
buildings, which must incorporate predefined means of evacuation facilities for the
building occupants. It should clearly indicate that the lifts are to be used by a fraction
of population such as the aged and disabled persons. The sign posted in front of lifts
“DO NOT USE LIFT IN CASE OF FIRE” can be replaced with “DO NOT USE
LIFT IN CASE OF FIRE, LIFTS ARE PERMITTED ONLY FOR ASSISTED
EVACUATION [AGED AND FRAIL PERSONS]”. There will be a necessity to
develop and evaluate a strategy for public education to ensure the effectiveness of the
proposed lift evacuation system. Without proper preparation and training (fire drills),
evacuees may become fearful of the dangerous conditions as uncertainties may arise
in relation to the proportion of population to be evacuated by lifts.
8.2.2 Double Protected Lift Lobby for the Entire Building Population
As shown in previous chapters, evacuation via double protected lift lobby provides is
safer than evacuation using stair systems. Details of the provision of double protected
lift lobby (conceptual design option C) are shown in Figure 8-2.
Figure 8-2 – Evacuation Option 2: Double Protected Lift Lobby
The BCA includes DTS provisions relating to travel via fire-isolated exits. The
requirements for fire-isolated exits typically require a smoke lobby, if more than two
FD
FD LIFT LOBBY
EX
IT
SD
LIFT
EX
IT
SD4
FD – Fire Door SD – Smoke Door
237
access doorways are open to a required fire-isolated exit in the same storey. In the
absence of pressurisation of lift shaft, a protected lift lobby can not be used as smoke
lobby as it permits smoke leakage to the lift shaft. For creating a smoke lobby in the
fire-isolated exit (lift system), one set of doors can be self-closing fire door while the
other can be a magnetically held open door or a normal smoke door. Smoke and hot
gases can be restricted to the fire-affected apartment unit (or corridor), i.e. away from
the lift shafts. The disadvantages of providing a double protected lift lobby are the
additional space requirement for smoke lobby between the two doors (additional
space is approximately 1.5 times more than that for the stair lobby). Evacuees may
also feel unease due to the double door opening and closing during their regular use
of lifts. Further, additional space is required in the lift lobby for accommodating the
entire population.
8.3 Redundancy Measures
This research has demonstrated the feasibility of the proposed two options for lift
evacuation. In this section, a number of redundancy measures are proposed that are
likely to enhance the reliability of proposed lift evacuation systems.
8.3.1 Common Lift and Stair Lobby
There are advantages in high-rise apartment buildings to provide access within the
same protected lobby to both stairs and lifts designed for operation during fire
emergencies. Building codes require fire compartmentation by providing a lift lobby
between the lift shaft and risk area and such lifts are often used as emergency or fire
fighter lifts. The conventional rectangular arrangement of lifts may optimize the
space requirement and allows common space to be utilised for both lift and stair
systems (see Figure 8-3).
238
Figure 8-3 – Rectangular Arrangement for a Common Lift Lobby
The scissor type stairs also satisfy the BCA DTS requirements as there is double
door protection (smoke lobby). With the provision of double door, pressurisation is
not a necessity (ABCB, 2005) and it will not cause any concern to lift door opening
and closing. The provision can be useful for lift evacuation. Stair is also
approachable as a redundancy measure for lifts. The provision can be used as fire
brigade staging area (as it provides an alternative evacuation facility). Other safety
features may also be incorporated into the system such as:
(a) a monitor in each lift lobby to indicate where the lifts are at any one time;
the time when the next lift will come; whether it is still safe to wait for the
lift and if one should move into the stairs; and
(b) the provision of an expanded landing to allow disabled persons to wait
there if the lifts cannot be used.
8.3.2 Refuge Area
Presently two types of systems are used for the aged and disabled persons, i.e.
staging area and horizontal separation. The staging areas (refuge area) are intended
as spaces in which people with disabilities can safely wait during a fire. Horizontal
separation consists of one or more barriers which divide a floor into separate areas
with the intent of restricting smoke and fire spread (compartments or landings in the
FD
FD LIFT LOBBY
EX
IT
EX
IT
SD
LIFT
LIFT
SCISSOR STAIRS
FD – Fire Door SD – Smoke Door
SD
239
stairwell). The refuge area provided in the building can be utilized as a redundancy
measure (see Figure 8-4). In this case, evacuees will have a choice to exit the
building using stairs or lifts or move to refuge area.
Figure 8-4 – Lifts located in the Refuge Area
8.3.3 Scattered Design of Lift System
During heavy traffic conditions, there are many hall calls to serve and the lifts have a
tendency to move side by side, which is called bunching of lifts. Lift stops at the
nearest floor. However, this phenomenon can be removed by giving priorities for the
long or timeout hall calls using electronic controls (Siikonen, 1997).
Lifts can be used effectively by scatter design lifts. This strategy involves
distribution of evacuees to different locations in the building for evacuation so that
the evacuees would not wait longer for lifts at one strategic location. Further, there
would be less door opening and closing due to which unsafe conditions do not arrive
quickly. This design provides a longer duration of tenable environment in the lift
lobby.
FD
FD LIFT LOBBY
EX
IT
EXIT
SD
FD
SD
REFUGE FLOOR
LIFT
FD – Fire Door SD – Smoke Door
240
8.3.4 Pressurization
The BCA does not mandate lift shaft pressurization for apartment buildings.
Sufficient pressure difference across the lift lobby and corridor can be maintained to
restrict the smoke spread in the unprotected and protected lift shafts. Pressure can be
maintained in the strategic location in the lift shaft or in the lift lobby of the fire-
affected floor as the occupants will be waiting there. The basic theory of this smoke
control method system may be similar to that of stair pressurisation which increases
the pressure inside by supplying sufficient amount of air to it. However, the
maximum pressure should not hinder the lift door opening and closing. The
pressurization system is generally interlinked with the smoke alarm system. Lift
landing door frequently opens and closes, hence the total leakage area is the leakage
area of both the lift landing door opening and closing (at the time of service) and the
leakage area between the lift landing door and lift landing wall frame at all the floors.
BS 5588, Part 4, Chapter 5 gives the required air flow for pressurised space as:
2/1827.0 EEE PAQ =
where
QE is the air supply to the pressurized space (m3/s)
AE is the total leakage area out of the space (m2)
PE is the pressurisation level in the pressurized space (Pa)
8.3.5 Smoke Seal in Lift Landing Door
If the leakage area of the lift doors is reduced with the help of smoke seal (or gasket)
then tenable conditions can be maintained in the lift shaft and upper floors. The
smoke seal can fit in between the lift landing door and lift landing wall frame (see
Figure 8-5). Warnock Hersey International Inc. (WHI, 1987) conducted a test to
determine the resistance to air leakage of brush type gasketing for lift landing doors.
The test was conducted after imparting 100 000 cycles to the gasket and the air
leakage was measured at several differential pressures at room temperature and at an
elevated temperature of over 400°F (204°C). The application of gasket showed at
241
least a 95% decrease in leakage rates. However, this option is applicable if the lifts
are provided in the unprotected lift shaft and the lifts are not served at the fire-
affected floor. Lift operating software is to be programmed with fire detection
system, however, this aspect was not analysed in this research.
Figure 8-5 – Smoke Seal in the Lift Landing Door and Wall Frame
8.4 Fire Protection Measures for Lift System
The fire protection systems in the lift system are provided to combat fires in lift
machine room and lift shafts involving lift equipment. During the fires in this system,
lift mechanism will not be operative and therefore it can not be used by the building
occupants. However, if the lift mechanism is still operative, safety provisions must
be made independently to cut off the lift system from regular operations after
returning the lift car to the designated landing (ground floor). A fire occurrence in
the lift machine room at the top may not necessarily require a building evacuation. A
fire occurrence in the lift shaft would give smoke and hot gases plume in the lift shaft
with leakages to the floors, which may require building evacuation. Primary route of
building evacuation (stair) is available in this case.
242
Automatic fire sprinkler protection is required in the lift machine room, and in the
top and bottom of lift shaft (NFPA 13, 1996). However, automatic fire sprinkler
protection is exempted in the machine rooms of traction-type lifts when the rooms
are located on the top of the lift shaft and are separated from other areas of the
building, other than the shaft, by not less than one-hour fire resistive occupancy
separation. The exception of the top and bottom of lift shaft is also provided, if the
lift shaft is of non-combustible construction and does not contain combustible
hydraulic fluid (NFPA 13, 1996). The system is provided with at least one smoke
detector and is designed in a separate and independent sprinkler branch.
8.5 Strategic Planning
The successful emergency evacuation of a building requires comprehensive
management procedures. Groner (2002) quoted - “Engineering the hardware to
prevent intrusions of heat, smoke and water is not the primary obstacle. The greater
challenge is in real world use of these systems-in the strategic planning, interface
design and operator training that will enable the safe, effective and efficient use of
elevators during emergencies”. If lifts are considered for emergency evacuation,
regular drills are required to be conducted to inculcate a sense of confidence among
the occupants. Building Fire Safety Regulation may incorporate the provisions of
conducting regular fire drills for the residents of apartment buildings. For a strategic
planning of partial lift evacuation, permanent residents can be taught the proper use
of stairs and lifts to avoid ambiguity in the evacuation procedure.
8.6 Conclusion
This chapter has demonstrated the feasibility of the two options of lift evacuation
systems. It has been shown that the lift system can be provided for evacuating part of
the population (25%) without reducing the level of fire safety. For the evacuation of
the entire population, lift systems need special design considerations such as double
door protection. The lifts require adherence of rules and norms relating to regular
electrical supply and precautionary measures for water spread. However, redundancy
measures can provide additional safety to the lift system. If lifts become a total or
243
partial means of alternative evacuation facility, the stair capacity should not be
reduced. Lifts are the mechanical means of transportation and require regular
maintenance. Lifts can not be a substitute for a non-mechanical and a non-
maintainable means of stair system. The stairs are required to maintain through an
effective and graded program of fire prevention inspections. The main thrust can be
diversified to discover and eliminate the fire hazards in the evacuation routes such as
stair and lift systems in the buildings.
Current building codes and regulations provide limited aspects of lift installation.
They are mainly related to the general requirements for normal lifts and certain
provisions for people with disabilities. If lifts are used as evacuation systems during
fire emergencies as proposed in this research, their design and installation in
apartment buildings deserve a special consideration as part of the overall design
process.
244
9. CONCLUSIONS
9.1 Summary
The main objective of this research was to study if lifts provide an acceptable means
for evacuation in apartment buildings during fire emergencies. Building evacuation
models, fire hazard models and risk assessment models have been developed for
systematic and in-depth analysis of lift operations. They were applied to a number of
hypothetical case studies for uncertain and complex factors. A brief summary of
their features, applications and evaluations are given as follows:
• A research methodology was developed. Risks associated with the lift
evacuation system were identified and risk cause-effect relationship was
established. The issues of human behavioural response, fire hazards and lift
operational mechanism were identified. Risk priorities were identified with
the help of Analytical Hierarchical Process. Considering that panic may be a
rare event and the outcome of the study of this phenomenon is inclusive, a
relatively low risk priority was give to this parameter in the risk analysis. A
research strategy was developed by reducing the level of consequences to an
acceptable level. Design options and evacuation strategies for lifts were
proposed for comparison with stairs.
• A pilot survey predicted the residents’ awareness level of emergency
evacuation procedure and their inclination toward the use of lifts. It also
demonstrated the necessity of alternative evacuation facility in high-rise
apartment buildings. Interviews also showed the concerns and controversy of
the unresolved issue (use of lifts for emergency evacuation).
• A stochastic simulation model was developed for the assessment of
parameters relating to lift evacuation system. The parameters included
relevant lift time periods (waiting, transportation, pre-evacuation and
evacuation times) and the number of evacuees in queues at various floor
levels. The model was developed for a 38 storey hypothetical building.
245
Results of hypothetical scenarios for the evacuation of the entire and one-
fourth (25%) of the building population using lifts under uncertainties were
generated. The output variables for one-fourth of the population were
determined for evaluating a safe and efficient lift evacuation strategy. Output
variables were plotted for obtaining the mean and standard deviation. The
simulation model was useful to comprehensively evaluate the risks relating to
human behavioural response and life safety in the lift system. It incorporated
effects of different conditions (social, physiological and psychological
characteristics, temporal and a priori heuristics of the lift domain), which may
vary during the emergency evacuations. The modeling results have provided
a base for evaluating relevant lift parameters for an acceptable level of safety
(for example, minimal human behavioural or life safety concerns during lift
evacuation by a smaller population).
• A stochastic model for stair system (similar to lift stochastic model) was also
developed. The parameters included stair time periods and number of
evacuees in queues at various floor levels. The stair simulation model was
useful for comparing the results with lift system. The output variables were
determined for the evacuation of the entire and three-fourth of the population
(75%) using stairs. The results from the stochastic models of lift and stair
systems revealed that the lift evacuation time for one-fourth of the building
population was within the acceptable limit of stair evacuation time. The
number of evacuees in queue for lifts was less than 2 persons. This showed
that the human behavioural response such as decision uncertainty would be
minimal during the evacuation of one-fourth of the population using the lift
system.
• Models of risks associated with fire and smoke hazards in evacuation routes
were proposed for estimating the smoke, asphyxiant toxic and heat exposure
hazards. The time, concentration and toxicity of the fire effluents were
considered in combination for predicting the probable time to incapacitation
or death. A 38-storey hypothetical building (used for stochastic evacuation
modelling) was analysed for fire hazards. Twenty four fire scenarios were
analysed after incorporating stochastic uncertainties relating to wind speed
246
and vertical location. Three compartmentation strategies along with
additional evacuation strategies were considered. The fire profile was
characterised in terms of smoke extinction coefficient (visibility), asphyxiant
gas CO concentration, CO2 concentration, O2 concentration (low oxygen
causing hypoxia), temperature and radiant heat flux. The output of the design
fires provided the initial conditions for a smoke movement analysis, including
the time when the lifts and stairs became unsafe and unusable by the
evacuees. These models calculated the incapacitating dose of fire effluents in
the lift and stair systems. The concept of fractional effective doses of safe
criterion (one-tenth of incapacitation) was applied for calculating the safety
index. Critical times for unsafe conditions after which evacuees could not use
lifts for emergency evacuation were determined. The influences of stack
effect and wind speed were also analysed. This study provided useful
information in the identification of safe lift evacuation systems. The use of
lifts with a protected or double protected lobby gave positive safety indices
whereas those with an unprotected lobby did not give a positive safety index,
indicating the greater safety of the first two options.
• Probabilistic risk assessment models were proposed for evaluating risks
associated with water spread. The development of these models was based on
the traditional probabilistic event tree method (a complex parallel and series
combination). After quantification of water spread, probabilities of water
spread were derived. It was determined that a copious amount of water
generated from fire fighting measures could damage the lift components. The
maximum quantity of water spread occurred from fire brigade intervention.
Unprotected lift lobbies are more likely to suffer from water damage whereas
protected lift lobbies can provide a barrier to water spread.
The spread of hot gases to the lift machine room (LMR) via unprotected lift
lobby increases the probability of lift malfunctioning. In unprotected lift
lobbies, temperature rise in LMR depends upon the location of fire. Protected
lift lobbies provide adequate safety against temperature rise in LMR.
247
Probabilistic risk assessment models were also proposed for evaluating risks
associated with electric power systems. The development of this method was
based on a traditional fault tree analysis. After quantification of electric
power failure issues for the hypothetical building, probabilities of risks were
derived. The reliabilities of lift design options were obtained. It was
determined that the probability of electric power failure due to fire in a
residential unit is small. Fire can disrupt electrical system if fire and hot gases
spread to unprotected lift lobbies. Electrical system adhering with relevant
codes and standards in protected lift lobbies can provide adequate safety.
• This research has introduced the concept of Decision Analysis for risk
assessment in the lift evacuation system. An integrated risk assessment
approach was developed to include the uncertainties associated with issues
relating to human behavioural response, fire hazards and lift operational
mechanism. This development was based on the ‘Multi-Objectives Decision
Analysis’ approach for the assessment of results achieved from stochastic
evacuation models, deterministic fire models, probabilistic event tree and
fault tree analyses. This study provided useful information for identifying the
overall safe evacuation system.
• The research provided a good background to the feasibility and design
considerations and evacuation strategies for safe and efficient lift operations.
Design and technical redundancy measures such as refuge area, scatter design
of lifts, pressurisation, gasket between lift landing door and wall frame were
proposed. An overview of fire protection systems for the lift system was also
presented.
9.2 Research Findings
This research has concluded that there are two options if lifts are considered for
emergency evacuation in apartment buildings. Lifts with protected lobby can be used
in the evacuation of one-fourth of the building population without reducing the level
of fire safety. For considering the evacuation of the entire population, lift system
needs special design considerations such as double door protection (or lifts protected
248
with double lobby). The performance based lift evacuation system is achievable. A
brief of advancements and contributions made to the field of research are given next.
9.2.1 Advancements in Systematic and In-Depth Risk Analysis
This research work presents a distinctively different approach over traditional
methods to risk assessment by:
(i) Incorporating related social, physical, temporal and a priori heuristics of
the lift domain for getting insight into lift evacuation system (with
downward, upward and inter-floor movement); and
(ii) Integration of issues relating to human behavioural response, fire hazards
and lift operational mechanism for risk assessment and reliability; and
(iii) Effective quantification of system uncertainties using statistical and
probabilistic techniques. Specifically, the proposed methods (including
MODA and AHP) could reasonably advance methodologies of risk
analysis and assessment for effectively addressing critical issues of
emergency lift operations.
Animation models based on the proposed methods are developed for resolving
obstacles before any lift evacuation plan becomes reality.
9.2.2 Contribution to Building Evacuation Strategy
This research work is an extension of earlier research work contributed by Klote
(1982). A framework for resolving the issues of lift system has been developed. The
concept of MODA is used for the first time in the field of fire safety in this research
project.
This research has resolved that the risk of protected lift lobby is within the acceptable
limit for evacuating a reduced population. Therefore the protected lift system is
recommended with the intent of providing an alternative evacuation facility for the
aged and disabled persons. It was determined that partial evacuation is possible with
the existing infrastructure of protected lift systems. With the current deemed-to-
249
satisfy infrastructure in the apartment buildings, the lift system can be used as a
solution alternative to stairs.
9.3 Scope for Future Work
Many variables have been considered to evaluate the performance of the lift
evacuation system during fire emergencies. These variables were subjected to
uncertainties. To obtain the best results, data were collected to reduce the uncertainty
in relevant variables (for example, reliability of fire protective measures, walking
speed of occupants, coping and response of occupants). These data were simplified
into statistical frequency distribution for the variables. Some factors are still not
possible to describe numerically and, therefore, have to be treated by other means for
example by judgment (such as relationship between the human behavioural response
issues – decision uncertainty and panic in Chapter 3). Therefore the results presented
in this thesis showed the general trend in the safety level. Future work should be
focused to establish a relationship of the human behavioural issues. More surveys are
required to be conducted to evaluate the response of residents toward the lift
evacuation system. The parameters relating to building features of fire safety and fire
protection systems and the level of the fire-affected floor can also be added in the
risk assessment using the Multi-Objectives Decision Analysis approach.
There are certain limitations in the computer model (FDS) used in this research. The
wind speed could not be increased above 7.43 m/s (25 kmph). This is because the
building size is large, which requires a multi-grid system. The information from one
mesh is not properly interpreted at the exterior boundary of a given mesh. This
causes a numerical instability during wind conditions and therefore wind speeds
higher than 7.43 m/s could not be modelled. A rational approach of single-grid
system is needed to study the parameter of higher wind speeds in detail.
This research has attempted to develop a rational procedure for the evaluation of lift
systems as an emergency evacuation facility in high-rise apartment buildings. The
efficacy of the proposed lifts-stairs evacuation strategy (such as 25% of the
population using lifts and for the rest using stairs) has to be evaluated in apartment
buildings during fire drills. Lifts protected with double lobby can be used for the
250
entire population. Double protection of lift shaft can be compared jointly with single
protection of lift shaft (lift protected with a lobby) and lift shaft pressurisation. In
regard to lift shaft pressurisation, smoke does not filter in the FDS model due to
positive pressure in the lift shaft. Therefore, tests relating to the efficacy of lift shaft
pressurisation and leakage areas should be conducted in apartment buildings in the
future.
The apartment buildings are provided with the simple lift group controller system.
The research project begins with the simple lift group controller system and can be
extended to the advanced lift group controller system to achieve a better performance
of the lift system. Further, the cost effectiveness analysis can also be conducted for
various designs of lift evacuation strategies.
The work relating to the use of lifts during fire emergencies is still in its initial stage.
This research concentrated on the use of lifts as an alternative evacuation facility in
apartment buildings. It can be extended in the future to other types of high-rise
buildings such as hotel, hospital and office buildings.
“It can be envisaged that the use of lifts will be inevitable for super
high-rise structures”
251
REFERENCES
ABCB (2005), “The Building Code of Australia – Volume 1”, 2005, Australian
Building Codes Board, Canberra, Australia
ABCB (2005b), “International Fire Engineering Guidelines”, Edition 2005,
Australian Building Codes Board, Canberra, Australia
ABS (2004), “Housing and Lifestyle: High Rise Living”, Australian Bureau of
Statistics, Report No. 4102.0
ABS (2005), “Australian Social Trends - Supply of Housing”, Australian Bureau of
Statistics, Report No. 4102.0
AFAC (1997), “Fire Brigade Intervention Model”, Australasian Fire Authorities
Council, Version 2.1, November, Australia
Aherns, M. (1998), “Batteries Not Included”, NFPA Journal, May/ June, pp. 99-109
Alexandris, N.A. (1977), “Statistical Models in Lift System”, Ph.D. Thesis,
University of Manchester Institute of Science and Technology, pp. 252
Allen, T. H. (1995), “Elevator Evacuation”, Building Performance Research
Institute, USA
Aloi, S. and Rogers, J. (2002), “Reach of the Sky”, Fire Prevention & Fire Engineers
Journal, Volume 62, No. 219, FPA, London, UK
Andersson, C., and Wadensten, K. (2000), “Elevator Evacuation of a High-rise
Building”, Ove Arup and Partners
AS 1735.1 (2003), “Lifts, Escalators and Moving Walks – General requirements”,
Australian Standards, Sydney, Australia
AS 1221 (1997), “Fire Hose Reels”, Australian Standards, Sydney, Australia
ASME A17.1 (2000), “Safety Code for Elevators & Escalators Handbook”, ASME,
USA
Barnett, C., Sfintesco, D., Charles Scawthorn, C. and Zicherman, J. (1992), “Fire
Safety in Tall Building”, Council on Tall Buildings and Urban Habitat
Committee 8A, pp. 3-8
Baum, H.R., and MaCaffery, B.J. (1989), “Fire Induced Flow Field – Theory and
Experiment”, Fire Safety Science – Proceedings of the Second International
Symposium, pp. 129-148
252
Beever, P., and Davy, B. (1999), “A Rational Approach to Firefighting Water
Supplies”, AFAC Conference, Australia
Bell, J. R. (1983), “Five Die in Rockefeller Park Towers Fire, Cleveland Ohio”, Fire
Journal, 77, March, pp. 63-65
Bennetts, I.D., Moinuddin, K.A.M., Goh, C.C., and Thomas, I.R. (1999), “Testing
and Factors Relevant to the Evaluation of the Structural Adequacy of Steel
Members within Fire-resistant Elevator Shafts”, Centre for Environmental
Safety and Risk Engineering, Victoria University of Technology, Melbourne
City MC, Australia
Beyler, C.L. (2004), “Overview of Tenability Criteria and Analysis Methods”, NFPA
Fall Technical Meeting, NFPA, USA
Bill, M. (2002), “World Trade Center Disaster: Initial Response”, Fire Engineering,
UK, September
Bounagui, A., Benichou, N., McCartney, C. and Kashef, A. (2004), “Grid
Optimization for the Full-Scale Test Facility to Evaluate the Fire Performance
fo Houses – Part I – Basement Fires”, Institute for Research in Construction,
National Research Council Canada, Ottawa, Canada
Brennan, P. (1999), “Successful Evacuation in Smoke: Good Luck, Good Health or
Good Management?” Interflam 1999, Proceedings of the 8th International Fire
Science and Engineering Conference, pp. 697-706
Brennan, P., and Thomas, I. (2001), “Victims of Fire, Predicting Outcomes in
Residential Fires”, In Human Behaviour in Fire: Proceedings of the 2nd
International Symposium, Boston. London: Inter Science Communications
Bryan, J.L. (1982), “Human behavior in the MGM Grand Hotel fire”, Fire Journal,
Vol. 76, 2 March, pp. 37-48
Bryan, J.L. (1983), “An Examination and Analysis of the Dynamics of the Human
Behaviour in the Westchase Hilton Hotel Fire”, National Fire Protection
Association, Quincy, USA
Buchanan, A.H. (2001), “Fire Engineering Design Guide”, 2nd
Edition, University of
Canterbury, New Zealand
Building Fire Safety Regulation (1991), “Building Act 1975, Fire and Rescue
Service Act 1990”, Queensland, Australia
Bukowski, R. W., Budnick, E. K. and Schemel, C. F. (1999), “Estimates of the
Operational Reliability of Fire Protection Systems”, International Conference
253
on Fire Research and Engineering (ICFRE3), 3rd Proceedings, SFPE, NIST
and IAFSS, October 4-8, 1999, Chicago, pp. 87-89
Butchen, K. (1995), “Relationship for Smoke Control Calculations: Technical
Memoranda”, TM 19, The Chartered Institution of Building Services
Engineers, USA
Canter, D., (1990), “Fire & Human Behaviour”, 2nd Edition, David Fulton Publishers
Ltd, UK, pp. 66
Caldwell, J. (1996), “Standby Power and Elevator Safety in Fire Emergencies”,
Elevator World Journal, on web
(http://www.elevator-world.com/magazine/archive01/9602-002.htm)
CEN (2003), “Safety Rules for the construction and Installation of Lifts - Part 72:
Fire-fighter Lifts”, CEN/TC10, Committee for European Standardization,
Brussels
Clemen, R.T. (1996), “Making Hard Decisions: An Introduction to Decision
Analysis”, 2nd Edition, Belmond, California – Duxbury Press
Cornell, C.A. (1969), “A Probability-Based Structural Code”, ACI Journal, Volume
66(12), pp. 974-985
Cortes, P., Larraneta, J., and Onieva, L. (2004), “Genetic Algorithm for Controllers
in Elevator Groups: Analysis and Simulation during Lunch Peak Traffic”,
Science Direct, Applied Soft Computing 4, pp. 159-174
Dalgliesh, W.A. and Boyd, D.W. (1962), “Wind on Buildings”, CBD-28, Canadian
Building Digest, Institute of Research in Construction, National Research
Council, Canada
Davis, S. (2000), “Fire Fighting Water: A Review of Fire Fighting Requirements – A
New Zealand Perspective”, Fire Engineering Research Report, ISSN 1173-
5996, University of Canterbury, New Zealand
Degenkolb, J. (1991), “Elevator Use During a Building Fire”, ASME Symposium on
Fire and Elevators, Baltimore, MD, pp. 76-79
Donegan, H.A. (2002), “Decision Analysis”, The SFPE Handbook of Fire Protection
Engineering, 3rd Edition, pp. 5:8-23
Duncanson, M., Woodward, A., Reid, P., and Langley, J. (2000), “Unintentional
House Fire Deaths in New Zealand 1991-1998”, New Zealand Public Health
Report, Vol. 7 No. 7/8, July/August 2000, pp. 31-34
254
Dunlop, K.E., Shields, T.J. and Silcock, G.W.H. (1996), “Toward the Quantification
of Emergency Egress Capabilities for Disabled People”, Fire Engineering and
Emergency Planning, E & FN Spon Publisher, pp. 154-161
Dunstan, K. and Thomson, N. (2006), “Demographic Aspects of New Zealand’s
Ageing Population”, Statistics New Zealand, Ministry of Social Development,
New Zealand pp. 1-44
Edwards, W. (1971), “Social Utilities”, Decision and Risk Analysis Powerful New
Tools for Management, The Engineering Economist – Summer Symposium
Series VI, American Society for Engineering Education, pp.119-129
Emporis (2005), “Emporis Buildings”, on web
(http://www.emporis.com/en/wm/bu/?id=101921)
FCRC (1996), “Fire Engineering Guidelines”, First Edition, Fire Code Reform
Centre, 1st Edition, Australia
Frantzich, H. (1997a), “Fire Safety Risk Analysis of a Health Care Facility”, Report
3085, Department of Fire Safety Engineering, Lund Institute of Technology,
Lund University, pp. 1-68
Frantzich, H. (1997b), “Fire Safety Risk Analysis of a Hotel”, Department of Fire
Safety Engineering, Lund Institute of Technology, Lund University, pp. 1-28
Frantzich, H. (1997c), “Design Based on Calculated Risk”, on web
(http://www.brand./th.se/englsh/research/risk.pdf)
Frantzich, H., Magnusson, S.E., Holmquist, B., and Ryden, J. (1997), “Derivation of
Partial Safety Factors for Fire Safety Evaluation Using the Reliability Index β
Method”, Fire Safety Science – Proceeding of the Fifth International
Symposium, Melbourne, Australia, pp. 667-678
Groner, N. (2002), “A compelling case for emergency elevator system”, Fire
Engineering, 155:10, October, pp. 126-128
Hakonen, H. (2003), “Simulation of Building Traffic and Evacuation by Elevators”,
Licentiate Thesis, Helsinki University of Technology, Department of
Engineering Physic and Mathematics, System Analysis Laboratory, Finland
Hall, J.R. (2001), “High-Rise Building Fires”, NFPA Fire Analysis and Research
Division, USA
Hartzell, G.E., Priest, D.N., Switzer, W.G. (1985), “Modeling of Toxicological
Effects of Fire Gases: II. Mathematical modeling of intoxication of rats by
carbon monoxide and hydrogen cyanide”, Fire Science, 3: pp. 115-28
255
Hartzell, G.E. (2001), “Engineering Analysis of Hazards to Life Safety in Fires: the
Fire Effluent Toxicity Component”, Safety Science, 38: pp. 147-155
Hasofer, A.M. and Lind, N.C. (1974), “An Exact and Invariant First Order
Reliability Format”, ASCE, J. Eng. Mech. Division
He, Y., Horason, M., Taylor, P., and Ramsay, C. and Lai, D. (2001), “A Probabilistic
Fire Safety Engineering Assessment of a Refurbished High Rise Office
Building”, Proceedings of the International Conference on Engineered Fire
Protection Design, Society of Fire Protection Engineers, San Francisco, pp.
211-228
He, Y., Horason, M., Taylor, P., and Ramsay, C. (2003), “Stochastic Modelling for
Risk Assessment”, Fire Safety Science – Proceedings of the Seventh
International Symposium, International Association for Fire Safety Science
(IAFSS), Worcester, MA, USA, Intl. Assoc. for Fire Safety Science, Boston,
MA, Evans, D. D., Editor, pp. 333-344
Honeybrook, J. (2002), Superintendent, NSWFB – slide presentation, on web
(http://sfs.au.com/Documents/NSW%20FB%20LIfts%20PP.pdf)
Howard, R.A. (1988), “Decision Analysis: Practice and Promise”, Management
Science, pp. 679-695
HMSO (1993), “Manual of Firemanship – Book 12”, Practical Firemanship II,
HMSO, London, pp. 25-50
Hoeksema, R.J. & Kitanidis, P.K. (1985), “Analysis of the spatial structure of
properties of selected aquifers”, Water Resources Research, 21(4), pp. 563-572
Hostikka, S. and Rahkonen O. K. (2003), “Probabilistic Simulation of Fire
Scenarios”, Science Direct, Nuclear Engineering and Design 224, pp. 301-311
Hultquist, H., and Karlsson, B. (2000), “Evaluation of a Fire Risk Index Method for
Multi-storey Apartment Buildings”, Department of Fire Safety Engineering
Lund University, Sweden
ISO TR 16765 (2003), “Comparison of Worldwide lifts (elevators) safety standards –
Fire fighters lifts (elevators)”, International Standardisation Organisation
ISO TS 13571 (2002), “Life-threatening components of fire — Guidelines for the
estimation of time available for escape using fire data”, International
Standardisation Organisation
Jin, T. and Yamada, T. (1985), “Irritating Effects of Fire Smoke on Visibility”, Fire
Science & Technology, Vol.5, No.1, pp. 79-89
256
Jin, T. and Yamada, T. (1989), “Experimental Study of Human Behaviour in Smoke
Filled Corridors”, Fire Safety Science – Proceeding of the Second International
Symposium, Hemisphere Publishing Corp., Washington, pp. 511-519
Jin, T. (2002), “Visibility and Human Behaviour in Fire Smoke”, The SFPE
Handbook of Fire Protection Engineering, Third Edition, pp. 2-42
Juillet, E. (1993), “Evacuating People with Disabilities”, Fire Engineering,
November, Vol. 126, No. 12, pp. 100-103
Karlsson, B. (2000), “Fire Risk Index Method for Multi-storey Apartment
Buildings”, Trätek AB, Report I0009025, Stockholm, Sweden
Keefer, D. L., Kirkwood, C.W. and Corner, J.L. (2002), “Summary of Decision
Analysis Applications in the Operations Research Literature, 1990-2001”,
Technical Report, Department of Supply Chain Management, Arizona State
University, USA
King, M. (1997), “Australian National Fire Statistics 1993-1994”, Technical Report
TR 97/2, CSIRO, Australia
Kisko, T.M, Francis, R.L. and Nobel, C.R. (1998), “EVACNET 4 User’s Guide”,
University of Florida, USA
Klote, J.H. (1982), “Elevators as a Means of Fire Escape”, American Society of
Heating Refrigerating and Air Conditioning Engineers Transactions, NBSIR
82-2507; pp. 1-40
Klote, J.H., and Tamura, G.T. (1986), “Elevator Piston Effect and the Smoke
Problem’’, Fire Safety Journal, Vol. 11, No. 3, pp. 227-233
Klote, J.H., Alvord, D.M., and Deal, S. (1991), “Routine for Analysis of the People
Movement Time for Elevator Evacuation - ELVAC”, NISTIR 4730, USA
Klote, J.H. and Milke, J.A. (1992), “Design of Smoke Management System”,
ASHRAE and SFPE
Klote, J. H., Deal, S. P., Levin, B. M., Groner, N. E., and Donoghue, E. A. (1993a),
“Workshop on Elevator use During Fires”, NIST, NISTR 4993
Klote, J.H., Deal, S.P., Donoghue, E.A., Levin, B.M., and Groner, N.E. (1993b),
“Fire Evacuation by Elevators”, Elevator World, June
Klote, J. H., Levin, B. M. and Groner, N. E. (1995), “Emergency Elevator
Evacuation Systems”, Proc. of ASME 2nd Symposium on Elevators, Fire and
Accessibility, Baltimore, MD, Feb 19-21, 1995, ASME, New York, NY, pp.
131-149
257
Klote, J. H. (2002), “Smoke Control”, The SFPE Handbook of Fire Protection
Engineering, 3rd Edition, pp. 274-291
Klote, J.H. (2003), “Analysis of the Life Safety Consequences of Smoke Migration
through Elevator Shafts”, American Society of Mechanical Engineers, ASME,
pp. 1-20
Koshak, J. (2003), “Elevator Evacuation in Emergency Situations”, ASME, pp. 1-6
Kuligowski, E. (2003), “Elevators for Occupant Evacuation and Fire Department
Access”, Proceedings of the CIB-CTBUH International Conference on Tall
Building, Malaysia
Kuligowski, E. and Bukowski, R.W. (2004), “Design of Occupants Egress Systems
for Tall Buildings”, National Institute of Standards and Technology, CIB, USA,
pp. 1-10
Lacey, R. (2000), “Lift Management in Emergencies”, Elevator World, May 2000,
pp. 126
Larsson, D. (2000), “Developing the Structure of a Fire Risk Index Method for
Multi-storey Apartment Buildings”, Department of Fire Safety Engineering”,
Lund University, Report 5062, Lund, Sweden
Lee, W. B., Lau, H., Liu, Z, and Tam, S. (2001), “A Fuzzy Analytical Hierarchy
Process Approach in Modular Product Design”, Expert System, Vol.18 No.1,
pp. 32-42
Levin, B.C., Paabo, M., Gurman, J.J., and Harris, S.E. (1987), “Effects of Exposure
to Single or Multiple Combinations of the Predominant Toxic Gases and Low
Oxygen Atmospheres Produced in Fires”, Fundamental and Applied
Toxicology, Vol. 9, pp. 236-250
Magnusson, S.E. (1997), “Risk Assessment”, Fire Safety Science – Proceeding of the
Fifth International Symposium, Melbourne, Australia, pp. 41-58
Magnusson, S. E. (1996), “How to Derive Safety Factors”, International Conference
on Performance-Based Codes and Fire Safety Design Methods, Ottawa, Canada
Marchitto, N. (1991), “High Temperature Operation of Elevators”, Elevators and
Fire, pp. 25-31
Marryatt, H.W. (1988), “Fire - A Century of Automatic Sprinkler Protection
Australia & New Zealand 1886-1986”, Australian Fire Protection Association,
Melbourne, Australia
258
Martin, M. (2003), “Challenges to Using Elevators in a Fire”, American Society of
Mechanical Engineers, USA, pp. 1-5
McGrattan, K.B., Forney, G.P., Floyd, J.E., Hostikka, S. and Prasad, K. (2004), “Fire
Dynamics Simulator - Version 4.07”, Fire Research Division, Building and Fire
Research laboratory, National Institute of Standards and Technology, USA
Meacham, B.J. (2002), “Building Fire Risk Analysis”, The SFPE Handbook of Fire
Protection Engineering, 3rd Edition, pp. 5:162
Melinek, S.J. and Booth, S. (1975), “An Analysis of Evacuation Times and the
Movement of Crowds in Buildings”, Borehamwood, Building Research
Establishment, Fire Research Station, UK
Miller, I. (2005), “Human Behaviour Contributing to Unintentional Residential Fire
Deaths 1997 -2003”, New Zealand Fire Service Commission Research Report
Number 47, New Zealand Fire Service Commission
Modarres, M., and Billoch, F.J. (2002), “Reliability”, The SFPE Handbook of Fire
Protection Engineering, 3rd Edition, pp. 5-34
Mori, B.C., and UHK (2002), “A Survey on Residents’ Responses to High-Rise
Living in Hong Kong”, Mori Building Co. Ltd and University of Hong Kong,
Hong Kong
NBS (2000), “The Building Regulations 2000, Approved Document B”, ISBN-
101859462014, Office of the Deputy Prime Minister, NBS Publications, UK
Nelson, H.E., and MacLennan, H.A. (1995), “Emergency Movement”, The SFPE
Handbook of Fire Protection Engineering, 2nd Edition, pp. 3:286-295
NFPA (1982), “Fire at the MGM Grand”, Fire Journal, National Fire Protection
Association (NFPA), pp. 19-37
NFPA (1998), “Fire Protection Handbook”, 18th Edition, Table 8-1P, pp. 8-17
NFPA 13 (1996), “Standard for the Installation of Sprinkler Systems”, NFPA,
Quincy, MA
NFPA (1999), “Special Data Information Package, High-rise Fatal Fires – Apartment
Buildings”, NFPA, Quincy, MA
NFPA 101 (2000), “Life Safety Code”, Volume 5, NFPA, Quincy, MA
NFPA 72 (2002), “National Fire Alarm Code”, NFPA, Quincy, MA
NFPA (2006), “Leading Causes of Structure Fires in Homes, 1999-2002 Annual
Averages”, NFPA, Fire Statistics – Causes, USA
259
NIST Special Publication 983 (2003), “Collected Publications Related to the Use of
Elevators during Fires”, Building and Fire Research Laboratory, NIST, USA
NUREG/CR-1278 (1980), “Handbook of Human Reliability Analysis with Emphasis
on Nuclear Power Plant Applications”, U.S. Atomic Energy Commission,
Washington, D C, USA
NZFS incident statistics, 1993-1997, “Table 8: All Fires – How They were
Extinguished”, New Zealand Fire Service, New Zealand
O’Donoghue, J.J. and O’Donnel J.K. (2003), “Why the Fire Service Must Maintain
Control of Elevators during a High-Rise Fire/ Emergency”, ASME, USA
Palisade Corp (1996), “@RISK 4.5 Software”, USA
Pauls, J.L. (1977), “Human Response to Tall Building”, Dowden, Hutchinson and
Ross, Stroudsburg, PA, USA
Pauls, J.L., Gatfield, A. J. and Juillet, E. (1991), “Elevator Use and for Egress: The
Human Factors Problems and Prospects”, ASME Symposium on Fire and
Elevators, Baltimore, MD, Feb 19-20, ASME, New York, USA, pp. 63-75
Pauls, J. (1995), “Movement of People”, The SFPE Handbook of Fire Protection
Engineering, Second Edition, pp. 3-278
Peakcock, R.D., Jones, W.W., Bukowski, R. W., and Forney, C.L. (1991),
“Technical Reference Guide for the HAZARD I Fire Hazard Assessment
Method”, Version 1.1, NIST Handbook 146, Volume II, NIST, MD, USA
Peakcock, R.D., Forney, G.P., Reneke, P., Portier, R., and Jones, W.W. (1993),
“CFAST, the Consolidated Model of Fire and Smoke Transport”, NIST
Technical Note 1299, NIST, MD, USA
Pfeiffer, D.U. (1997), “Decision Analysis and Risk Analysis”, Risk Analysis and
Animal Health Course Manual, International Training Course, Diibendorf,
Switzerland, pp. 861-877
Phillips, W.G.B. (1995), “Computer Simulation for Fire Protection Engineering”,
The SFPE Handbook of Fire Protection Engineering, Second Edition, pp. 5-7
Proulx, G. (1993), “A Stress Model for People Facing a Fire”, Journal of
Environmental Psychology, 13, pp. 137-147
Proulx, G. (1995), “Evacuation Timing in Apartment Buildings”, Fire Safety Journal,
Vol. 24, pp. 229-246
Proulx, G. (2000), “Is it Wise to Evacuate during a High-Rise Fire?”, Presentation,
NFPA Conference, Denver, USA.
260
Proulx, G. (2003), “Human Behaviour Study: Cook County Administration Building
Fire October 17, 2003, Chicago”, Research Report 181, Institute for Research
in Construction, National Research Council, Ottawa, Canada.
Purser, D.A. (2002), “Toxicity Assessment of Combustion Products”, The SFPE
Handbook of Fire Protection Engineering, 3rd Edition, pp. 2:83-171
Ramachandran, G. (1991), “Human Behaviour in Fires-A Review of Research in the
United Kingdom”, Fire Technology, 47, 1, pp. 66-81
Ramachandran, G. (2002), “Stochastic Models of Fire Growth”, The SFPE
Handbook of Fire Protection Engineering, 3rd Edition, pp. 381-401
Reber, A.S. (1995), “The Penguin Dictionary of Psychology”, Second Edition,
Penguin Books Ltd, London, UK
Robibero, V.P. (1991), “High Temperature Operation of Elevators”, Elevators and
Fire, pp. 37-43
Rockwell (2000), “ARENA 5.0 Version”, Rockwell Software Inc., USA
Rohr, K. (2001), “U.S. Experience with Sprinklers”, NFPA, September, on web
(http://www.afsaflorida.org/os-sprinkler01.pdf)
Roytman, M.Y. (1969), “Principles of Fire Safety Standards for Building
Construction”, Construction Literature Publishing House, Moscow, English
translation (TT 71-580002) from National Technical Information Service
Saaty, T.L. (1977), “A Scaling Method for Priorities in Hierarchical Structures”,
Journal of Mathematical Psychology, Vol. 15, No.3, pp. 234-281
Saaty, T.L. (1980), “The Analytic Hierarchy Process”, McGraw-Hill, New York,
USA.
Schroeder, J. (1986), “The Energy Consumption of Elevators”, Elevator Technology,
(Editor Dr. G. Barney), Ellis Horwood
Schultz, D.P. (1968), “Contract Report NR 170-274”, University of North Carolina,
Charlotte, USA
Sekizawa, A. I. (1991), “Statistical Analyses on Fatalities Characteristics of
Residential Fires”, Fire Safety Science – Proceedings of the 3rd International
Symposium (Editors: Cox, G. and Langford), B., Edinburgh, UK, pp. 475-484
Sekizawa, A., Nakahama, S., Notake, H., Ebihara, M., and Ikehata Y. (1996), “Study
on Feasibility of Evacuation using Elevators in a High-rise Building”,
University of Tokyo, Taisei Corporation and Shimizu Corporation, Japan
261
Shestopal, V.O. (1993), “Computer Program for an Uninhibited Smoke Plume and
Associated Computer Software”, Fire Technology, Vol. 29, No. 3, pp. 246-267
Shields, T.J., and Silcock, G.W.H. (1987), “Means of Escape”, Buildings and Fire,
Longman Scientific & Technical, 397
Siikonen, M.L. (1989), “Hissiliikenteen Ja Ohjaukesen Mallintaminen
Tietokoneella”, Licentiate Thesis, Helsinki University of Technology,
Department of Engineering Physic and Mathematics, System Analysis
Laboratory, Finland
Siikonen, M.L. (1997), “Planning and Control Models for Elevators in High-Rise
Buildings”, Research Report (KONE Corporation), Helsinki University of
Technology, System Analysis Laboratory, Finland
Siikonen, M.L., and Hakonen, H. (2003), “Efficient Evacuation Methods in Tall
Buildings”, Elevator World, July, pp. 78-83
Sime, J.D. (1983), “Affiliative Behaviour during Escape to Building Exits”, Journal
of Environmental Psychology, Vol. 3 No. 1, pp. 21-41
Smith, D. (2003), “Lift could be used for Evacuation during Emergencies! Fact or
Fiction?”, ASME, pp. 1-20
So, A., and Yu, J. (2003), “Lift Logic”, Fire Prevention – Fire Engineers Journal,
August Issue, Gloucestershire, UK, pp. 26-28
Straskosch, G.R. (1998), “The Vertical Transportation Handbook”, Third Edition,
ISBN: 978-0471162919, John Viley & Sons, Inc, NJ, USA
Tamura, G.T. and Shaw, C.Y. (1976), “Air Leakage Data for the Design of Elevator
and Stair Shaft Pressurisation System”, ASHRAE Transactions, Vol. 82 (Part
2), pp. 179-190
Tamura, G. (1994), “Smoke Movement and Control in High-Rise Buildings”, NFPA,
USA
Taylor, R. E. (1975) “The Carlyle Apartment Fire: Study of a Pressurised Corridor”,
ASHRAE Journal, Vol. 17, pp. 52-55
Teo, H.P. (2001), Key Note Address, The International Association of Elevator
Engineers in Elevator Technology 11, Proceedings of Elevcon 2001, the 11th
International Congress on Vertical Transportation Technologies, Singapore
Proceedings of ELEVCON 2001, Singapore, July 2001
262
Tyni, T. and Ylinen, J. (2006), “Evolutionary Bi-Objective Optimisation in the
Elevator Car Routing Problem”, European Journal of Operational Research
169(3), pp. 960-977
Vistnes, J., Grubits, S. J., and He, Y. (2005), “A Stochastic Approach to Occupant
Pre-Movement in Fires”, Proceedings of the 8th International Symposium on
Fire Safety Science, pp. 531-542
Vrijling, J.K., Van Hengel, W., and Houben, R.J. (1995), “A Framework for Risk
Evaluation”, Journal of Hazardous Materials, 43, pp. 245–261
Walton, G. N. (1993) “CONTAMW 1.0 User Manual”, Multi-zone Airflow and
Contaminant Transport Analysis Software, NISTIR 6476, NIST
WASH 1400 (1974), “Reactor Safety Study, An Assessment of Accident Risks in
U.S. Commercial Nuclear Power Plants, Appendix III, Failure Data”, U.S.
Atomic Energy Commission, Washington, D.C.
Watts, J.M. Jr. (1991), “Criteria for Fire Risk Ranking”, Fire Safety Science,
Proceedings of the Third International Symposium, pp. 457-466
Watts, J.M. Jr. (2002), “Fire Risk Ranking”, The SFPE Handbook of Fire Protection
Engineering, 3rd Edition, pp. 5:125-142
Watts, J.M. Jr. and Hall, J. R. (2002), “Introduction to Fire Risk Analysis”, The
SFPE Handbook of Fire Protection Engineering, 3rd Edition, pp. 5:1-7
WHI (1987), “Test of Gasketing for Resistance to Air Flow at Elevated Temperature
to Determine Effectiveness as a Smoke Control Door Seal”, WHI Test Report
# 630-611300, Warnock Hersey International, Inc.
Willey, A. (1973), “Fire Baptist Towers Housing For the Elderly”, Fire Journal, May
1973, NFPA, pp. 28-36
Zhao, L. (1998), “Technical Report FCRC-TR 98-05”, Project 4, Part A
263
Appendix A
International Listing of Major Fires, where Lifts were used
The National Fire Protection Association (NFPA) has maintained an international
listing of high-rise building fires since 1911 (a few incidents reproduced in book
‘Fire Safety in Tall Building’ published by Council on Tall Buildings and Urban
Habitat Committee 8A). Some of the major incidents, where occupants tried to use
lift for evacuation, are given in Table A1:
Table A1: Use of Lifts during Major Building Fires
Date Building Occupancy Number of
storeys
(Origin of
fire)
Number
of
deaths
A brief of lift use
24.01.1969 Hawthorne House, Chicago
Apartment 39 (36) 4 Use of lifts by residents for building evacuation, which causes delay in fire fighting operation
5.08.1970 One New York Plaza
Office 50 (33) 2 Deaths from lift stopping at fire floor
4.12.1970 Office building NY
Office 47 (5) 3 Deaths from lift stopping at fire floor
2.07.1971 Motor hotel complex NY
Hotel 17 (12) 6 Victims mistakenly tried lifts versus stairs
1.02.1974 Joelana building Brazil
25 (12) 179 Lifts used heavily for rescue
12.11.1974 Century City LA
Office 15 (8) - Lifts used for evacuation
23.06.1980 Westvaco building NY
Office 42 (20) - Elevator malfunction (a severe problem)
21.11.1980 MGM Grand Hotel LA
Hotel 23 (1) 85 Smoke spread through air system, stairs and lift hoist way. Death from smoke in rooms, corridor and lifts.
10.02.1981 Hilton Hotel LA
Hotel 30 (8) 8 Fire rated lift vestibules. Fire started in lift lobby (arson) and spread to exterior
31.12.1986 Dupont Plaza Hotel Puerto Rico
Hotel 20 (1) 97 Smoke spread through lifts, air system, utility passage ways, stairs and building exterior.
24.12.1989 John Sevier Retirement Centre
Old age residential
11 (1) 16 Typifies evacuation problems in elderly housing
Note: In spite of notices not to use lifts, lifts were used for evacuation. Some of them
lead to a disastrous end.
264
Appendix B
Risk Priorities and Matrix Consistency Ratio
Calculation of Risk Priorities (3 × 3 Matrix)
Decision Uncertainty Panic Nonfatal Injuries Average Risk
Priorities
Decision
Uncertainty 1
0667.0951
1=
++
1/5
0625.021
51
51
=
++
1/9
0683.01
21
91
91
=
++
3
0683.00625.00667.0 ++ 0.066
Panic 5
3333.0951
5=
++
1
3125.021
51
1=
++
½
3105.01
21
91
21
=
++
3
3105.03125.03333.0 ++ 0.319
Nonfatal
Injuries 9
6.0951
9=
++
2
625.021
51
2=
++
1
6211.01
21
91
1=
++
3
6211.0625.06.0 ++
0.615
265
Calculation of Eigenvalue, Consistency Index and Consistency Ratio
Decision Uncertainty Panic Nonfatal Injuries
Decision Uncertainty 1 – λ 1/5 1/9
Panic 5 1 – λ ½
Nonfatal Injuries 9 2 1 – λ
To determine Eigenvalue λ, cubical equation is resolved as shown next:
[ ] ( ) ( ) ( ) ( ) ( ) ( )[ ]1112/129/19(5/1511123 −−−×+×+×−++− λλλ( ) ( ) ( ) ( ) ( ) ( )[ ] 019/192/15/1929/15155/122/11111 =××−××+××+××−××−××+
( ) ( ) ( )[ ] 010/99/1023 23 =++−−− λλ
( )[ ] 090/13 23 =−− λλ
10.3=λ
The Consistency Index (CI) is calculated as given below:
05.013
310.3
1
max =−−
=−
−=n
nCI
λ
The Consistency Ratio (CR) is calculated as given below (RI is 0.58):
086.058.0
05.0===
RI
CICR
266
Appendix C
Survey Questionnaire
1. Your age ……………
2. Floor level of your flat in the multi-storey building…………….level
3. Do you use the stairs for normal building access?
Always Sometimes Rarely Never
4. Do you use the lift for normal building access?
Always Sometimes Rarely Never
5. Do you use the stairs as a normal building exit?
Always Sometimes Rarely Never
6. Do you use the lift as a normal building exit?
Always Sometimes Rarely Never
7. How would you evacuate the building during a fire emergency?
Stairs only
Lift only
Both stairs and lift
Remain in the building
By fire brigade appliance
8. Have you ever been trained in fire emergency evacuation procedure?
Yes No
9. If you have used the stairs in the fire drill, have you had any difficulty in
using the stairs?
Yes No
10. Any specific information on difficulty encountered during fire drills/ any
other comments
……………………………….………………………….……...
‘THANK YOU FOR YOUR ACTIVE PARTICIPATION IN THE INTEREST
OF PUBLIC SAFETY’
267
Questionnaire Results:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1 Age 26 52 58 47 35 52 28 58 43 55 54 65 25 34 42 25 36 54 42 40 36 56 60 40 57
2 Floor level 1 2 2 3 4 5 6 6 6 10 10 11 11 11 12 13 13 15 15 16 17 18 18 19 19
3 Stair for normal access
Always
Sometimes 1 1 1
Rarely 1 1
Never 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
4 Lift for normal access
Always 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Sometimes 1 1 1
Rarely
Never
5 Stair for exit
Always 1
Sometimes 1 1 1
Rarely 1 1
Never 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
6 Lift for exit
Always 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Sometimes 1 1
Rarely
Never
7 How would you evacuate
Stair only 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Lift only
Both stair and lift 1
Stay in place
By fire brigade
8 Are you trained in EEP
Yes 1 1 1 1 1 1 1
268
No 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
9 Difficulty in stair in EEP
Yes 1 1 1 1
No 1 1 1 1 1 1 1 1 1 1 1 1
10 Any specific information
(stair locked) 1 1
Other details are given on next page
--2--
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
1 Age 30 19 42 30 45 50 35 20 35 69 31 73 26 32 57 66 24 50 44 59 24 58 44 61 24
2 Floor level 20 21 21 21 21 22 23 25 26 26 27 27 28 28 29 29 30 31 31 32 33 33 34 37 38
3 Stair for normal access
Always
Sometimes
Rarely 1 1
Never 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
4 Lift for normal access
Always 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Sometimes
Rarely
Never
5 Stair for exit
Always
Sometimes
Rarely 1 1
Never 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
6 Lift for exit
Always 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Sometimes
Rarely
Never
7 How would you evacuate
269
Stair only 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Lift only
Both stair and lift 1
Stay in place 1
By fire brigade
8 Are you trained in EEP
Yes 1 1 1 1 1 1 1 1 1
No 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
9 Difficulty in stair in EEP
Yes 1 1 1 1 1
No 1 1 1 1 1 1 1 1 1 1 1 1 1
10 Any specific information
(Stair locked) 1 1
Comments from residents are given below:
Fire drills are tiring
Some had knee trouble
Problem with weaker sex for long traveling
More people downward as one travels
Long evacuation time
Lots of stairs to walk
Some have to stay in the building
High level of concern for evacuation due to high altitude
Never had a fire drill
Doors are generally locked – 8% indicated that stair doors were always locked.
270
Calculation of Confidence Interval:
For calculating the confidence interval in Excel, the Beta Distribution is used to
study variation in the percentage of sample across the data. The details for
calculating the confidence interval with the help of BETAINV are given below:
Syntax: BETAINV(p, α, β, A, B)
where
p is a probability associated with the beta distribution
α is a parameter of the distribution
β is a parameter of the distribution
A is an optional lower bound to the interval of x =0
B is an optional upper bound to the interval of x = 1
α = y + 1 and β = n – y + 1
where y is the number of successes in n trials
To obtain the 95% Confidence Interval for p, let x = 0.9275 and 0.0725
There are mean number of 46.375 occupants intend to use lifts in 50 occupants, then:
α = 47.375 and β = 4.625
Using Excel,
BETAINV(0.9275, 47.375, 4.625, 0, 1) = 0.961
BETAINV(0.0725, 47.375,4.625, 0, 1) = 0.849
Hence, a 95% Confidence Interval for the true proportion is 0.849 to 0.961.
271
Appendix D
Interview
Interviews were conducted with the Fire Engineer from NSW Fire Brigade, Sydney
(Person 1), two Community Safety Officers from QFRS, Brisbane (Persons 2 and 3).
These officers have rendered mixed services of regular fire fighting and fire-
prevention. The following questions and answers were covered although these may
not be the same exact words used by the fire personnel:
Questions and Replies – Person 1:
1. Q: Do you notice any problem with the use of stairs during fire emergencies
or drills?
A: As a fire-fighter in the U.S., I often noticed that space was an issue even
when the stairways were only used for fire fighting operations (not enough
space for equipment and hose). The problem is exacerbated due to the
conflicting needs of the evacuees and the fire fighting personnel.
2. Q: How would you recommend means of evacuation for disabled, aged
people, ladies and children from high-rise building during a fire emergency?
A: Refuge areas are separated from the building, which can be used for
temporary staging then use lifts and/or stairs with assistance as needed.
3. Q: During fire emergencies, how is the occupant behaviour – irrational or
rational? What about their priority on upper levels – stairs or lift? What about
visitors in office building?
A: It depends on the scenario and the people involved. I assume that people
would usually avoid lifts if there was evidence of smoke and/or fire.
Anything different in evacuation strategy would need a change in
culture/training/acceptance. If there was minimal evidence of fire and/or
smoke then they might use the lifts, but my concern is that it’s the “big ones”
that result in deaths and/or injury upon which a lot of our building codes are
based.
4. Q: Do you feel that lift should be considered as an option (secondary) for
means of evacuation?
A: Only if the detailed design and commissioning and maintenance and use
can be proven to support such a function. This would also require a change in
occupant acceptance and attitude, since they would need to feel safe in doing
so (in addition to the system working under the conditions it would need to).
One big advantage of stairs is that even if people are waiting, they are
flowing, whereas with lifts they will egress in increments and have not
control over their destiny between lift arrivals (as per your rational behaviour
query it’s my understanding that when people feel they have control over
272
their outcome during tense emergency situations then they will tend to act
rationally but when they feel that the situation is deteriorating more rapidly
than they can react appropriately then they become irrational (like when they
might be waiting for the lift to arrive with rapidly deteriorating environmental
conditions).
5. Q: Do you feel that “fire emergency evacuation drills” are regularly held in
NSW?
A: No idea, I believe they might be in offices but I know when I was
consulting in a high-rise downtown I never did one.
6. Q: In your opinion, what is acceptable limit for comfortable travelling in
building? Whether one can come down in stairs – let say 10 storey or 20
storey or 30 storey or more.
A: If I felt safe in the stair then it wouldn’t matter.
7. Q: How the fire fighters approach to the upper levels? Do they use stairs or
lift?
A: In a working fire environment I would expect them to use the stairs until
they felt the fire floor was safe (that’s my understanding of how they do
things here).
8. Q: How far fire fighter can climb in building with fire fighting gears?
A: I believe as far as they would need to, the time needed might be more of
the issue.
9. Q: Are power supply is reliable in buildings?
A: My opinion would be that unless it is specifically designed as “Emergency
Power” with on-site emergency back up generation then it cannot be assumed
to be. I’ve seen too many blocks in the CBD out of power for hours on end
even though the power has been “deemed” to be reliable enough for back-up
for emergencies (based on girded supply). This is because there are still
single point failure modes and the assessment is not done “in-depth” as would
be needed to address all single point failure modes and effects and providing
solutions to the same.
10. Q: Any specific information, you would like to give.
A: Keep up the good work!
273
Questions and Replies – Person 2:
1. Q: Do you notice any problem with the use of stairs during fire emergencies
or drills?
A: Bottleneck movement, fire fighters have entrance problem in the building
2. Q: How would you recommend means of evacuation for disabled, aged
people, ladies and children from high-rise building during a fire emergency?
A: Such occupants should go next level downward or go to top terrace or stair
landing
3. Q: During fire emergencies, how is the occupant behaviour – irrational or
rational? What about their priority on upper levels – stairs or lift? What about
visitors in office building?
A: Nearly 50% are occupants are rational and remaining 50% are irrational.
Normally occupants try to use lifts in office buildings, whereas occupants try
to use stairs in shopping centre. Visitors would like to go in car park
straightway from the upper levels.
4. Q: Do you feel that lift should be considered as an option (secondary) for
means of evacuation?
A: It can be considered for the disabled, aged and fire fighters. Normally fire
hydrants locations are near to lift lobby, which can be made as a staging area.
5. Q: Do you feel that “fire emergency evacuation drills” are regularly held in
QLD?
A: It is must in commercial buildings, but not in residential buildings.
Building Fire Safety Regulation permits for workers only.
6. Q: In your opinion, what is acceptable limit for comfortable travelling in
building? Whether one can come down in stairs – let say 10 storey or 20
storey or 30 storey or more.
A: At every 4th level, provision is made for re-entry. So occupants could go at
every alternate 4th level.
7. Q: How the fire fighters approach to the upper levels? Do they use stairs or
lift?
A: Whatever is convenient to us.
8. Q: How far fire fighter can climb in building with fire fighting gears?
A: Generally they can climb up to 8th floor.
9. Q: Are power supply is reliable in buildings?
A: It is normally reliable in Brisbane.
10. Q: Any specific information, you would like to give.
A: The provision such as pressurisation (zone smoke management) can be
considered for lift system.
274
Questions and Replies – Person 3:
1. Q: Do you notice any problem with the use of stairs during fire emergencies
or drills?
A: Overcrowding, problematic for disabled occupants, difficulties for staging
area during drills
2. Q: How would you recommend means of evacuation for disabled, aged
people, ladies and children from high-rise building during a fire emergency?
A: Lifts can be explored as an option in high-rise buildings.
3. Q: During fire emergencies, how is the occupant behaviour – irrational or
rational? What about their priority on upper levels – stairs or lift? What about
visitors in office building?
A: Experienced people acts rationally. Generally occupants use stairs for
emergency evacuation. Occupants are given instructions through public
address system, what they have to use for evacuation.
4. Q: Do you feel that lift should be considered as an option (secondary) for
means of evacuation?
A: It must be considered as an option for elderly and disabled occupants
5. Q: Do you feel that “fire emergency evacuation drills” are regularly held in
QLD?
A: Periodically fire drills are conducted in building normally once a year.
6. Q: In your opinion, what is acceptable limit for comfortable travelling in
building? Whether one can come down in stairs – let say 10 storey or 20
storey or 30 storey or more.
A: It depends on the individual strength.
7. Q: How the fire fighters approach to the upper levels? Do they use stairs or
lift?
A: Fire fighters should use lifts for upper level access.
8. Q: How far fire fighter can climb in building with fire fighting gears?
A: Again it depends on the individual strength.
9. Q: Are power supply is reliable in buildings?
A: Yes, they are reliable in Australia.
10. Q: Any specific information, you would like to give.
A: Stack arrangements for emergency evacuation can be researched for
evacuation purpose.
275
Appendix E
Occupant Response and Coping Times
The following response and coping scores for residential occupants are based on
Table 7.9 – Chapter 7 – Fire Engineering Guidelines:
Table E1 – Weighting Factor in Occupant Response Time
Factor Weighting
Alertness 2
Role 1
Commitment 4
Focal Point 1
Mobility 3
Social Affiliation 1
Position 2
Familiarity 4
Primary factors are shown in bold
RESPONSE TIME CALCULATION (a sum of primary factors is multiplied by 2
and secondary factors by 0.4)
((10 x 2) + (8 x 0.4)) / 8 = 2.9;
6 – 2.9 = 3.1 minutes (186 seconds)
Table E2 – Weighting Factor in Occupant Coping Time
Factor Weighting
Role 1
Mobility 2
Social Affiliation 1
Alertness 3
Position 4
Commitment 2
Focal Point 2
Familiarity 4
Primary factors are shown in bold
COPING TIME CALCULATION
((4 x 2) + (15 x 0.4)) / 8 = 1.75;
6 – 1.75 = 4.25 minutes (255 seconds)
• Residential occupant response score = 186 seconds
• Residential occupant coping score = 255 seconds
276
Appendix F
SIMAN Language
SIMAN Language Used in Lift Simulation Model (only a small portion)
Simulation model has generated SIMAN output files. Various modules indicating
their attributes, variables and distribution functions are given below:
; Model statements for module: Create 1 214$ CREATE, 4,SecondstoBaseTime(0.0),Entity 1:SecondstoBaseTime(EXPO(1)),1:NEXT(215$); 215$ ASSIGN: Elevator Arival.NumberOut=Elevator Arival.NumberOut + 1:NEXT(1$); ; Model statements for module: Assign 1 1$ ASSIGN: Elevator_Number=0: Elevator_Load=0: FloorNumber=38:NEXT(0$); ; Model statements for module: Station 1 0$ STATION, Floors; 218$ ASSIGN: WhatFloor=MEMIDX(Floors,M); 220$ DELAY: 0.0,,VA:NEXT(2$); ; Model statements for module: Decide 1 2$ BRANCH, 1: If,Elevator_Number>=1,221$,Yes: Else,222$,Yes; 221$ ASSIGN: Decide 1.NumberOut True=Decide 1.NumberOut True + 1:NEXT(15$); 222$ ASSIGN: Decide 1.NumberOut False=Decide 1.NumberOut False + 1:NEXT(30$); ; Model statements for module: Decide 2 15$ BRANCH, 1: If,ReleasedAtFloor==WhatFloor,16$,Yes: If,FloorNumber==WhatFloor,60$,Yes: If,(Elevator_Load < Elevator_Capacity) * (WhatFloor > FloorNumber),60$,Yes: Else,32$,Yes; ; Model statements for module: Assign 8 32$ ASSIGN: direction=WhatFloor < FloorNumber: Accl=( ABS( FloorNumber - WhatFloor ) == 1) * 0.95: WhatFloor= ((WhatFloor < FloorNumber) * (WhatFloor + 1)) + ((WhatFloor > FloorNumber) * (WhatFloor - 1)): ReleasedAtFloor=DISC(0.01, WhatFloor, 1.0, FloorNumber):NEXT(29$); ; Model statements for module: Route 5 29$ ROUTE: TravelTime,MEMBER(Floors,WhatFloor); ; Model statements for module: Assign 2 16$ ASSIGN: ReleasedAtFloor=EWhatFloor(2) * Elevator_Load + EWhatFloor(1) * ANINT(Elevator_Load * 7 / 100): Elevator_Load=Elevator_Load - ReleasedAtFloor:NEXT(22$); ; Model statements for module: Dropoff 1 22$ DROPOFF, 1,ReleasedAtFloor:18$,
277
FloorNumber:NEXT(60$); ; Model statements for module: Delay 10 60$ DELAY: (EXPO(10) + (DISC(0.25,0,1.0,1) * EXPO(10))),,Other:NEXT(17$); ; Model statements for module: Decide 3 17$ BRANCH, 1: If,Elevator_Load + NQ(Queues_2(WhatFloor)) <= Elevator_Capacity,225$,Yes: Else,226$,Yes; 225$ ASSIGN: Can_Fit_1_Passenger.NumberOut True=Can_Fit_1_Passenger.NumberOut True + 1:NEXT(19$); 226$ ASSIGN: Can_Fit_1_Passenger.NumberOut False=Can_Fit_1_Passenger.NumberOut False + 1:NEXT(20$); ; Model statements for module: Assign 3 19$ ASSIGN: Elevator_RemainingCap=NQ(Queues_2(WhatFloor)): Elevator_Load=Elevator_Load + Elevator_RemainingCap:NEXT(31$); 31$ PICKUP: Queues_2(WhatFloor),1,Elevator_RemainingCap; 58$ WHILE: TWhatFloor(1) || (NQ(Queues_2(FloorNumber)) < 1) * (Dispose 1.Numberout<1210) * EWhatFloor(2) :NEXT(21$); ; Model statements for module: Assign 5 21$ ASSIGN: FloorNumber= TWhatFloor(1)+(EWhatFloor(2)*DISC(0.027,38,0.054,37,0.081,36,0.108,35,0.135,34,0.162,33,0.189,32,0.216,31,0.243,30,0.27,29,0.297,28,0.324,27,0.351,26,0.378,25,0.405,24,0.432,23,0.459,22,0.486,21,0.513,20,0.54,19,0.567,18,0.594,17,0.621,16,0.648,15,0.675,14,0.702,13,0.729,12,0.756,11,0.783,10,0.81,9,0.837,8,0.864,7,0.891,6,0.918,5,0.945,4,0.972,3,1,2))+(TWhatFloor(2)*FloorNumber): ReleasedAtFloor=FloorNumber:NEXT(61$); 61$ SCAN: (QE(1) + QE(2) + QE(3) + QE(4) + QE(4) + QE(5) + QE(6) + QE(7)) + Elevator_Load >= 1; 59$ ENDWHILE:NEXT(213$); ; Model statements for module: Assign 11 213$ ASSIGN: FloorNumber= ((TWhatFloor(1)+(EWhatFloor(2)*DISC(0.027,38,0.054,37,0.081,36,0.108,35,0.135,34,0.162,33,0.189,32,0.216,31,0.243,30,0.27,29,0.297,28,0.324,27,0.351,26,0.378,25,0.405,24,0.432,23,0.459,22,0.486,21,0.513,20,0.54,19,0.567,18,0.594,17,0.621,16,0.648,15,0.675,14,0.702,13,0.729,12,0.756,11,0.783,10,0.81,9,0.837,8,0.864,7,0.891,6,0.918,5,0.945,4,0.972,3,1,2))+(TWhatFloor(2)*FloorNumber))*(WhatFloor == FloorNumber)) + (FloorNumber * (WhatFloor <> FloorNumber)): ReleasedAtFloor=FloorNumber:NEXT(32$); ; Model statements for module: Assign 4 20$ ASSIGN: Elevator_RemainingCap=Elevator_Capacity - Elevator_Load: Elevator_Load=Elevator_Capacity:NEXT(31$); ; Model statements for module: Route 2 18$ ROUTE: EXPO( 20 ),Apartments(FloorNumber); ; Model statements for module: Assign 6 30$ ASSIGN: Elevator_Number=LastElevatorNumber + 1: LastElevatorNumber=Elevator_Number:NEXT(32$); ; Model statements for module: Create 2 227$ CREATE, 1,SecondstoBaseTime(ResponseTime),Person:SecondstoBaseTime(POIS( 23.5 )),22:NEXT(228$); 228$ ASSIGN: Passenger Arrival Floor 01.NumberOut=Passenger Arrival Floor 01.NumberOut + 1:NEXT(14$); 14$ QUEUE, Floor 01.Queue:DETACH;
278
; Model statements for module: Create 3 231$ CREATE, 1,SecondstoBaseTime(ResponseTime),Person:SecondstoBaseTime(POIS( 6.81 )),32:NEXT(232$); 232$ ASSIGN: Passenger Arrival Floor 02.NumberOut=Passenger Arrival Floor 02.NumberOut + 1:NEXT(13$); 13$ QUEUE, Floor 02.Queue:DETACH; ; Model statements for module: Create 4 207$ ASSIGN: StartTime=TNOW:NEXT(210$); ; Model statements for module: Hold 1 210$ SCAN: Dispose 1.Numberout == 1206:NEXT(208$); ; Model statements for module: Record 1 208$ TALLY: Total System Time,INT(StartTime),1:NEXT(212$); ; Model statements for module: Assign 10 212$ ASSIGN: FinishReplication=1:NEXT(211$); ; Model statements for module: Dispose 2 211$ ASSIGN: Dispose 2.NumberOut=Dispose 2.NumberOut + 1; 609$ DISPOSE: Yes;
279
A r iv a lE le v a t o r
S t a t io n 1A s s ig n 1
F lo o r 0 1A r r iv a l
P a s s e n g e r
F lo o r 0 2A r r iv a l
P a s s e n g e r
F lo o r 0 3A r r iv a l
P a s s e n g e r
F lo o r 0 4A r r iv a l
P a s s e n g e r
T r u e
F a ls e
D e c id e 1
2S t a t io n
D is p o s e 1
0 2
A p a r t m e n t
0 3
A p a r t m e n t
0 4
A p a r t m e n t
D e la y 1 R o u t e 1
6
S t a t io n
F lo o r 0 4 . Q u e u e
Q u e u e
F lo o r 0 3 . Q u e u e
Q u e u e
F lo o r 0 2 . Q u e u e
Q u e u e
F lo o r 0 1 . Q u e u e
Q u e u e
R e le a s e d A t F lo o r = = W h a t F lo o r
F lo o r N u m b e r = = W h a t F lo o r
( E le v a t o r _ L o a d < E le v a t o r _ C a p a c it y ) * ( W h a t F lo o r > F lo o r N u m b e r )E ls e
D e c id e 2
A s s ig n 2
C a n _ F it _ 1 _ P a s s e n g e r
T r u e
F a ls e
R o u t e 2
A s s ig n 3
A s s ig n 4
A s s ig n 5
O r ig in a l
M e m b e r s
1
D r o p o f f
7
S t a t io n
8
S t a t io n
D e la y 2 R o u t e 3
D e la y 3 R o u t e 4
Passenger Zone
Controlle r and L ift Zone
Passenger Re-entry Zone
F ina l Exit
R o u t e 5A s s ig n 6
E le v a t o r _ R e m a in in g C a pQ u e u e s _ 2 ( W h a t F lo o r )
P ic k u p
A s s ig n 8
F lo o r 0 5A r r iv a l
P a s s e n g e r
F lo o r 0 6A r r iv a l
P a s s e n g e r
F lo o r 0 6 . Q u e u e
Q u e u e
F lo o r 0 5 . Q u e u e
Q u e u e
9S t a t io n
1 0
S t a t io n
F lo o r 0 7A r r iv a l
P a s s e n g e r
F lo o r 0 8A r r iv a l
P a s s e n g e r
F lo o r 0 8 . Q u e u e
Q u e u e
F lo o r 0 7 . Q u e u e
Q u e u e
1 1
S t a t io n
1 2
S t a t io n
F lo o r 0 9A r r iv a l
P a s s e n g e r
F lo o r 0 9 . Q u e u e
Q u e u e
1 3S t a t io n
0 6
A p a r t m e n t
0 5
A p a r t m e n t
D e la y 5 R o u t e 6
D e la y 6 R o u t e 7
0 8
A p a r t m e n t
0 7
A p a r t m e n t
D e la y 7 R o u t e 8
D e la y 8 R o u t e 9
0 9A p a r t m e n t
D e la y 91 0
R o u t e
Animation
T W h a t F lo o r ( 1 ) | | ( N Q ( Q u e u e s _ 2 ( F lo o r N u m b e r ) ) < 1 ) * ( D is p o s e 1 . N u m b e r o u t < 1 2 1 0 ) * E W h a t F lo o r ( 2 )
W h ile E n d W h ile
F lo o r D r o p P ic k D e la y
( Q E ( 1 ) + Q E ( 2 ) + Q E ( 3 ) + Q E ( 4 ) + Q E ( 4 ) + Q E ( 5 ) + Q E ( 6 ) + Q E ( 7 ) ) + E le v a t o r _ L o a d > = 1
S c a n
F lo o r 1 0A r r iv a l
P a s s e n g e r
F lo o r 1 1A r r iv a l
P a s s e n g e r
F lo o r 1 1 . Q u e u e
Q u e u e
F lo o r 1 0 . Q u e u e
Q u e u e
2 2
S t a t io n
2 3
S t a t io n
F lo o r 1 2A r r iv a l
P a s s e n g e r
F lo o r 1 2 . Q u e u e
Q u e u e
2 4S t a t io n
1 1
A p a r t m e n t
1 0
A p a r t m e n t
D e la y 1 11 1
R o u t e
D e la y 1 21 2
R o u t e
1 2
A p a r t m e n tD e la y 1 3
1 3
R o u t e
F lo o r 1 3A r r iv a l
P a s s e n g e r
F lo o r 1 4A r r iv a l
P a s s e n g e r
F lo o r 1 5A r r iv a l
P a s s e n g e r
2 8
S t a t io n
F lo o r 1 5 . Q u e u e
Q u e u e
F lo o r 1 4 . Q u e u e
Q u e u e
F lo o r 1 3 . Q u e u e
Q u e u e
2 9
S t a t io n
3 0
S t a t io n
F lo o r 1 6A r r iv a l
P a s s e n g e r
F lo o r 1 7A r r iv a l
P a s s e n g e r
F lo o r 1 7 . Q u e u e
Q u e u e
F lo o r 1 6 . Q u e u e
Q u e u e
3 1
S t a t io n
3 2
S t a t io n
F lo o r 1 8A r r iv a l
P a s s e n g e r
F lo o r 1 9A r r iv a l
P a s s e n g e r
F lo o r 1 9 . Q u e u e
Q u e u e
F lo o r 1 8 . Q u e u e
Q u e u e
3 3S t a t io n
3 4
S t a t io n
F lo o r 2 0A r r iv a l
P a s s e n g e r
F lo o r 2 0 . Q u e u e
Q u e u e
3 5
S t a t io n
F lo o r 2 1A r r iv a l
P a s s e n g e r
F lo o r 2 2A r r iv a l
P a s s e n g e r
F lo o r 2 2 . Q u e u e
Q u e u e
F lo o r 2 1 . Q u e u e
Q u e u e
3 6
S t a t io n
3 7
S t a t io n
F lo o r 2 3A r r iv a l
P a s s e n g e r
F lo o r 2 3 . Q u e u e
Q u e u e
3 8
S t a t io n
F lo o r 2 4A r r iv a l
P a s s e n g e r
F lo o r 2 4 . Q u e u e
Q u e u e
3 9S t a t io n
F lo o r 2 5A r r iv a l
P a s s e n g e r
F lo o r 2 5 . Q u e u e
Q u e u e
4 0
S t a t io n
1 3
A p a r t m e n t
1 4
A p a r t m e n t
1 5
A p a r t m e n t
D e la y 1 41 4
R o u t e
D e la y 1 51 5
R o u t e
D e la y 1 61 6
R o u t e
1 7A p a r t m e n t
1 6
A p a r t m e n t
D e la y 1 71 7
R o u t e
D e la y 1 81 8
R o u t e
1 9
A p a r t m e n t
1 8
A p a r t m e n t
D e la y 1 91 9
R o u t e
D e la y 2 02 0
R o u t e
2 0
A p a r t m e n tD e la y 2 1
2 1
R o u t e
2 2
A p a r t m e n t
2 1
A p a r t m e n t
D e la y 2 22 2
R o u t e
D e la y 2 32 3
R o u t e
2 3
A p a r t m e n tD e la y 2 4
2 4
R o u t e
2 4
A p a r t m e n tD e la y 2 5
2 5
R o u t e
2 5
A p a r t m e n tD e la y 2 6
2 6
R o u t e
F lo o r 2 6A r r iv a l
P a s s e n g e r
F lo o r 2 7A r r iv a l
P a s s e n g e r
5 4
S t a t io n
F lo o r 2 7 . Q u e u e
Q u e u e
F lo o r 2 6 . Q u e u e
Q u e u e
5 5
S t a t io n
F lo o r 2 8A r r iv a l
P a s s e n g e r
F lo o r 2 9A r r iv a l
P a s s e n g e r
F lo o r 2 9 . Q u e u e
Q u e u e
F lo o r 2 8 . Q u e u e
Q u e u e
5 6
S t a t io n
5 7
S t a t io n
F lo o r 3 0A r r iv a l
P a s s e n g e r
F lo o r 3 1A r r iv a l
P a s s e n g e r
F lo o r 3 1 . Q u e u e
Q u e u e
F lo o r 3 0 . Q u e u e
Q u e u e
5 8S t a t io n
5 9
S t a t io n
F lo o r 3 2A r r iv a l
P a s s e n g e r
F lo o r 3 2 . Q u e u e
Q u e u e
6 0
S t a t io n
F lo o r 3 3A r r iv a l
P a s s e n g e r
F lo o r 3 4A r r iv a l
P a s s e n g e r
F lo o r 3 4 . Q u e u e
Q u e u e
F lo o r 3 3 . Q u e u e
Q u e u e
6 1S t a t io n
6 2
S t a t io n
F lo o r 3 5A r r iv a l
P a s s e n g e r
F lo o r 3 5 . Q u e u e
Q u e u e
6 3
S t a t io n
F lo o r 3 6A r r iv a l
P a s s e n g e r
F lo o r 3 6 . Q u e u e
Q u e u e
6 4
S t a t io n
F lo o r 3 7A r r iv a l
P a s s e n g e r
F lo o r 3 7 . Q u e u e
Q u e u e
6 5
S t a t io n
F lo o r 3 8A r r iv a l
P a s s e n g e r
F lo o r 3 8 . Q u e u e
Q u e u e
6 6
S t a t io n
2 6
A p a r t m e n t
2 7
A p a r t m e n t
2 8
A p a r t m e n t
D e la y 2 72 7
R o u t e
D e la y 2 82 8
R o u t e
D e la y 2 92 9
R o u t e
3 0
A p a r t m e n t
2 9
A p a r t m e n t
D e la y 3 03 0
R o u t e
D e la y 3 13 1
R o u t e
3 2
A p a r t m e n t
3 1
A p a r t m e n t
D e la y 3 23 2
R o u t e
D e la y 3 33 3
R o u t e
3 3A p a r t m e n t
D e la y 3 43 4
R o u t e
3 5
A p a r t m e n t
3 4
A p a r t m e n t
D e la y 3 53 5
R o u t e
D e la y 3 63 6
R o u t e
3 6
A p a r t m e n tD e la y 3 7
3 7
R o u t e
3 7
A p a r t m e n tD e la y 3 8
3 8
R o u t e
3 8
A p a r t m e n tD e la y 3 9
3 9
R o u t e
C r e a t e 4 3 A s s ig n 9 R e c o r d 1H o ld 1
D is p o s e 2
A s s ig n 1 0
A s s ig n 1 1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
Figure F1 – Overall Model used for Lift Simulation
280
SIMAN Language Used in Stair Evacuation Model (only a small portion) ; Model statements for module: Create 3 931$ CREATE, 1,SecondstoBaseTime(Response1),Person:SecondstoBaseTime(POIS( 34 )),16:NEXT(932$); 932$ ASSIGN: Passenger Arrival Floor 02.NumberOut=Passenger Arrival Floor 02.NumberOut + 1:NEXT(878$); ; Model statements for module: Assign 73 878$ ASSIGN: Floor_entered=02: AllTWay=DISC(0.99, 1, 1.0, 0): WantedFloor= (AllTWay == 1) + (AllTWay <> 1) * DISC(0.028,38,0.056,37,0.084,36,0.112,35,0.14,34,0.168,33,0.196,32,0.224,31,0.252,30,0.28,29,0.308,28,0.336,27,0.364,26,0.392,25,0.42,24,0.448,23,0.476,22,0.504,21,0.532,20,0.56,19,0.588,18,0.616,17,0.644,16,0.672,15,0.7,14,0.728,13,0.756,12,0.784,11,0.812,10,0.84,9,0.868,8,0.896,7,0.924,6,0.95,5,0.98,4,1,3) :NEXT(189$); ; Model statements for module: Station 116 189$ STATION, Stair_lobby02; 937$ DELAY: 0.0,,VA:NEXT(190$); ; Model statements for module: Route 112 190$ ROUTE: 0.000000000000000,Floor 02; ; Model statements for module: Dispose 1 1$ ASSIGN: Dispose 1.NumberOut=Dispose 1.NumberOut + 1; 955$ DISPOSE: Yes; ; Model statements for module: Station 111 179$ STATION, Stair_lobby07; 985$ DELAY: 0.0,,VA:NEXT(180$); ; Model statements for module: Route 102 180$ ROUTE: 0.000000000000000,Floor 07; ; Model statements for module: Create 9 986$ CREATE, 1,SecondstoBaseTime(Response),Person:SecondstoBaseTime(POIS( 14.5 )),16:NEXT(987$); 987$ ASSIGN: Passenger Arrival Floor 08.NumberOut=Passenger Arrival Floor 08.NumberOut + 1:NEXT(782$); ; Model statements for module: Assign 67 ; Model statements for module: Decide 1 198$ BRANCH, 1: If,WantedFloor == 37,211$,Yes: If,Floor_entered == 37 && WantedFloor > 37,202$,Yes: If,Floor_entered == 37 && WantedFloor < 37,205$,Yes: If,WantedFloor > 37,208$,Yes: Else,196$,Yes; ; Model statements for module: Seize 79 205$ QUEUE, Floor 37_2.Queue; SEIZE, 2,Other: Floor37_36Stairs,1:NEXT(1269$); 1269$ DELAY: 0.0,,VA:NEXT(204$); ; Model statements for module: Route 117 204$ ROUTE: TRIA( 13.4, 19.6, 24.6 ),Floor 36; ; Model statements for module: Seize 80 208$ QUEUE, Floor 37.Queue; SEIZE, 2,Other: Floor38_37Stairs,1:NEXT(1271$);
281
1271$ DELAY: 0.0,,VA:NEXT(210$); ; Model statements for module: Assign 82 887$ ASSIGN: Floor_entered=29: WantedFloor=1:NEXT(90$); ; Model statements for module: Route 31 90$ ROUTE: EXPO( 20 ),Stair_lobby29; ; Model statements for module: Release 145 871$ RELEASE: Floor02_01Stairs,1:NEXT(868$); ; Model statements for module: Route 293 868$ ROUTE: TRIA( 13.4, 19.6, 24.6 ),Floor 03; ; Model statements for module: Create 43 1787$ CREATE, 1,SecondstoBaseTime(Response),Person:SecondstoBaseTime(POIS( 33 )),6:NEXT(1788$); 1788$ ASSIGN: Passenger Arrival Floor 01.NumberOut=Passenger Arrival Floor 01.NumberOut + 1:NEXT(920$); ; Model statements for module: Assign 110 920$ ASSIGN: Floor_entered=01: WantedFloor= DISC(0.027,38,0.054,37,0.081,36,0.108,35,0.135,34,0.162,33,0.189,32,0.216,31,0.243,30,0.27,29,0.297,28,0.324,27,0.351,26,0.378,25,0.405,24,0.432,23,0.459,22,0.486,21,0.517,20,0.54,19,0.567,18,0.594,17,0.621,16,0.648,15,0.675,14,0.702,13,0.729,12,0.756,11,0.783,10,0.81,9,0.837,8,0.864,7,0.891,6,0.918,5,0.945,4,0.972,3,1,2) :NEXT(918$); ; Model statements for module: Station 155 918$ STATION, Stair_lobby01; 1793$ DELAY: 0.0,,VA:NEXT(919$); ; Model statements for module: Route 296 919$ ROUTE: 0.000000000000000,Floor 01; ; Model statements for module: Create 44 1794$ CREATE, 1,HoursToBaseTime(0.0),TimerEntity:HoursToBaseTime(EXPO(1)):NEXT(1795$); 1795$ ASSIGN: Create 44.NumberOut=Create 44.NumberOut + 1:NEXT(925$); ; Model statements for module: Assign 111 925$ ASSIGN: StartTime=TNOW:NEXT(928$); ; Model statements for module: Hold 1 928$ SCAN: Dispose 1.Numberout == 598:NEXT(926$); ; Model statements for module: Record 1 926$ TALLY: Total System Time,INT(StartTime),1:NEXT(930$); ; Model statements for module: Assign 112 930$ ASSIGN: FinishReplication=1:NEXT(929$); ; Model statements for module: Dispose 2 929$ ASSIGN: Dispose 2.NumberOut=Dispose 2.NumberOut + 1; 1798$ DISPOSE: Yes;
282
0 2F lo o r
A r r iv a lP a s s e n g e r
0 3F lo o r
A r r iv a lP a s s e n g e r
0 4F lo o r
A r r iv a lP a s s e n g e r
2
S t a t io n1
D is p o s e
0 2
A p a r t m e n t
0 3A p a r t m e n t
0 4
A p a r t m e n t
D e la y 1
1R o u t e
D e la y 2
3R o u t e
D e la y 3 4
R o u t e
F in a l E xi t
0 5F lo o r
A r r iv a lP a s s e n g e r
0 6F lo o r
A r r iv a l
P a s s e n g e r
0 7F lo o r
A r r iv a lP a s s e n g e r
0 8F lo o r
A r r iv a lP a s s e n g e r
0 9F lo o r
A r r iv a lP a s s e n g e r
0 6
A p a r t m e n t
0 5
A p a r t m e n t
D e la y 5 6
R o u t e
D e la y 67
R o u t e
0 8
A p a r t m e n t
0 7
A p a r t m e n t
D e la y 78
R o u t e
D e la y 8 9
R o u t e
0 9
A p a r t m e n tD e la y 9 1 0
R o u t e
1 0F lo o r
A r r iv a lP a s s e n g e r
1 1F lo o r
A r r iv a lP a s s e n g e r
1 2F lo o r
A r r iv a lP a s s e n g e r
1 1
A p a r t m e n t
1 0A p a r t m e n t
1 1
D e la y
1 1
R o u t e
1 2D e la y
1 2R o u t e
1 2
A p a r t m e n t
1 3
D e la y
1 3
R o u t e
1 3F lo o r
A r r iv a lP a s s e n g e r
1 4F lo o r
A r r iv a lP a s s e n g e r
1 5F lo o r
A r r iv a lP a s s e n g e r
1 6F lo o r
A r r iv a lP a s s e n g e r
1 7F lo o r
A r r iv a lP a s s e n g e r
1 8F lo o r
A r r iv a lP a s s e n g e r
1 9F lo o r
A r r iv a lP a s s e n g e r
2 0F lo o r
A r r iv a lP a s s e n g e r
2 1F lo o r
A r r iv a lP a s s e n g e r
2 2
F lo o rA r r iv a l
P a s s e n g e r
2 3F lo o r
A r r iv a lP a s s e n g e r
2 4F lo o r
A r r iv a lP a s s e n g e r
2 5F lo o r
A r r iv a lP a s s e n g e r
1 3
A p a r t m e n t
1 4
A p a r t m e n t
1 5
A p a r t m e n t
1 4
D e la y
1 4R o u t e
1 5
D e la y
1 5
R o u t e
1 6
D e la y
1 6
R o u t e
1 7
A p a r t m e n t
1 6
A p a r t m e n t
1 7
D e la y
1 7
R o u t e
1 8
D e la y1 8
R o u t e
1 9
A p a r t m e n t
1 8
A p a r t m e n t
1 9
D e la y
1 9
R o u t e
2 0
D e la y
2 0
R o u t e
2 0
A p a r t m e n t
2 1
D e la y2 1
R o u t e
2 2A p a r t m e n t
2 1A p a r t m e n t
2 2D e la y
2 2
R o u t e
2 3D e la y
2 3
R o u t e
2 3A p a r t m e n t
2 4D e la y
2 4
R o u t e
2 4
A p a r t m e n t
2 5
D e la y
2 5
R o u t e
2 5
A p a r t m e n t
2 6
D e la y
2 6
R o u t e
2 6F lo o r
A r r iv a lP a s s e n g e r
2 7F lo o r
A r r iv a lP a s s e n g e r
2 8F lo o r
A r r iv a lP a s s e n g e r
2 9F lo o r
A r r iv a lP a s s e n g e r
3 0F lo o r
A r r iv a l
P a s s e n g e r
3 1F lo o r
A r r iv a lP a s s e n g e r
3 2F lo o r
A r r iv a lP a s s e n g e r
3 3F lo o r
A r r iv a lP a s s e n g e r
3 4
F lo o rA r r iv a l
P a s s e n g e r
3 5F lo o r
A r r iv a lP a s s e n g e r
3 6F lo o r
A r r iv a lP a s s e n g e r
3 7F lo o r
A r r iv a lP a s s e n g e r
6 5S t a t io n
3 8F lo o r
A r r iv a lP a s s e n g e r
6 6
S t a t io n
2 6
A p a r t m e n t
2 7A p a r t m e n t
2 8A p a r t m e n t
2 7
D e la y2 7
R o u t e
2 8D e la y
2 8R o u t e
2 9D e la y
2 9R o u t e
3 0A p a r t m e n t
2 9A p a r t m e n t
3 0D e la y
3 0R o u t e
3 1D e la y
3 1
R o u t e
3 2
A p a r t m e n t
3 1
A p a r t m e n t
3 2
D e la y3 2
R o u t e
3 3
D e la y
3 3
R o u t e
3 3
A p a r t m e n t
3 4
D e la y3 4
R o u t e
3 5
A p a r t m e n t
3 4A p a r t m e n t
3 5
D e la y3 5
R o u t e
3 6D e la y
3 6R o u t e
3 6
A p a r t m e n t
3 7
D e la y 3 7
R o u t e
3 7
A p a r t m e n t
3 8
D e la y
3 8
R o u t e
3 8
A p a r t m e n t3 9
D e la y
3 9
R o u t e
4 0
R o u t e
8 0S t a t io n
4 1
R o u t e
8 1
S t a t io n4 2
R o u t e
4 3R o u t e
8 2
S t a t io n
4 4
R o u t e
8 3
S t a t io n4 6
R o u t e
8 4
S t a t io n
4 8
R o u t e
8 5S t a t io n
5 0R o u t e
8 6
S t a t io n5 2
R o u t e
8 7
S t a t io n5 4
R o u t e
8 8S t a t io n 5 6
R o u t e
8 9
S t a t io n
5 8
R o u t e
9 0
S t a t io n6 0
R o u t e
9 1S t a t io n
6 2R o u t e
9 2
S t a t io n
6 4R o u t e
9 3S t a t io n
6 6
R o u t e
9 4S t a t io n
6 8R o u t e
9 5
S t a t io n
7 0
R o u t e
9 6
S t a t io n
7 2
R o u t e
9 7
S t a t io n
7 4
R o u t e
9 8S t a t io n
7 6R o u t e
9 9
S t a t io n7 8
R o u t e
1 0 0S t a t io n
8 0R o u t e
1 0 1
S t a t io n
8 2
R o u t e
1 0 2
S t a t io n8 4
R o u t e
1 0 3S t a t io n
8 6R o u t e
1 0 4S t a t io n
8 8R o u t e
1 0 5
S t a t io n9 0
R o u t e
1 0 6
S t a t io n
9 2
R o u t e
1 0 7
S t a t io n
9 4
R o u t e
1 0 8
S t a t io n9 6
R o u t e
1 0 9
S t a t io n9 8
R o u t e
1 1 0
S t a t io n
1 0 0
R o u t e
1 1 1
S t a t io n
1 0 2
R o u t e
1 1 2
S t a t io n
1 0 4
R o u t e
1 1 3
S t a t io n1 0 6
R o u t e
1 1 4
S t a t io n
1 0 8
R o u t e
1 1 5
S t a t io n1 1 0
R o u t e
1 1 6
S t a t io n1 1 2
R o u t e
1 1 7
S t a t io n
1 1 4R o u t e
1
S e iz e
2S e iz e
1R e le a s e
D e c id e 1
W a n t e d F lo o r = = 3 7F lo o r _ e n t e r e d = = 3 7 & & W a n t e d F lo o r > 3 7F lo o r _ e n t e r e d = = 3 7 & & W a n t e d F lo o r < 3 7W a n t e d F lo o r > 3 7E ls e
2A s s ig n
3 9
R e le a s e
1 1 5
R o u t e
7 8
S e iz e
1 1 7
R o u t e
7 9
S e iz e
1 1 8
R o u t e
8 0S e iz e 4 0
R e le a s e
4 1
R e le a s e
1 1 9
R o u t e
3 8A s s ig n
1 2 0
S t a t io n
1 2 0
R o u t e
8 1
S e iz e4 2
R e le a s e
D e c id e 3 8
W a n t e d F lo o r = = 3 6F lo o r _ e n t e r e d = = 3 6 & & W a n t e d F lo o r > 3 6F lo o r _ e n t e r e d = = 3 6 & & W a n t e d F lo o r < 3 6W a n t e d F lo o r > 3 6E ls e
1 2 1
R o u t e8 2
S e iz e
1 2 2
R o u t e
8 3
S e iz e
1 2 3
R o u t e
8 4
S e iz e4 3
R e le a s e
4 4
R e le a s e
1 2 4
R o u t e
1 2 1
S t a t io n
1 2 5
R o u t e
8 5
S e iz e4 5
R e le a s e
D e c id e 3 9
W a n t e d F lo o r = = 3 5F lo o r _ e n t e r e d = = 3 5 & & W a n t e d F lo o r > 3 5F lo o r _ e n t e r e d = = 3 5 & & W a n t e d F lo o r < 3 5W a n t e d F lo o r > 3 5E ls e
1 2 6
R o u t e
8 6
S e iz e
1 2 7
R o u t e
8 7
S e iz e
1 2 8
R o u t e
8 8
S e iz e4 6
R e le a s e
4 7
R e le a s e1 2 9
R o u t e
1 2 2S t a t io n
1 3 0
R o u t e
8 9
S e iz e4 8
R e le a s e
D e c id e 4 0
W a n t e d F lo o r = = 3 4F lo o r _ e n t e r e d = = 3 4 & & W a n t e d F lo o r > 3 4F lo o r _ e n t e r e d = = 3 4 & & W a n t e d F lo o r < 3 4W a n t e d F lo o r > 3 4E ls e
1 3 1R o u t e
9 0S e iz e
1 3 2R o u t e
9 1S e iz e
1 3 3R o u t e
9 2S e iz e
4 9R e le a s e
5 0R e le a s e
1 3 4R o u t e
1 2 3
S t a t io n
1 3 5
R o u t e
9 3
S e iz e5 1
R e le a s e
D e c id e 4 1
W a n t e d F lo o r = = 3 3
F lo o r _ e n t e r e d = = 3 3 & & W a n t e d F lo o r > 3 3
F lo o r _ e n t e r e d = = 3 3 & & W a n t e d F lo o r < 3 3
W a n t e d F lo o r > 3 3E ls e
1 3 6
R o u t e
9 4
S e iz e
1 3 7
R o u t e
9 5
S e iz e
1 3 8
R o u t e
9 6
S e iz e5 2
R e le a s e
5 3
R e le a s e1 3 9
R o u t e
1 2 4
S t a t io n
1 4 0
R o u t e
9 7
S e iz e5 4
R e le a s e
D e c id e 4 2
W a n t e d F lo o r = = 3 2
F lo o r _ e n t e r e d = = 3 2 & & W a n t e d F lo o r > 3 2
F lo o r _ e n t e r e d = = 3 2 & & W a n t e d F lo o r < 3 2
W a n t e d F lo o r > 3 2E ls e
1 4 1
R o u t e
9 8
S e iz e
1 4 2
R o u t e
9 9
S e iz e
1 4 3
R o u t e
1 0 0
S e iz e5 5
R e le a s e
5 6
R e le a s e1 4 4
R o u t e
3 9A s s ig n
4 0A s s ig n
4 1A s s ig n
4 2A s s ig n
4 3A s s ig n
1 2 5
S t a t io n
1 4 5
R o u t e
1 0 1
S e iz e5 7
R e le a s e
D e c id e 4 3
W a n t e d F lo o r = = 3 1
F lo o r _ e n t e r e d = = 3 1 & & W a n t e d F lo o r > 3 1F lo o r _ e n t e r e d = = 3 1 & & W a n t e d F lo o r < 3 1
W a n t e d F lo o r > 3 1E ls e
1 4 6R o u t e
1 0 2S e iz e
1 4 7R o u t e
1 0 3S e iz e
1 4 8R o u t e
1 0 4S e iz e
5 8R e le a s e
5 9R e le a s e
1 4 9
R o u t e
1 2 6
S t a t io n
1 5 0
R o u t e
1 0 5
S e iz e6 0
R e le a s e
D e c id e 4 4
W a n t e d F lo o r = = 3 0
F lo o r _ e n t e r e d = = 3 0 & & W a n t e d F lo o r > 3 0
F lo o r _ e n t e r e d = = 3 0 & & W a n t e d F lo o r < 3 0
W a n t e d F lo o r > 3 0E ls e
1 5 1
R o u t e
1 0 6
S e iz e
1 5 2
R o u t e
1 0 7
S e iz e
1 5 3
R o u t e
1 0 8
S e iz e6 1
R e le a s e
6 2
R e le a s e1 5 4
R o u t e
1 2 7
S t a t io n
1 5 5
R o u t e
1 0 9
S e iz e6 3
R e le a s e
D e c id e 4 5
W a n t e d F lo o r = = 2 9
F lo o r _ e n t e r e d = = 2 9 & & W a n t e d F lo o r > 2 9
F lo o r _ e n t e r e d = = 2 9 & & W a n t e d F lo o r < 2 9
W a n t e d F lo o r > 2 9E ls e
1 5 6
R o u t e
1 1 0
S e iz e
1 5 7
R o u t e
1 1 1
S e iz e
1 5 8
R o u t e
1 1 2
S e iz e6 4
R e le a s e
6 5
R e le a s e1 5 9
R o u t e
1 2 8
S t a t io n
1 6 0
R o u t e
1 1 3
S e iz e6 6
R e le a s e
D e c id e 4 6
W a n t e d F lo o r = = 2 8
F lo o r _ e n t e r e d = = 2 8 & & W a n t e d F lo o r > 2 8
F lo o r _ e n t e r e d = = 2 8 & & W a n t e d F lo o r < 2 8W a n t e d F lo o r > 2 8E ls e
1 6 1
R o u t e
1 1 4
S e iz e
1 6 2
R o u t e
1 1 5
S e iz e
1 6 3R o u t e
1 1 6
S e iz e6 7
R e le a s e
6 8
R e le a s e1 6 4
R o u t e
1 2 9
S t a t io n
1 6 5
R o u t e
1 1 7
S e iz e6 9
R e le a s e
D e c id e 4 7
W a n t e d F lo o r = = 2 7
F lo o r _ e n t e r e d = = 2 7 & & W a n t e d F lo o r > 2 7
F lo o r _ e n t e r e d = = 2 7 & & W a n t e d F lo o r < 2 7W a n t e d F lo o r > 2 7E ls e
1 6 6
R o u t e
1 1 8
S e iz e
1 6 7
R o u t e
1 1 9
S e iz e
1 6 8
R o u t e
1 2 0
S e iz e7 0
R e le a s e
7 1
R e le a s e1 6 9
R o u t e
1 3 0
S t a t io n
1 7 0
R o u t e
1 2 1S e iz e
7 2
R e le a s e
D e c id e 4 8
W a n t e d F lo o r = = 2 6F lo o r _ e n t e r e d = = 2 6 & & W a n t e d F lo o r > 2 6
F lo o r _ e n t e r e d = = 2 6 & & W a n t e d F lo o r < 2 6
W a n t e d F lo o r > 2 6E ls e
1 7 1
R o u t e
1 2 2
S e iz e
1 7 2R o u t e
1 2 3S e iz e
1 7 3
R o u t e
1 2 4
S e iz e7 3
R e le a s e
7 4
R e le a s e1 7 4
R o u t e
4 4A s s ig n
4 5
A s s ig n
4 6A s s ig n
4 7A s s ig n
4 8A s s ig n
4 9A s s ig n
1 3 1S t a t io n
1 7 5R o u t e
1 2 5S e iz e
7 5
R e le a s e
D e c id e 4 9
W a n t e d F lo o r = = 2 5
F lo o r _ e n t e r e d = = 2 5 & & W a n t e d F lo o r > 2 5F lo o r _ e n t e r e d = = 2 5 & & W a n t e d F lo o r < 2 5W a n t e d F lo o r > 2 5E ls e
1 7 6R o u t e
1 2 6S e iz e
1 7 7
R o u t e
1 2 7
S e iz e
1 7 8
R o u t e
1 2 8
S e iz e7 6
R e le a s e
7 7
R e le a s e1 7 9
R o u t e
1 3 2
S t a t io n
1 8 0
R o u t e
1 2 9
S e iz e7 8
R e le a s e
D e c id e 5 0
W a n t e d F lo o r = = 2 4
F lo o r _ e n t e r e d = = 2 4 & & W a n t e d F lo o r > 2 4F lo o r _ e n t e r e d = = 2 4 & & W a n t e d F lo o r < 2 4W a n t e d F lo o r > 2 4E ls e
1 8 1
R o u t e
1 3 0
S e iz e
1 8 2
R o u t e
1 3 1
S e iz e
1 8 3
R o u t e
1 3 2
S e iz e7 9
R e le a s e
8 0
R e le a s e1 8 4
R o u t e
1 3 3
S t a t io n
1 8 5
R o u t e
1 3 3
S e iz e8 1
R e le a s e
D e c id e 5 1
W a n t e d F lo o r = = 2 3
F lo o r _ e n t e r e d = = 2 3 & & W a n t e d F lo o r > 2 3F lo o r _ e n t e r e d = = 2 3 & & W a n t e d F lo o r < 2 3W a n t e d F lo o r > 2 3E ls e
1 8 6
R o u t e
1 3 4
S e iz e
1 8 7
R o u t e
1 3 5
S e iz e
1 8 8
R o u t e
1 3 6
S e iz e8 2
R e le a s e
8 3
R e le a s e1 8 9
R o u t e
1 3 4S t a t io n
1 9 0R o u t e
1 3 7S e iz e
8 4R e le a s e
D e c id e 5 2
W a n t e d F lo o r = = 2 2F lo o r _ e n t e r e d = = 2 2 & & W a n t e d F lo o r > 2 2F lo o r _ e n t e r e d = = 2 2 & & W a n t e d F lo o r < 2 2
W a n t e d F lo o r > 2 2E ls e
1 9 1
R o u t e
1 3 8
S e iz e
1 9 2
R o u t e
1 3 9
S e iz e
1 9 3
R o u t e
1 4 0
S e iz e8 5
R e le a s e
8 6
R e le a s e1 9 4
R o u t e
1 3 5
S t a t io n
1 9 5
R o u t e
1 4 1
S e iz e8 7
R e le a s e
D e c id e 5 3
W a n t e d F lo o r = = 2 1F lo o r _ e n t e r e d = = 2 1 & & W a n t e d F lo o r > 2 1F lo o r _ e n t e r e d = = 2 1 & & W a n t e d F lo o r < 2 1W a n t e d F lo o r > 2 1E ls e
1 9 6
R o u t e
1 4 2
S e iz e
1 9 7
R o u t e
1 4 3
S e iz e
1 9 8
R o u t e
1 4 4
S e iz e8 8
R e le a s e
8 9
R e le a s e1 9 9
R o u t e
1 3 6
S t a t io n
2 0 0
R o u t e
1 4 5
S e iz e9 0
R e le a s e
D e c id e 5 4
W a n t e d F lo o r = = 2 0F lo o r _ e n t e r e d = = 2 0 & & W a n t e d F lo o r > 2 0F lo o r _ e n t e r e d = = 2 0 & & W a n t e d F lo o r < 2 0W a n t e d F lo o r > 2 0E ls e
2 0 1
R o u t e
1 4 6
S e iz e
2 0 2
R o u t e
1 4 7
S e iz e
2 0 3
R o u t e
1 4 8
S e iz e9 1
R e le a s e
9 2
R e le a s e2 0 4
R o u t e
1 3 7
S t a t io n
2 0 5
R o u t e
1 4 9
S e iz e9 3
R e le a s e
D e c id e 5 5
W a n t e d F lo o r = = 1 9
F lo o r _ e n t e r e d = = 1 9 & & W a n t e d F lo o r > 1 9
F lo o r _ e n t e r e d = = 1 9 & & W a n t e d F lo o r < 1 9
W a n t e d F lo o r > 1 9E ls e
2 0 6R o u t e
1 5 0S e iz e
2 0 7R o u t e
1 5 1S e iz e
2 0 8R o u t e
1 5 2S e iz e
9 4R e le a s e
9 5R e le a s e
2 0 9
R o u t e
5 0A s s ig n
5 1A s s ig n
5 2A s s ig n
5 3A s s ig n
5 4A s s ig n
5 5
A s s ig n
5 6A s s ig n
1 3 8
S t a t io n
2 1 0
R o u t e
1 5 3
S e iz e9 6
R e le a s e
D e c id e 5 6
W a n t e d F lo o r = = 1 8F lo o r _ e n t e r e d = = 1 8 & & W a n t e d F lo o r > 1 8
F lo o r _ e n t e r e d = = 1 8 & & W a n t e d F lo o r < 1 8
W a n t e d F lo o r > 1 8E ls e
2 1 1
R o u t e
1 5 4
S e iz e
2 1 2
R o u t e
1 5 5
S e iz e
2 1 3
R o u t e
1 5 6
S e iz e9 7
R e le a s e
9 8
R e le a s e2 1 4
R o u t e
1 3 9
S t a t io n
2 1 5
R o u t e
1 5 7
S e iz e9 9
R e le a s e
D e c id e 5 7
W a n t e d F lo o r = = 1 7
F lo o r _ e n t e r e d = = 1 7 & & W a n t e d F lo o r > 1 7
F lo o r _ e n t e r e d = = 1 7 & & W a n t e d F lo o r < 1 7
W a n t e d F lo o r > 1 7E ls e
2 1 6
R o u t e
1 5 8
S e iz e
2 1 7
R o u t e
1 5 9
S e iz e
2 1 8
R o u t e
1 6 0
S e iz e1 0 0
R e le a s e
1 0 1
R e le a s e2 1 9
R o u t e
1 4 0
S t a t io n
2 2 0
R o u t e
1 6 1
S e iz e1 0 2
R e le a s e
D e c id e 5 8
W a n t e d F lo o r = = 1 6
F lo o r _ e n t e r e d = = 1 6 & & W a n t e d F lo o r > 1 6
F lo o r _ e n t e r e d = = 1 6 & & W a n t e d F lo o r < 1 6
W a n t e d F lo o r > 1 6E ls e
2 2 1
R o u t e
1 6 2
S e iz e
2 2 2
R o u t e
1 6 3
S e iz e
2 2 3
R o u t e
1 6 4
S e iz e1 0 3
R e le a s e
1 0 4
R e le a s e2 2 4
R o u t e
1 4 1
S t a t io n
2 2 5
R o u t e
1 6 5
S e iz e1 0 5
R e le a s e
D e c id e 5 9
W a n t e d F lo o r = = 1 5
F lo o r _ e n t e r e d = = 1 5 & & W a n t e d F lo o r > 1 5
F lo o r _ e n t e r e d = = 1 5 & & W a n t e d F lo o r < 1 5W a n t e d F lo o r > 1 5E ls e
2 2 6
R o u t e
1 6 6
S e iz e
2 2 7
R o u t e
1 6 7
S e iz e
2 2 8R o u t e
1 6 8S e iz e
1 0 6R e le a s e
1 0 7R e le a s e
2 2 9R o u t e
1 4 2
S t a t io n
2 3 0
R o u t e
1 6 9
S e iz e1 0 8
R e le a s e
D e c id e 6 0
W a n t e d F lo o r = = 1 4
F lo o r _ e n t e r e d = = 1 4 & & W a n t e d F lo o r > 1 4
F lo o r _ e n t e r e d = = 1 4 & & W a n t e d F lo o r < 1 4W a n t e d F lo o r > 1 4E ls e
2 3 1R o u t e
1 7 0S e iz e
2 3 2R o u t e
1 7 1S e iz e
2 3 3R o u t e
1 7 2S e iz e
1 0 9R e le a s e
1 1 0R e le a s e
2 3 4R o u t e
1 4 3
S t a t io n
2 3 5R o u t e
1 7 3S e iz e
1 1 1
R e le a s e
D e c id e 6 1
W a n t e d F lo o r = = 1 3F lo o r _ e n t e r e d = = 1 3 & & W a n t e d F lo o r > 1 3
F lo o r _ e n t e r e d = = 1 3 & & W a n t e d F lo o r < 1 3W a n t e d F lo o r > 1 3E ls e
2 3 6
R o u t e
1 7 4
S e iz e
2 3 7R o u t e
1 7 5
S e iz e
2 3 8R o u t e
1 7 6S e iz e
1 1 2
R e le a s e
1 1 3
R e le a s e2 3 9
R o u t e
1 4 4
S t a t io n
2 4 0R o u t e
1 7 7S e iz e
1 1 4R e le a s e
D e c id e 6 2
W a n t e d F lo o r = = 1 2
F lo o r _ e n t e r e d = = 1 2 & & W a n t e d F lo o r > 1 2
F lo o r _ e n t e r e d = = 1 2 & & W a n t e d F lo o r < 1 2
W a n t e d F lo o r > 1 2E ls e
2 4 1
R o u t e
1 7 8
S e iz e
2 4 2
R o u t e
1 7 9
S e iz e
2 4 3R o u t e
1 8 0
S e iz e1 1 5
R e le a s e
1 1 6
R e le a s e 2 4 4
R o u t e
5 7A s s ig n
5 8A s s ig n
5 9A s s ig n
6 0A s s ig n
6 1A s s ig n
6 2
A s s ig n
6 3A s s ig n
1 4 5
S t a t io n
2 4 5
R o u t e
1 8 1
S e iz e1 1 7
R e le a s e
D e c id e 6 3
W a n t e d F lo o r = = 1 1F lo o r _ e n t e r e d = = 1 1 & & W a n t e d F lo o r > 1 1F lo o r _ e n t e r e d = = 1 1 & & W a n t e d F lo o r < 1 1W a n t e d F lo o r > 1 1E ls e
2 4 6
R o u t e
1 8 2
S e iz e
2 4 7
R o u t e
1 8 3
S e iz e
2 4 8R o u t e
1 8 4S e iz e
1 1 8R e le a s e
1 1 9
R e le a s e2 4 9
R o u t e
1 4 6
S t a t io n
2 5 0
R o u t e
1 8 5
S e iz e1 2 0
R e le a s e
D e c id e 6 4
W a n t e d F lo o r = = 1 0F lo o r _ e n t e r e d = = 1 0 & & W a n t e d F lo o r > 1 0F lo o r _ e n t e r e d = = 1 0 & & W a n t e d F lo o r < 1 0W a n t e d F lo o r > 1 0E ls e
2 5 1
R o u t e
1 8 6
S e iz e
2 5 2
R o u t e
1 8 7
S e iz e
2 5 3
R o u t e
1 8 8
S e iz e1 2 1
R e le a s e
1 2 2
R e le a s e2 5 4
R o u t e
1 4 7
S t a t io n
2 5 5
R o u t e
1 8 9
S e iz e1 2 3
R e le a s e
D e c id e 6 5
W a n t e d F lo o r = = 0 9F lo o r _ e n t e r e d = = 0 9 & & W a n t e d F lo o r > 0 9F lo o r _ e n t e r e d = = 0 9 & & W a n t e d F lo o r < 0 9W a n t e d F lo o r > 0 9E ls e
2 5 6
R o u t e
1 9 0
S e iz e
2 5 7
R o u t e
1 9 1
S e iz e
2 5 8
R o u t e
1 9 2
S e iz e1 2 4
R e le a s e
1 2 5
R e le a s e
2 5 9
R o u t e
1 4 8
S t a t io n
2 6 0
R o u t e
1 9 3
S e iz e1 2 6
R e le a s e
D e c id e 6 6
W a n t e d F lo o r = = 0 8F lo o r _ e n t e r e d = = 0 8 & & W a n t e d F lo o r > 0 8
F lo o r _ e n t e r e d = = 0 8 & & W a n t e d F lo o r < 0 8W a n t e d F lo o r > 0 8E ls e
2 6 1
R o u t e
1 9 4
S e iz e
2 6 2
R o u t e
1 9 5
S e iz e
2 6 3
R o u t e
1 9 6
S e iz e1 2 7
R e le a s e
1 2 8
R e le a s e2 6 4
R o u t e
1 4 9
S t a t io n
2 6 5
R o u t e1 9 7
S e iz e
1 2 9
R e le a s e
D e c id e 6 7
W a n t e d F lo o r = = 0 7
F lo o r _ e n t e r e d = = 0 7 & & W a n t e d F lo o r > 0 7F lo o r _ e n t e r e d = = 0 7 & & W a n t e d F lo o r < 0 7W a n t e d F lo o r > 0 7E ls e
2 6 6R o u t e
1 9 8S e iz e
2 6 7
R o u t e
1 9 9
S e iz e
2 6 8
R o u t e
2 0 0
S e iz e1 3 0
R e le a s e
1 3 1
R e le a s e2 6 9
R o u t e
6 4A s s ig n
6 5A s s ig n
6 6A s s ig n
6 7A s s ig n
6 8A s s ig n
1 5 0
S t a t io n
2 7 0
R o u t e2 0 1
S e iz e
1 3 2
R e le a s e
D e c id e 6 8
W a n t e d F lo o r = = 0 6
F lo o r _ e n t e r e d = = 0 6 & & W a n t e d F lo o r > 0 6
F lo o r _ e n t e r e d = = 0 6 & & W a n t e d F lo o r < 0 6
W a n t e d F lo o r > 0 6E ls e
2 7 1
R o u t e
2 0 2
S e iz e
2 7 2
R o u t e
2 0 3
S e iz e
2 7 3
R o u t e
2 0 4
S e iz e1 3 3
R e le a s e
1 3 4
R e le a s e2 7 4
R o u t e
1 5 1S t a t io n
2 7 5R o u t e
2 0 5S e iz e
1 3 5R e le a s e
D e c id e 6 9
W a n t e d F lo o r = = 0 5
F lo o r _ e n t e r e d = = 0 5 & & W a n t e d F lo o r > 0 5
F lo o r _ e n t e r e d = = 0 5 & & W a n t e d F lo o r < 0 5
W a n t e d F lo o r > 0 5E ls e
2 7 6R o u t e
2 0 6S e iz e
2 7 7R o u t e
2 0 7S e iz e
2 7 8R o u t e
2 0 8S e iz e
1 3 6R e le a s e
1 3 7R e le a s e 2 7 9
R o u t e
1 5 2
S t a t io n
2 8 0
R o u t e2 0 9
S e iz e
1 3 8
R e le a s e
D e c id e 7 0
W a n t e d F lo o r = = 0 4
F lo o r _ e n t e r e d = = 0 4 & & W a n t e d F lo o r > 0 4
F lo o r _ e n t e r e d = = 0 4 & & W a n t e d F lo o r < 0 4
W a n t e d F lo o r > 0 4E ls e
2 8 1
R o u t e
2 1 0
S e iz e
2 8 2
R o u t e
2 1 1
S e iz e
2 8 3
R o u t e
2 1 2
S e iz e1 3 9
R e le a s e
1 4 0
R e le a s e2 8 4
R o u t e
1 5 3
S t a t io n
2 8 5
R o u t e2 1 3
S e iz e
1 4 1
R e le a s e
D e c id e 7 1
W a n t e d F lo o r = = 0 3
F lo o r _ e n t e r e d = = 0 3 & & W a n t e d F lo o r > 0 3
F lo o r _ e n t e r e d = = 0 3 & & W a n t e d F lo o r < 0 3
W a n t e d F lo o r > 0 3E ls e
2 8 6
R o u t e
2 1 4
S e iz e
2 8 7
R o u t e
2 1 5
S e iz e
2 8 8
R o u t e2 1 6
S e iz e1 4 2
R e le a s e
1 4 3
R e le a s e2 8 9
R o u t e
1 5 4
S t a t io n
2 9 0R o u t e
2 1 7
S e iz e
1 4 4
R e le a s e
D e c id e 7 2
W a n t e d F lo o r = = 0 2
F lo o r _ e n t e r e d = = 0 2 & & W a n t e d F lo o r > 0 2
F lo o r _ e n t e r e d = = 0 2 & & W a n t e d F lo o r < 0 2W a n t e d F lo o r > 0 2E ls e
2 9 1
R o u t e
2 1 8
S e iz e
2 9 2
R o u t e
2 1 9
S e iz e
2 9 3
R o u t e
2 2 0
S e iz e1 4 5
R e le a s e
1 4 6
R e le a s e2 9 4
R o u t e
6 9A s s ig n
7 0A s s ig n
7 1A s s ig n
7 2A s s ig n
7 3A s s ig n
7 4A s s ig n
7 5A s s ig n
7 6A s s ig n
7 7A s s ig n
7 8A s s ig n
7 9A s s ig n
8 0A s s ig n
8 1A s s ig n
8 2A s s ig n
8 3A s s ig n
8 4A s s ig n
8 5A s s ig n
8 6A s s ig n
8 7A s s ig n
8 8A s s ig n
8 9A s s ig n
9 0A s s ig n
9 1A s s ig n
9 2A s s ig n
9 3A s s ig n
9 4A s s ig n
9 5A s s ig n
9 6A s s ig n
9 7A s s ig n
9 8A s s ig n
9 9
A s s ig n
1 0 0A s s ig n
1 0 1A s s ig n
1 0 2A s s ig n
1 0 3A s s ig n
1 0 4A s s ig n
1 0 5A s s ig n
1 0 6A s s ig n
1 0 7A s s ig n
1 0 8A s s ig n
1 0 9A s s ig n
D e c id e 7 3
W a n t e d F lo o r = = 3 8
E ls e
1 4 7
R e le a s e
2 9 5
R o u t e
0 1F lo o r
A r r iv a lP a s s e n g e r
1 5 5
S t a t io n2 9 6
R o u t e
1 1 0A s s ig n
D e c id e 7 4
F lo o r _ e n t e r e d = = 0 1 & & W a n t e d F lo o r > 0 1
E ls e
2 9 7R o u t e
2 2 1S e iz e
4 4C r e a t e
1 1 1A s s ig n
1R e c o r d
H o ld 12
D is p o s e
1 1 2A s s ig n
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
Figure F2 – Overall Model Used for Stair Simulation
283
Appendix G
ARENA Results
Table G1: Lift Waiting Time, tLW (100 replications)
Lift Waiting Time-100% population Lift Waiting Time-25% population
2nd
floor fire 19th
floor fire 38th
floor fire 2nd
floor fire 19th
floor fire 38th
floor fire
717.06 707.48 680.26 313.37 298.74 343.14
725.01 722.97 718.92 312.24 246.39 244.45
658.28 644.88 675.10 335.08 264.66 272.30
770.99 804.44 766.92 302.16 296.53 287.19
760.50 653.93 725.87 259.76 304.24 275.68
601.69 679.86 678.05 211.23 276.04 266.97
731.41 743.00 705.11 303.64 316.57 311.80
731.52 751.88 658.73 296.09 243.80 298.96
705.94 683.51 689.57 273.11 245.58 253.95
667.20 727.76 726.46 301.28 325.45 269.07
695.39 720.55 813.62 283.76 299.83 283.49
702.97 656.49 804.02 322.37 311.53 348.73
672.46 677.19 715.59 265.54 279.50 277.75
754.20 640.13 718.01 299.77 309.55 307.19
669.93 729.94 685.37 306.17 299.44 246.41
716.58 708.03 630.57 308.97 267.09 300.64
700.94 783.57 800.97 283.03 307.39 265.46
800.59 733.33 764.56 345.33 351.36 309.30
696.85 659.46 688.20 259.62 295.17 315.40
721.70 625.47 658.39 321.09 289.32 273.38
703.36 737.21 788.47 264.79 323.27 304.78
681.01 747.12 660.68 344.61 327.18 315.63
734.25 760.89 715.62 345.32 320.99 303.44
733.91 746.61 706.69 368.90 317.54 335.41
723.56 716.90 700.41 326.93 268.60 273.90
722.59 792.07 743.59 284.95 267.11 380.21
789.46 697.49 800.16 326.82 301.01 312.27
710.00 763.75 796.73 292.52 304.03 258.58
757.26 732.68 702.82 378.35 305.61 332.54
741.60 732.17 739.09 343.29 301.91 378.46
676.32 732.37 694.68 247.31 329.95 301.63
721.48 758.92 777.81 324.62 325.85 337.77
723.91 766.84 745.47 352.79 321.28 285.83
699.70 717.51 766.56 305.60 318.27 328.38
743.45 753.32 747.34 365.18 304.96 323.90
737.23 810.20 728.20 302.93 260.89 266.89
704.74 682.95 647.58 424.95 303.55 292.86
723.70 696.30 704.78 251.19 281.30 319.17
716.62 724.86 687.46 295.39 283.77 309.67
671.66 710.74 801.20 307.98 285.45 299.79
701.37 711.95 679.68 310.75 262.04 254.31
733.87 741.76 771.82 321.03 248.50 301.17
801.11 798.97 727.75 320.67 282.36 315.73
284
775.36 758.59 692.43 305.14 336.14 325.10
795.21 805.87 727.97 301.11 387.43 320.72
778.93 774.57 728.59 276.53 232.71 299.75
672.04 735.83 756.77 310.32 326.09 351.58
748.42 758.38 695.23 271.16 252.91 314.67
773.02 684.56 757.03 289.44 285.53 327.43
701.26 670.80 719.57 250.85 266.40 328.78
767.12 739.28 750.75 361.09 309.77 260.17
744.89 680.45 804.48 264.26 284.56 320.59
775.04 746.46 839.67 257.48 361.35 316.69
716.03 724.70 712.29 259.36 294.52 297.74
770.43 746.73 690.23 280.91 332.62 286.41
726.98 708.12 691.51 251.51 267.15 253.86
827.20 702.55 749.26 293.79 357.46 307.16
717.48 699.50 714.20 293.18 256.43 257.46
639.08 737.43 744.89 248.45 294.64 403.91
708.36 710.21 731.75 286.54 252.37 323.18
721.75 713.94 693.47 245.33 249.17 242.10
691.20 669.66 735.78 265.54 324.78 281.92
677.66 743.48 745.23 279.27 308.13 265.23
709.14 702.62 720.63 260.34 337.19 270.83
678.89 706.64 676.16 313.17 280.54 289.55
663.70 736.38 744.95 288.58 240.41 299.71
682.50 714.74 709.72 351.29 324.11 249.33
724.83 736.44 705.94 324.07 374.12 327.41
693.76 631.04 725.52 273.63 244.73 284.36
800.03 610.03 706.41 316.83 317.17 290.15
701.96 726.58 760.93 292.17 363.09 267.22
729.07 721.77 717.04 268.67 324.04 288.30
725.58 729.48 712.90 300.11 265.53 279.31
691.39 728.75 730.41 313.85 271.47 305.21
719.16 709.72 773.34 289.39 317.90 332.44
770.10 731.21 701.21 377.65 242.89 241.85
694.05 672.24 730.05 374.72 373.00 324.86
737.17 695.17 714.54 306.79 344.81 321.72
734.95 724.36 741.33 314.24 313.30 289.00
688.88 717.12 740.35 353.24 270.71 318.10
694.42 679.82 645.94 272.21 287.42 279.04
736.61 716.94 715.91 308.34 316.05 346.02
674.29 681.96 652.29 315.76 361.87 298.28
715.95 735.89 657.46 317.10 306.95 325.93
723.45 723.34 809.09 267.81 258.25 312.98
682.29 708.60 674.14 284.29 237.85 257.28
785.61 740.96 702.46 312.22 318.80 345.54
718.56 772.53 711.36 289.91 233.23 284.86
693.59 736.10 655.71 244.52 329.05 244.72
725.73 666.55 669.65 295.80 208.79 244.16
796.37 747.45 761.53 373.63 298.12 301.48
704.39 761.96 767.03 278.06 304.32 362.18
723.76 708.64 657.46 256.01 297.85 266.25
730.90 704.17 727.93 283.14 229.12 279.18
770.62 753.80 768.18 279.22 311.67 280.99
285
675.95 761.74 680.12 252.89 276.68 299.11
791.87 692.73 754.97 319.38 298.01 281.13
636.12 687.96 605.19 281.61 265.73 308.02
707.97 698.94 710.22 289.56 270.65 268.41
709.86 692.60 736.57 249.04 226.63 279.65
Table G2: Lift Transportation Time, tLT (100 replications)
Lift Transportation Time -100% population Lift Transportation Time - 25% population
2nd
floor fire 19th
floor fire 38th
floor fire 2nd
floor fire 19th
floor fire 38th
floor fire
42.68 40.06 42.90 42.28 37.96 39.35
41.45 42.37 42.62 41.29 39.84 41.81
41.85 39.95 41.22 38.32 38.73 39.59
43.60 43.79 41.72 38.08 39.11 38.78
42.92 41.32 42.06 42.13 40.44 40.37
40.96 41.55 41.04 39.68 40.89 39.88
41.54 42.03 40.66 39.83 40.76 39.76
42.98 42.74 40.37 40.48 40.80 39.04
42.23 42.78 42.40 39.84 42.93 40.23
40.83 41.56 41.27 39.21 43.00 40.06
43.14 42.24 40.96 40.14 38.98 38.08
40.56 39.80 41.74 37.81 41.64 40.06
41.21 44.30 44.47 40.33 40.00 41.50
43.80 41.90 41.91 39.40 43.36 40.92
41.57 42.08 40.73 37.76 39.45 40.07
41.86 43.10 40.91 40.50 40.61 40.68
40.60 43.20 42.71 39.63 40.29 40.50
42.59 42.34 41.06 41.56 40.37 38.69
41.47 42.37 42.90 40.65 43.24 38.74
41.57 40.37 40.51 38.61 41.39 38.45
42.95 41.82 43.93 40.16 39.31 44.73
41.28 42.94 43.00 38.25 40.89 40.18
42.47 42.79 41.74 41.09 41.52 42.24
41.89 42.89 42.55 40.23 38.44 41.32
41.02 39.68 40.58 38.92 36.48 37.64
41.03 43.23 41.90 39.78 38.92 39.71
42.08 41.77 44.37 40.50 39.30 39.07
40.89 42.31 41.54 37.72 43.90 41.69
42.39 41.33 41.32 40.35 39.16 40.34
42.89 40.19 41.86 39.44 39.77 40.75
41.76 42.82 41.47 40.68 37.80 45.43
42.45 44.36 44.77 37.72 40.75 44.73
41.39 42.28 42.50 41.63 40.61 38.60
40.87 43.82 42.98 43.36 40.59 39.19
42.37 41.18 40.15 41.52 39.39 37.24
41.42 39.83 43.71 40.27 40.00 38.37
40.90 42.11 40.58 40.12 38.90 38.86
40.68 42.08 41.15 39.41 38.61 39.15
41.97 41.71 42.62 42.03 40.99 40.16
41.73 42.27 44.65 44.17 41.41 41.57
286
42.21 42.06 41.17 40.03 39.44 41.73
39.74 40.77 42.35 40.81 41.04 39.81
42.53 43.47 40.91 41.60 39.22 39.81
43.17 41.45 42.70 37.73 39.04 42.51
41.43 43.27 40.39 39.09 41.67 41.72
41.86 41.98 41.87 38.43 40.45 38.49
41.06 41.89 40.15 39.19 37.80 40.51
41.48 43.84 41.96 40.93 43.36 41.47
44.31 41.51 42.40 39.41 39.86 39.82
40.44 39.75 40.42 39.36 40.08 40.83
43.55 42.83 45.12 37.87 39.97 38.48
42.58 42.53 42.39 39.50 40.75 41.18
42.20 40.10 42.68 41.08 41.29 40.60
40.82 40.84 41.89 39.37 42.06 40.45
42.60 43.83 41.66 41.36 40.18 44.70
40.76 41.45 40.24 43.24 39.48 38.73
42.96 40.10 40.33 39.25 41.07 40.08
43.51 41.78 39.72 38.63 39.20 36.92
41.54 42.82 42.81 39.90 40.33 41.14
40.21 42.46 42.18 40.16 38.49 39.52
41.83 41.77 40.85 40.17 42.12 40.37
41.78 42.65 40.00 41.41 42.76 40.35
41.64 43.79 41.10 39.72 38.97 40.58
42.14 41.49 42.99 39.61 40.46 41.75
40.42 43.43 39.73 40.98 40.02 38.89
42.33 43.23 39.63 40.39 40.48 39.72
42.61 44.38 41.49 39.36 43.59 42.07
41.21 42.18 42.02 40.35 40.03 41.39
41.38 41.14 42.77 38.99 39.94 38.93
42.40 42.20 41.69 42.65 39.06 45.29
40.84 42.18 40.15 40.40 40.66 37.83
43.86 43.63 41.06 38.84 43.71 39.50
43.12 41.32 43.25 41.60 38.87 42.43
40.70 40.45 42.06 41.12 37.32 40.82
39.78 42.36 42.48 40.11 40.10 41.67
42.67 41.59 41.76 39.98 39.98 38.70
42.61 41.70 41.17 44.08 40.44 42.90
42.88 40.85 42.37 39.15 39.94 38.99
42.71 41.42 41.57 43.00 43.08 41.10
41.25 41.69 41.99 40.53 41.85 44.87
41.87 41.50 40.61 41.30 39.66 38.71
42.39 41.22 42.79 39.52 39.84 39.04
43.04 42.38 40.28 40.44 41.50 40.06
41.75 41.08 41.59 39.41 41.22 39.50
40.39 41.18 43.20 38.46 41.32 38.69
40.25 40.83 42.20 38.72 38.06 38.75
43.12 42.14 42.12 41.83 40.54 40.44
41.75 41.46 41.07 40.31 40.83 39.72
40.45 42.59 41.88 39.01 39.32 40.86
40.15 40.76 40.37 38.62 40.22 39.78
43.25 42.45 43.84 39.44 41.74 39.84
40.66 43.23 44.16 41.03 38.86 41.55
287
41.28 39.50 41.63 38.91 42.25 40.08
42.55 41.82 43.27 37.55 39.75 40.31
43.91 41.96 44.63 36.62 38.29 38.86
42.44 40.95 41.58 40.44 38.93 39.03
40.13 40.81 41.28 39.73 40.56 40.71
40.56 41.38 39.58 39.25 37.67 41.10
40.17 40.60 40.10 40.07 39.70 43.34
42.77 41.64 42.11 44.14 41.61 41.60
Table G3: Lift Evacuation Time, tET (100 replications)
Lift Evacuation Time -100% population Lift Evacuation Time - 25% population
2nd
floor fire 19th
floor fire 38th
floor fire 2nd
floor fire 19th
floor fire 38th
floor fire
2748.98 1829.56 2398.57 1275.75 1479.93 1333.93
2635.33 2274.70 2688.76 1669.52 1535.24 1279.55
1404.26 1428.23 1374.50 1702.19 1461.67 1479.06
2228.93 2126.23 2649.70 1443.70 1065.72 1290.06
2803.96 2396.17 2379.92 1086.68 1194.01 1755.74
1265.64 1499.16 1796.80 1036.52 905.33 881.96
1451.18 2051.85 1901.64 1563.82 987.10 939.05
2698.57 2826.74 2647.91 1730.38 1839.23 1824.76
2124.91 2091.09 2447.28 1476.18 1667.07 1680.70
2347.72 2223.33 1536.06 1467.12 1174.27 926.69
1349.78 1532.20 2008.87 1066.32 1003.41 885.08
2173.08 2663.32 2711.77 2208.90 1635.43 1452.51
2345.70 1687.92 1641.86 947.56 1296.58 1682.40
1957.81 1736.60 1900.42 1313.38 1157.68 1195.61
2192.14 2091.93 2185.29 1352.64 1257.25 1123.65
3052.42 2526.62 2418.75 2319.05 1848.81 1773.11
2667.24 3151.79 3403.47 2144.00 1806.90 1904.40
2918.06 2931.97 2815.44 1921.98 1715.97 2289.45
2611.97 2723.74 2256.57 1910.74 1940.82 2179.01
1642.97 1447.17 1331.00 1836.94 1972.33 1818.53
2098.70 2303.09 2195.80 1085.96 1145.53 1290.91
1598.25 1536.03 1450.76 1300.11 1002.93 1943.75
1450.30 2834.46 2865.11 1971.33 1274.44 1309.57
2004.97 1988.23 2291.11 1187.24 1073.05 1132.88
2432.96 2461.74 2473.95 1698.10 1877.70 1480.63
2369.74 2888.65 3004.48 1787.52 1972.95 2039.00
2260.39 2078.97 2162.86 1555.88 1418.15 1023.99
1158.68 1682.84 1688.64 1104.77 1793.97 1048.06
1998.88 2083.99 1784.71 1688.79 1626.60 1216.72
2724.10 3218.29 2840.22 1845.61 1637.61 1926.93
2860.65 3259.40 3000.33 2197.94 1706.69 2229.96
2688.81 1763.12 2199.51 1292.89 1722.17 1772.07
2058.49 2183.99 1180.71 1134.09 929.34 1101.41
2292.07 2073.96 2075.12 982.78 1319.35 876.05
2618.82 1782.50 2468.49 1517.73 933.53 979.03
2449.68 1294.04 2090.32 879.14 1326.09 1088.38
3123.82 2729.22 3176.55 2083.59 1956.05 1784.64
288
2100.82 2856.92 1696.81 1314.72 1585.28 1380.83
2968.50 3134.87 3119.61 2189.22 1677.14 1475.24
2791.37 2401.12 2960.13 1792.10 2059.58 1770.51
3038.19 2821.86 2889.85 2008.04 1620.32 1675.65
2810.35 2783.68 2706.93 1892.29 1550.87 1807.92
1414.38 1984.56 1398.76 1604.93 1754.31 1308.60
3169.29 3146.26 3142.99 1996.30 1817.96 1880.70
2327.73 1528.70 2028.06 1311.58 2136.79 1960.63
1898.25 2163.22 1290.36 1664.52 1749.13 1007.37
2742.92 2689.70 2576.40 1700.24 1834.16 1875.44
2999.73 2980.68 2725.57 1948.80 1654.67 2238.84
2688.22 3205.78 2872.52 1802.84 1772.29 1783.66
2548.78 1591.29 1860.39 1744.16 1165.15 973.96
2346.99 1684.07 1603.14 1779.27 2051.86 1143.36
2122.99 1922.12 1456.00 1017.99 1713.43 1379.04
2378.67 1375.47 1459.78 1551.53 2401.98 1667.39
2328.07 2230.80 1389.28 1450.69 1491.30 1422.67
2574.80 2640.07 3125.31 2242.60 1822.34 1626.70
2926.59 3021.87 2870.62 1720.54 1559.02 1930.16
2598.03 2788.79 2557.93 1515.18 1663.16 2200.65
2071.92 1983.43 1615.47 1629.82 1226.48 1360.20
2779.99 2766.19 3035.95 1968.01 1715.24 2154.41
3053.41 2859.22 2651.62 1999.52 1858.62 1912.35
2171.84 2504.12 2235.32 1218.39 1245.64 1338.76
1930.69 1634.92 1832.97 1979.95 1675.04 1274.52
2599.34 1815.35 1727.01 1635.20 1631.10 1263.74
1821.21 1330.27 1514.86 1572.93 1760.72 1813.69
2457.27 2825.51 2638.22 2085.83 1411.87 1449.43
2868.34 2761.14 2921.05 1667.21 1496.94 1755.88
2846.87 2823.32 2974.45 2003.51 1846.52 2051.28
2087.56 1749.80 1936.19 1514.57 1021.52 1196.39
1499.52 1358.98 1603.47 1782.39 1953.74 1677.68
1356.48 1176.37 1302.67 1829.33 2170.80 2206.20
3080.85 2548.14 2871.00 1308.35 1887.17 1908.06
1714.32 2511.72 2543.82 1613.34 1163.76 1201.35
2668.95 2747.69 2922.56 1858.40 2040.17 1899.46
1484.19 1539.31 1487.31 1721.73 1886.33 2306.55
1348.12 2004.62 1366.32 1833.20 2076.86 2044.47
3109.78 3291.86 2652.86 1983.47 1632.10 1599.81
1605.60 1740.47 2125.40 1010.74 1365.36 1097.46
1997.76 2056.48 1913.77 987.61 1354.51 1717.49
2388.80 2414.14 1919.91 1461.63 2007.37 1269.63
2682.85 2466.06 3053.14 2160.09 1988.92 1841.18
2622.52 2784.52 2436.25 1512.33 1951.86 1756.12
2611.89 2521.68 2644.98 2216.76 1962.19 2128.72
2156.90 1912.13 2399.05 1123.52 742.13 874.57
3082.70 2567.67 2902.78 2051.41 2059.86 2132.75
1795.42 1993.60 1419.23 1585.41 1361.48 1403.25
2043.86 2397.74 2325.84 1314.55 1301.58 1431.19
1948.83 1879.29 2765.37 1922.20 1364.03 1366.65
3008.02 2677.97 2430.14 1435.40 1768.96 1548.09
2123.80 2948.91 2426.60 918.71 1561.59 912.21
289
1625.44 1777.03 1613.06 1199.74 706.84 993.96
2703.63 2619.85 2690.75 1983.81 2096.21 1720.44
2043.86 1941.13 2213.71 1125.29 1194.34 1398.99
2628.61 2580.49 2889.92 1304.82 1804.65 1732.62
1636.04 1318.32 1496.65 1864.18 1852.78 2146.17
2598.83 2449.73 2598.38 1732.65 1773.12 1384.40
2892.66 3124.70 2805.89 2138.55 1568.79 2293.19
1592.05 2234.02 2470.24 785.97 1630.34 1216.28
2330.46 1533.93 1662.40 1608.34 1404.90 1638.59
3145.59 2566.52 3167.82 2052.15 2118.76 1914.47
2372.74 1836.73 1622.27 1584.53 1273.48 1312.46
Table G4: Stair Travelling Time, tST (100 replications)
Stair Travelling Time (100% Population) Stair Travelling Time (75% Population)
2nd
floor fire 19th
floor fire 38th
floor fire 2nd
floor fire 19th
floor fire 38th
floor fire
374.24 374.19 373.41 365.31 371.29 372.29
368.62 369.37 373.11 371.86 364.79 368.14
374.29 367.4 371.39 366.47 366.17 368.85
371.9 378.32 375.11 366.01 368.78 365.33
372.24 373.46 371.75 366.91 365.05 362.42
373.57 370.77 372.24 366.39 365.41 365.23
372.85 370.9 376.57 369.31 365.34 363.38
375.04 374.28 371.39 366.83 366.99 363.63
371.03 371.39 375.62 366.45 364.10 365.43
372.67 372.23 370.67 370.40 367.70 371.09
373.71 372.65 376.57 367.25 369.05 368.52
370.36 372.98 368.72 368.18 366.86 367.61
372.88 370.75 375.17 367.10 365.18 363.92
372.74 375.48 373.18 370.29 366.37 370.35
372.76 371.92 370.58 366.83 367.17 364.99
372.79 376.13 373.01 366.39 365.28 367.41
372.11 372.79 372.81 369.44 369.59 368.78
371.82 372.26 370.56 365.03 366.99 364.34
371.08 371.58 376.14 365.22 369.45 367.52
375.48 375.42 374.99 362.63 364.57 364.39
368.95 370.02 373.91 366.12 367.97 364.08
373.75 375.1 376.27 368.08 368.19 365.72
369.64 373.19 374.08 366.36 367.20 368.53
372.81 372.78 372.31 369.30 370.70 366.44
372.32 374.68 372.26 365.09 365.52 370.40
370.92 375.87 372.79 367.32 367.06 365.96
372.87 373.41 374.45 372.44 364.39 367.91
372.17 373.65 372.93 367.89 368.77 369.58
377.12 374.45 376.18 368.16 372.06 369.15
371.56 376.31 375.45 368.90 369.13 368.47
375.69 373.25 378.1 367.19 369.41 373.72
377.34 372.14 371.68 371.12 366.73 364.89
372.6 369.02 371.9 368.55 367.23 370.39
290
375.32 371.65 371.48 366.54 366.05 367.97
373.17 370.19 372.93 369.56 368.68 364.25
372.06 375.81 373.75 368.53 365.83 367.42
372.15 370.58 371.41 367.52 364.69 364.35
377.32 371.46 374.22 367.32 366.27 367.08
371.44 371.97 373.88 364.95 366.03 364.79
371.41 370.01 375.91 368.04 368.93 369.43
376.85 378.49 374.88 368.11 369.01 370.43
371.68 370.42 371.55 365.89 367.42 367.27
373.21 374.03 370.46 368.44 368.20 365.87
374.27 372.88 371.92 368.80 367.56 366.18
371.81 372.86 369.54 368.56 364.10 366.57
373.52 372.85 371.39 370.16 366.93 363.73
376.24 371.68 374.7 368.47 367.42 367.17
372.45 374.76 369.88 367.91 366.22 366.41
373.65 372.39 373.13 366.04 368.74 364.77
374.3 372.31 371.54 367.40 363.40 369.63
373.79 379.44 375.17 367.41 368.03 369.43
375.9 373.15 370.29 369.39 370.25 367.29
369.93 369.18 374.83 366.87 366.10 365.80
373.36 374.83 377.51 363.41 368.63 373.26
372.12 374.58 370.89 373.80 370.11 370.60
375.08 371.69 372.18 369.06 368.33 370.39
375.39 374.92 374.87 368.38 368.09 367.36
370.62 369.59 372.02 367.49 367.14 365.19
371.99 374.55 376.25 366.76 365.87 366.00
369.96 376.19 373.9 367.56 364.74 368.81
370.58 372.09 371.92 365.23 366.22 367.88
375.13 369.39 368.72 368.39 365.88 367.71
371.29 369.89 367.12 369.42 366.98 366.04
375.87 370.05 374.14 368.15 365.85 366.22
373.77 376.02 371.9 371.71 366.90 365.78
369.79 372.59 375 365.48 365.06 364.92
372.59 373.23 375.79 364.84 366.75 369.43
370.04 374.98 371.63 367.68 365.73 366.90
372.39 372.91 373.45 366.15 364.41 364.90
371 370.12 368.2 366.84 364.95 367.81
373.79 372.33 371.82 369.40 364.96 363.41
377.54 369.11 366.3 365.42 365.57 367.31
372.79 372.08 378.38 369.27 366.27 366.11
373.63 371.52 373.62 369.45 367.35 367.83
375.88 371.54 373.34 370.87 368.04 366.80
371.33 376.03 368.16 368.94 365.57 366.05
371.29 370.31 370.96 369.15 367.56 366.56
367.77 374.34 374.07 365.43 364.68 369.97
376.83 376.43 371.89 368.57 365.64 370.82
372.54 371.96 373.1 368.59 369.77 370.11
374.36 373.59 372.92 367.61 366.67 364.74
375.55 369.57 369.75 367.50 368.10 365.58
291
373.43 376.45 374.76 370.26 366.63 368.70
372.65 370.39 373.31 369.81 366.49 363.45
375.7 369.8 372.78 365.32 367.88 362.99
373.37 371.13 374.72 368.75 368.19 365.80
371.48 375.42 372.78 365.73 368.38 373.34
371.02 368.64 370.52 367.52 364.92 368.54
376.64 375.77 374.81 365.72 368.26 366.40
371.31 375.35 373.58 366.73 367.23 366.47
374 368.81 370.5 367.95 366.83 369.19
373.76 372.11 369.87 366.74 369.14 365.43
377.9 371.11 374.06 367.67 368.16 370.41
375.41 374.95 372.54 370.21 365.49 369.01
370.33 371.47 372.15 366.66 365.77 367.37
377.14 371.43 370.86 369.49 367.34 367.98
372.34 371.67 368.88 367.10 363.99 363.73
372.12 372.92 373.33 367.71 367.04 364.04
368.67 370.47 372.04 365.34 366.48 364.34
377.86 370.06 372.79 366.19 367.84 368.10
Table G5: Stair Evacuation Time, tSE (100 replications)
Stair Evacuation Time (100% Population) Stair Evacuation Time (75% Population)
2nd
floor fire 19th
floor fire 38th
floor fire 2nd
floor fire 19th
floor fire 38th
floor fire
1812.73 1731.27 1799.73 1476.53 1219.15 1540.93
1717.4 1936.55 1920.57 1204.92 1246.38 1623.37
1681.98 1321.1 1796.19 1187.36 1223.95 1520.77
1488.45 926.1 909.4 1204.72 1186.40 1446.89
1165.35 1102.33 822.04 1174.28 1225.69 1234.67
1711.33 1947.65 1676.86 1273.49 1210.89 1477.21
1827.44 1761.46 1826.44 1226.55 1265.24 1462.48
1829.56 1740.89 1605.64 1217.78 1232.12 1461.53
1105.69 1132.48 1314.98 820.43 705.61 779.43
1433.73 1287.19 1317.24 906.18 988.71 1146.36
1078.54 1587.65 1609.51 1036.58 1002.47 1100.91
1779.85 1734.67 1497.09 1219.89 1257.19 1359.00
1797.3 1730.81 2002.17 1244.41 1332.18 1431.55
1431.76 1743.24 1782.24 1242.17 1336.34 1618.81
1644.63 1766.27 1711.58 1199.76 1318.45 1439.13
1830.24 1857.11 1807.4 1212.00 1187.24 1470.11
1649.33 1618.56 1610.44 1168.81 1263.72 1428.99
1813.49 2021.96 1773.55 1238.98 1234.27 1622.78
1807.62 1609.3 1770.46 1353.38 1215.94 1438.25
1657.57 1695.44 1709.49 1158.86 1207.65 1572.65
1842.99 1595.22 1790.21 1237.21 1384.31 1278.49
1526.73 1641.66 1869.02 1261.29 1401.78 1275.23
1596.84 1931.41 1834.06 1196.19 1213.72 1582.10
1569.91 1632.72 1694.72 1296.11 1202.82 1588.84
1270.52 937.93 1196.22 854.19 833.11 1048.02
1337.3 1283.97 1550.94 787.94 731.12 858.91
292
1786.05 1607.7 1669.21 1217.97 1418.73 1408.21
1756.07 1871.31 1718.2 1232.38 1368.74 1310.75
1846.28 1823.98 1847.1 1402.89 1316.26 1267.93
946.26 1178.67 1227.99 805.02 847.82 755.34
1804.8 1784.13 1782.68 1249.71 1401.90 1586.49
1540.61 1314.18 1438.08 810.81 1276.07 936.90
1502.74 1549.45 1695.18 1080.03 1152.09 1322.43
1679.7 1708.54 1456.99 1177.35 1228.32 1286.19
1581.98 928.38 1099.34 1217.03 1215.32 1473.08
1220.8 1616.96 1585.92 1066.02 1065.56 982.34
800.35 1186.54 1417.15 901.95 952.96 921.10
1440.84 1446.33 1380.63 651.48 763.87 937.98
1532.72 1757.8 1784.9 1170.27 1177.89 1217.45
1528.53 1233.08 1407.8 878.89 916.11 1008.60
1779.33 1726.49 1847.52 1282.20 1258.69 1701.60
1579.68 1306.61 1453.33 951.37 951.34 1329.33
1687.17 1737.22 1799.89 1195.00 1333.04 1230.14
1772.34 1649.96 1678.97 1178.30 1411.98 1271.66
1471.23 1671.28 1576.19 1032.91 1292.43 1262.42
1132.14 1386.38 1643.8 1026.24 1056.20 1068.65
1265.19 1023.95 832.81 723.05 770.22 1010.43
1783.24 1753.82 1760.4 1223.47 1234.07 1551.95
1900.23 1813.73 1743.49 1238.33 1327.16 1579.26
1743.68 1789.07 1777.84 1217.58 1257.39 1671.25
1738.2 1920.56 1775.94 1222.56 1228.13 1494.19
921.73 1434.89 1369.77 838.07 983.13 897.27
1726.35 1590.21 1761.8 1146.66 1188.74 1285.84
1103.01 1101.83 1243.4 867.18 865.40 1018.33
1551.73 1795.99 1682.55 1181.77 1247.20 1533.23
1307.59 1364.29 1358.53 737.47 932.28 856.36
1061.4 1067.93 1204.06 1292.50 1252.30 1432.23
1158.44 1527.86 1374.3 1022.88 1044.51 1190.33
1299.32 1642.9 1544.1 992.79 992.75 988.84
884.51 1066.58 1030.77 1154.74 1323.33 1444.44
1716.23 1755.84 1876.95 1203.21 1224.78 1581.35
1854.11 1714.04 1657.84 1216.90 1245.85 1498.74
1947.13 1687.36 1757.79 1260.86 1424.52 1599.75
1950.26 1895.88 1716.87 1187.65 1329.41 1505.14
1643.46 1838.51 1514.96 1187.85 1462.94 1277.18
904.69 1209.38 870.81 884.45 898.58 906.43
1453.16 1170.31 1342.15 950.30 968.35 1275.96
1466.16 1763.61 1640.3 1115.24 1131.81 1342.11
1283.14 951.58 1422.12 945.66 956.28 1295.34
1392.56 1394.79 1527.45 1051.17 1054.40 1358.31
911.44 1562.76 1775.24 1210.55 1252.58 1264.59
1649.44 917.38 964.42 816.24 836.97 925.60
1828.68 2157.91 1673.66 1176.94 1289.24 1291.78
1728.43 933.48 1813.49 1222.37 1389.64 1472.43
1288.82 1099.92 953.29 729.12 749.48 878.47
293
1831.43 1565.72 1540.54 1162.66 1182.41 1454.51
852.84 1229.04 1425.11 815.73 854.38 1010.00
1282.06 995.86 846.67 1193.62 1211.64 1487.70
1846.93 1850.2 1872.71 1254.07 1265.70 1640.43
2048.76 1844.99 1813.83 1249.79 1291.67 1502.21
1549.31 1167.29 890.19 875.89 681.79 1010.70
1689.54 1471.84 1606.69 1217.04 1342.32 1400.00
1457.25 1741.25 1740.36 1195.50 1241.08 1494.85
1334.15 989.57 1144.95 619.42 731.00 672.45
1664.12 1720.43 1799.59 1192.71 1198.43 1487.35
1833.19 2002.72 1778.73 1157.91 1294.38 1414.68
1737.25 1846.32 1559.62 1207.21 1163.85 1375.09
1505.17 1823.33 1820.77 1225.37 1306.26 1638.20
2114.19 1605.02 1605.39 1244.24 1220.67 1389.25
1777.25 1601.88 1785.26 1178.83 1370.89 1391.34
1792.84 1695.71 1673.96 1185.76 1301.11 1535.04
1774.57 1372.89 1524.51 1249.50 1260.91 1360.70
1783.22 1754.42 1775.94 1228.15 1245.58 1639.45
1949.24 1832.03 1622 1214.91 1344.33 1387.66
1692.48 1709.71 1736.39 1127.12 1155.42 1371.82
1502.1 1107.54 1401.84 709.69 989.69 940.32
1812.81 1863.41 1656.92 1222.18 1231.59 1489.96
1730.7 1798.44 1954.55 1192.57 1212.97 1527.18
1058.44 1388 1109.07 1202.90 1201.40 874.57
1723.27 1728.74 1599.02 1182.78 1196.98 1484.52
Table G6 – Mean Lift Time Periods (average values of three levels)
Lift Waiting
Time (100%)
Lift
Transportation
Time (100%)
Lift
Evacuation
Time (100%)
Lift
Waiting
Time (25%)
Lift
Transportation
Time (25%)
Lift
Evacuation
Time (25%)
701.60 41.88 2325.70 318.42 39.86 1363.20
722.30 42.15 2532.93 267.69 40.98 1494.77
659.42 41.00 1402.33 290.68 38.88 1547.64
780.79 43.04 2334.95 295.30 38.66 1266.49
713.43 42.10 2526.68 279.89 40.98 1345.47
653.20 41.18 1520.53 251.41 40.15 941.27
726.51 41.41 1801.56 310.67 40.12 1163.32
714.04 42.03 2724.41 279.62 40.11 1798.12
693.01 42.47 2221.10 257.55 41.00 1607.98
707.14 41.22 2035.70 298.60 40.76 1189.36
743.19 42.11 1630.29 289.03 39.07 984.94
721.16 40.70 2516.06 327.54 39.84 1765.61
688.41 43.32 1891.82 274.26 40.61 1308.85
704.11 42.53 1864.95 305.50 41.23 1222.22
695.08 41.46 2156.45 284.01 39.09 1244.51
685.06 41.96 2665.93 292.23 40.60 1980.32
761.83 42.17 3074.17 285.29 40.14 1951.77
766.16 42.00 2888.49 335.33 40.21 1975.80
294
681.50 42.25 2530.76 290.06 40.88 2010.19
668.52 40.82 1473.71 294.60 39.48 1875.93
743.01 42.90 2199.19 297.61 41.40 1174.13
696.27 42.41 1528.35 329.14 39.77 1415.60
736.92 42.34 2383.29 323.25 41.62 1518.45
729.07 42.45 2094.77 340.61 39.99 1131.06
713.62 40.43 2456.22 289.81 37.68 1685.48
752.75 42.05 2754.29 310.76 39.47 1933.16
762.37 42.74 2167.41 313.37 39.62 1332.68
756.82 41.58 1510.05 285.04 41.10 1315.60
730.92 41.68 1955.86 338.83 39.95 1510.70
737.62 41.65 2927.54 341.22 39.98 1803.38
701.12 42.02 3040.13 292.96 41.30 2044.87
752.74 43.86 2217.15 329.41 41.07 1595.71
745.41 42.06 1807.73 319.97 40.28 1054.95
727.93 42.56 2147.05 317.42 41.05 1059.39
748.04 41.23 2289.94 331.35 39.38 1143.43
758.54 41.65 1944.68 276.90 39.54 1097.87
678.42 41.19 3009.86 340.45 39.29 1941.43
708.26 41.30 2218.18 283.89 39.06 1426.94
709.65 42.10 3074.33 296.28 41.06 1780.53
727.87 42.89 2717.54 297.74 42.38 1874.06
697.66 41.81 2916.63 275.70 40.40 1768.00
749.15 40.95 2766.99 290.24 40.56 1750.36
775.94 42.30 1599.23 306.25 40.21 1555.94
742.13 42.44 3152.85 322.13 39.76 1898.32
776.35 41.70 1961.49 336.42 40.83 1803.00
760.70 41.90 1783.94 269.67 39.12 1473.67
721.55 41.03 2669.68 329.33 39.17 1803.28
734.01 42.43 2901.99 279.58 41.92 1947.44
738.20 42.74 2922.17 300.80 39.69 1786.26
697.21 40.20 2000.15 282.01 40.09 1294.42
752.38 43.83 1878.07 310.35 38.77 1658.17
743.27 42.50 1833.70 289.80 40.48 1370.15
787.06 41.66 1737.97 311.84 40.99 1873.63
717.67 41.18 1982.72 283.87 40.62 1454.89
735.80 42.70 2780.06 299.98 42.08 1897.21
708.87 40.82 2939.69 257.51 40.49 1736.57
759.67 41.13 2648.25 319.47 40.13 1793.00
710.39 41.67 1890.27 269.02 38.25 1405.50
707.13 42.39 2860.71 315.67 40.46 1945.89
716.77 41.62 2854.75 287.36 39.39 1923.49
709.72 41.48 2303.76 245.53 40.89 1267.60
698.88 41.48 1799.53 290.74 41.51 1643.17
722.12 42.18 2047.23 284.21 39.76 1510.01
710.80 42.21 1555.45 289.45 40.61 1715.78
687.23 41.19 2640.34 294.42 39.96 1649.04
715.01 41.73 2850.18 276.23 40.20 1640.01
702.32 42.82 2881.55 308.24 41.67 1967.10
295
722.40 41.80 1924.52 341.87 40.59 1244.16
683.44 41.76 1487.32 267.57 39.29 1804.61
705.49 42.10 1278.50 308.05 42.33 2068.78
729.82 41.06 2833.33 307.49 39.63 1701.20
722.63 42.85 2256.62 293.67 40.68 1326.15
722.65 42.56 2779.73 281.65 40.97 1932.68
716.85 41.07 1503.60 296.84 39.75 1971.54
734.07 41.54 1573.02 313.24 40.63 1984.85
734.18 42.01 3018.17 287.46 39.55 1738.46
698.78 41.83 1823.82 357.53 42.47 1157.86
715.62 42.03 1989.34 324.44 39.36 1353.20
733.55 41.90 2240.95 305.51 42.39 1579.54
715.45 41.64 2734.02 314.02 42.42 1996.73
673.39 41.33 2614.43 279.56 39.89 1740.10
723.15 42.14 2592.85 323.47 39.47 2102.56
669.51 41.90 2156.02 325.30 40.67 913.41
703.10 41.47 2851.05 316.66 40.04 2081.34
751.96 41.59 1736.08 279.68 39.49 1450.05
688.34 41.09 2255.81 259.81 38.51 1349.11
743.01 42.46 2197.83 325.52 40.94 1550.96
734.15 41.43 2705.38 269.34 40.28 1584.15
695.13 41.64 2499.77 272.76 39.73 1130.84
687.31 40.42 1671.84 249.58 39.54 966.85
768.45 43.18 2671.41 324.41 40.34 1933.49
744.46 42.69 2066.23 314.85 40.48 1239.54
696.62 40.80 2699.68 273.37 40.41 1614.03
721.00 42.55 1483.67 263.81 39.21 1954.38
764.20 43.50 2548.98 290.63 37.92 1630.06
705.94 41.65 2941.08 276.23 39.47 2000.18
746.52 40.74 2098.77 299.51 40.34 1210.86
643.09 40.51 1842.26 285.12 39.34 1550.61
705.71 40.29 2959.98 276.21 41.04 2028.46
713.01 42.17 1943.92 251.78 42.45 1390.16
296
Table G7 – Mean Stair Time Periods (average values of three levels)
Stair Travelling
Time (100%
population)
Stair Evacuation
Time (100%
population)
Stair Travelling
Time (75%
population)
Stair Evacuation
Time (75%
population)
373.95 1781.24 369.63 1412.20
370.37 1858.17 368.26 1358.22
371.03 1599.76 367.16 1310.69
375.11 1107.98 366.70 1279.34
372.48 1029.91 364.79 1211.54
372.19 1778.61 365.68 1320.53
373.44 1805.12 366.01 1318.09
373.57 1725.36 365.82 1303.81
372.68 1184.38 365.33 768.49
371.86 1346.05 369.73 1013.75
374.31 1425.24 368.27 1046.65
370.69 1670.54 367.55 1278.69
372.93 1843.42 365.40 1336.04
373.8 1652.41 369.00 1399.11
371.75 1707.49 366.33 1319.11
373.98 1831.59 366.36 1289.79
372.57 1626.11 369.27 1287.17
371.55 1869.67 365.45 1365.34
372.93 1729.13 367.40 1335.86
375.3 1687.5 363.86 1313.05
370.96 1742.8 366.06 1300.00
375.04 1679.14 367.33 1312.76
372.3 1787.43 367.36 1330.67
372.63 1632.45 368.81 1362.59
373.08 1134.89 367.00 911.77
373.2 1390.74 366.78 792.66
373.58 1687.65 368.25 1348.30
372.91 1781.86 368.74 1303.96
375.92 1839.12 369.79 1329.03
374.44 1117.64 368.83 802.73
375.68 1790.54 370.11 1412.70
373.72 1430.96 367.58 1007.93
371.17 1582.46 368.72 1184.85
372.81 1615.07 366.85 1230.62
372.1 1203.23 367.50 1301.81
373.87 1474.56 367.26 1037.97
371.38 1134.68 365.52 925.34
374.34 1422.6 366.89 784.45
372.43 1691.81 365.25 1188.54
372.45 1389.8 368.80 934.53
376.74 1784.45 369.18 1414.16
371.22 1446.54 366.86 1077.35
372.57 1741.43 367.50 1252.73
297
373.02 1700.43 367.52 1287.31
371.4 1572.9 366.41 1195.92
372.59 1387.44 366.94 1050.36
374.21 1040.65 367.69 834.57
372.36 1765.82 366.85 1336.50
373.06 1819.15 366.52 1381.58
372.72 1770.2 366.81 1382.07
376.13 1811.57 368.29 1314.96
373.11 1242.13 368.98 906.16
371.31 1692.79 366.26 1207.08
375.23 1149.41 368.43 916.97
372.53 1676.75 371.50 1320.73
372.98 1343.47 369.26 842.04
375.06 1111.13 367.94 1325.67
370.74 1353.53 366.61 1085.91
374.27 1495.44 366.21 991.46
373.35 993.95 367.04 1307.50
371.53 1783.01 366.44 1336.45
371.08 1742 367.33 1320.50
369.43 1797.43 367.48 1428.38
373.35 1854.34 366.74 1340.73
373.9 1665.64 368.13 1309.32
372.46 994.96 365.15 896.49
373.87 1321.87 367.01 1064.87
372.22 1623.35 366.77 1196.39
372.92 1218.95 365.15 1065.76
369.77 1438.27 366.54 1154.63
372.65 1416.48 365.92 1242.57
370.99 1177.08 366.10 859.60
374.42 1886.75 367.22 1252.65
372.92 1491.8 368.21 1361.48
373.58 1114.01 368.57 785.69
371.84 1645.9 366.85 1266.53
370.86 1169 367.75 893.37
372.06 1041.53 366.69 1297.65
375.05 1856.62 368.34 1386.73
372.54 1902.53 369.49 1347.89
373.63 1202.26 366.34 856.13
371.62 1589.35 367.06 1319.79
374.88 1646.29 368.53 1310.48
372.12 1156.22 366.59 674.29
372.76 1728.05 365.40 1292.83
373.07 1871.55 367.58 1288.99
373.23 1714.4 369.15 1248.72
370.06 1716.43 366.99 1389.95
375.74 1774.87 366.79 1284.72
373.41 1721.46 366.81 1313.69
371.1 1720.84 367.99 1340.64
371.91 1557.32 367.10 1290.37
298
374.36 1771.19 368.75 1371.06
374.3 1801.09 368.24 1315.63
371.31 1712.86 366.60 1218.12
373.14 1337.16 368.27 879.90
370.96 1777.71 364.94 1314.58
372.79 1827.9 366.26 1310.91
370.39 1185.17 365.38 1092.96
373.57 1683.68 367.38 1288.10
Table G8: Floor-Wise Lift Waiting Time (fire on 2nd
floor)
Waiting Time Average Half
Width
Minimum
Average
Maximum
Average
Minimum
Value
Maximum
Value
Floor 01.Queue 26.16 2.15 10.63 69.01 0.01 238.00
Floor 02.Queue 486.32 63.37 55.74 1655.52 0.08 1868.33
Floor 03.Queue 627.53 74.25 43.60 1738.81 0.21 1982.33
Floor 04.Queue 641.85 75.50 74.58 1645.42 0.05 2046.17
Floor 05.Queue 656.92 73.66 62.88 1607.76 0.43 1964.06
Floor 06.Queue 710.06 79.71 88.35 1767.21 0.31 1922.58
Floor 07.Queue 615.65 65.51 96.61 1524.02 1.32 1631.03
Floor 08.Queue 636.33 78.24 68.31 1728.73 2.37 1988.79
Floor 09.Queue 711.80 62.15 106.15 1309.53 0.02 1928.64
Floor 10.Queue 750.00 72.97 110.13 1494.49 1.34 1895.74
Floor 11.Queue 726.09 72.96 120.95 1695.20 3.13 1908.29
Floor 12.Queue 695.72 69.37 70.16 1575.41 2.56 1900.89
Floor 13.Queue 700.95 78.47 82.85 1781.33 0.36 1937.83
Floor 14.Queue 728.34 78.07 64.73 1725.82 1.48 1901.43
Floor 15.Queue 743.67 74.18 120.17 1755.58 0.15 1840.10
Floor 16.Queue 762.71 77.52 66.63 1610.28 0.14 1817.28
Floor 17.Queue 739.79 72.67 190.18 1805.31 0.89 2015.29
Floor 18.Queue 765.46 79.78 98.85 1771.46 0.68 1891.45
Floor 19.Queue 895.56 81.42 75.46 1916.37 1.36 2089.71
Floor 20.Queue 823.65 74.56 125.69 1800.62 0.43 1945.83
Floor 21.Queue 803.23 78.91 86.21 2225.58 0.44 2298.87
Floor 22.Queue 684.25 75.60 92.28 1709.34 1.59 2000.60
Floor 23.Queue 812.56 87.66 102.08 1930.13 1.12 1998.80
Floor 24.Queue 811.48 83.38 53.13 1751.92 0.92 1914.64
Floor 25.Queue 754.58 77.20 109.11 1615.62 0.07 1892.65
Floor 26.Queue 723.81 80.21 62.70 1561.69 0.58 1980.61
Floor 27.Queue 740.11 78.04 79.74 1967.46 0.06 2064.60
Floor 28.Queue 623.84 77.22 49.74 1762.76 0.09 1899.19
Floor 29.Queue 588.45 74.40 28.44 1700.80 0.01 2039.30
Floor 30.Queue 590.60 83.71 42.79 1628.52 0.09 1895.64
Floor 31.Queue 598.33 81.84 32.10 1898.37 0.01 2011.71
Floor 32.Queue 645.16 77.50 29.27 1644.15 0.00 1838.97
Floor 33.Queue 646.00 72.76 36.30 1637.27 0.01 2074.18
Floor 34.Queue 737.30 81.05 38.78 1572.82 0.03 1978.26
Floor 35.Queue 904.95 80.96 136.98 1751.41 0.02 2206.09
Floor 36.Queue 887.52 77.12 97.37 1689.04 0.11 2195.09
Floor 37.Queue 983.78 86.23 115.76 2016.86 0.00 2289.98
Floor 38.Queue 1004.49 97.85 135.84 2319.36 0.55 2371.36
299
Table G9: Floor-Wise Number of Evacuees in Queue (fire on 2nd floor)
Number Waiting Average Half
Width
Minimum
Average
Maximum
Average
Minimum
Value
Maximum
Value
Floor 01.Queue 0.21 0.02 0.08 0.56 0.00 11.00
Floor 02.Queue 5.61 0.76 0.52 20.39 0.00 32.00
Floor 03.Queue 7.21 0.88 0.46 21.17 0.00 32.00
Floor 04.Queue 7.36 0.88 0.95 18.04 0.00 32.00
Floor 05.Queue 7.62 0.90 0.66 20.50 0.00 32.00
Floor 06.Queue 8.12 0.93 1.08 21.69 0.00 32.00
Floor 07.Queue 7.08 0.78 1.09 18.51 0.00 33.00
Floor 08.Queue 7.28 0.91 0.88 21.21 0.00 33.00
Floor 09.Queue 8.14 0.72 1.27 15.89 0.00 33.00
Floor 10.Queue 8.65 0.88 1.24 19.53 0.00 32.00
Floor 11.Queue 8.39 0.87 1.20 21.00 0.00 33.00
Floor 12.Queue 8.04 0.83 0.74 19.16 0.00 33.00
Floor 13.Queue 8.09 0.92 0.95 21.12 0.00 32.00
Floor 14.Queue 8.39 0.94 0.78 22.85 0.00 33.00
Floor 15.Queue 8.63 0.91 1.37 20.59 0.00 33.00
Floor 16.Queue 8.80 0.91 0.70 20.08 0.00 33.00
Floor 17.Queue 8.53 0.86 1.98 20.48 0.00 33.00
Floor 18.Queue 8.83 0.94 1.07 21.13 0.00 33.00
Floor 19.Queue 10.32 0.95 0.82 21.49 0.00 33.00
Floor 20.Queue 9.51 0.89 1.53 22.34 0.00 34.00
Floor 21.Queue 9.27 0.92 1.06 24.73 0.00 32.00
Floor 22.Queue 7.84 0.85 1.09 18.14 0.00 32.00
Floor 23.Queue 9.39 1.04 1.08 21.56 0.00 33.00
Floor 24.Queue 9.34 0.96 0.64 19.89 0.00 33.00
Floor 25.Queue 8.69 0.89 1.19 20.03 0.00 32.00
Floor 26.Queue 8.35 0.93 0.69 19.22 0.00 33.00
Floor 27.Queue 8.48 0.89 0.91 21.65 0.00 34.00
Floor 28.Queue 7.18 0.90 0.59 21.27 0.00 34.00
Floor 29.Queue 6.78 0.86 0.32 20.52 0.00 32.00
Floor 30.Queue 6.75 0.94 0.54 18.57 0.00 33.00
Floor 31.Queue 6.84 0.90 0.39 19.17 0.00 32.00
Floor 32.Queue 7.43 0.89 0.35 18.39 0.00 33.00
Floor 33.Queue 7.41 0.82 0.39 17.70 0.00 33.00
Floor 34.Queue 8.60 0.96 0.41 19.11 0.00 33.00
Floor 35.Queue 10.50 0.95 1.68 23.56 0.00 34.00
Floor 36.Queue 10.27 0.88 1.14 20.71 0.00 34.00
Floor 37.Queue 11.30 0.97 1.41 22.97 0.00 32.00
Floor 38.Queue 11.66 1.17 1.52 23.90 0.00 33.00
300
Table G10: Floor-Wise Lift Waiting Time (fire on 19th floor)
Waiting Time Average Half
Width
Minimum
Average
Maximum
Average
Minimum
Value
Maximum
Value
Floor 01.Queue 26.97 2.27 13.53 61.75 0.02 224.60
Floor 02.Queue 683.20 78.09 100.65 1759.93 1.61 2062.88
Floor 03.Queue 653.12 64.28 85.38 1404.80 0.21 1997.93
Floor 04.Queue 658.34 76.64 59.74 1625.67 0.36 1833.95
Floor 05.Queue 630.88 72.55 46.98 1808.77 0.50 2086.69
Floor 06.Queue 691.11 68.74 85.97 1666.60 0.34 1814.71
Floor 07.Queue 665.67 74.26 45.22 1683.63 0.45 1999.93
Floor 08.Queue 685.93 78.24 144.63 2096.00 2.29 2260.06
Floor 09.Queue 686.45 73.64 88.20 1689.76 1.04 2102.50
Floor 10.Queue 755.85 82.21 87.22 1783.25 0.64 2024.93
Floor 11.Queue 758.26 71.70 118.78 1723.59 0.87 2106.54
Floor 12.Queue 751.37 74.17 127.66 1547.48 0.65 2029.59
Floor 13.Queue 761.62 76.78 81.83 1901.58 0.92 2015.35
Floor 14.Queue 773.69 76.72 71.81 1831.71 0.89 2140.82
Floor 15.Queue 778.43 73.05 73.10 1750.03 0.15 2161.87
Floor 16.Queue 751.06 82.09 145.23 1950.21 0.54 2076.69
Floor 17.Queue 728.02 68.85 163.15 1699.30 2.58 1849.20
Floor 18.Queue 658.77 75.52 60.62 1705.65 1.65 1928.71
Floor 19.Queue 661.78 69.76 126.25 1508.59 0.17 1737.91
Floor 20.Queue 783.63 76.03 129.80 1733.56 0.54 2029.08
Floor 21.Queue 798.39 72.88 94.77 1583.46 1.43 1944.72
Floor 22.Queue 804.79 74.40 65.24 1837.08 0.34 1897.02
Floor 23.Queue 819.30 73.68 198.28 1788.42 2.58 1867.20
Floor 24.Queue 826.84 78.66 113.16 1859.88 1.36 2054.16
Floor 25.Queue 740.62 76.84 103.82 1633.84 0.24 1991.94
Floor 26.Queue 711.78 76.09 63.38 1588.32 0.18 1863.37
Floor 27.Queue 689.25 80.11 62.22 1692.42 0.48 1924.10
Floor 28.Queue 638.02 69.10 70.06 1481.95 0.05 1838.55
Floor 29.Queue 520.74 77.89 31.35 1494.43 0.03 1960.52
Floor 30.Queue 543.67 81.64 27.54 1832.21 0.02 1981.69
Floor 31.Queue 527.83 73.38 45.56 1662.53 0.03 1914.93
Floor 32.Queue 595.23 77.08 31.96 1783.42 0.06 2250.99
Floor 33.Queue 704.95 76.90 44.73 1650.98 0.16 2202.82
Floor 34.Queue 771.98 75.19 50.04 1597.23 0.03 2077.66
Floor 35.Queue 806.33 75.79 103.78 1601.22 0.04 2084.80
Floor 36.Queue 897.59 82.28 156.75 1946.84 0.11 2314.22
Floor 37.Queue 906.66 90.95 120.38 2063.47 0.15 2323.26
Floor 38.Queue 1077.01 89.29 151.29 2178.60 16.53 2402.26
301
Table G11: Floor-Wise Number of Evacuees in Queue (fire on 19th floor)
Number Waiting Average Half
Width
Minimum
Average
Maximum
Average
Minimum
Value
Maximum
Value
Floor 01.Queue 0.21 0.02 0.09 0.52 0.00 10.00
Floor 02.Queue 7.80 0.90 1.18 19.76 0.00 32.00
Floor 03.Queue 7.51 0.76 1.06 16.79 0.00 32.00
Floor 04.Queue 7.60 0.92 0.67 19.91 0.00 32.00
Floor 05.Queue 7.16 0.82 0.57 20.20 0.00 32.00
Floor 06.Queue 7.88 0.78 1.01 19.35 0.00 33.00
Floor 07.Queue 7.70 0.90 0.51 20.84 0.00 32.00
Floor 08.Queue 7.90 0.92 1.54 23.05 0.00 32.00
Floor 09.Queue 7.88 0.86 0.93 20.02 0.00 32.00
Floor 10.Queue 8.77 1.01 0.98 22.19 0.00 32.00
Floor 11.Queue 8.71 0.84 1.31 21.48 0.00 32.00
Floor 12.Queue 8.58 0.84 1.47 18.76 0.00 32.00
Floor 13.Queue 8.69 0.87 1.00 19.74 0.00 32.00
Floor 14.Queue 8.87 0.90 0.90 21.29 0.00 33.00
Floor 15.Queue 8.94 0.86 0.90 19.88 0.00 32.00
Floor 16.Queue 8.66 0.97 1.59 23.20 0.00 33.00
Floor 17.Queue 8.36 0.81 2.14 21.06 0.00 32.00
Floor 18.Queue 7.59 0.90 0.75 20.13 0.00 32.00
Floor 19.Queue 7.70 0.86 1.26 19.44 0.00 32.00
Floor 20.Queue 9.03 0.89 1.51 20.64 0.00 33.00
Floor 21.Queue 9.20 0.87 1.02 19.05 0.00 32.00
Floor 22.Queue 9.25 0.88 0.76 22.13 0.00 33.00
Floor 23.Queue 9.47 0.88 2.21 20.98 0.00 34.00
Floor 24.Queue 9.62 0.95 1.24 19.90 0.00 34.00
Floor 25.Queue 8.58 0.92 1.40 19.36 0.00 33.00
Floor 26.Queue 8.25 0.89 0.63 19.16 0.00 32.00
Floor 27.Queue 7.93 0.91 0.66 19.42 0.00 33.00
Floor 28.Queue 7.29 0.77 0.79 15.00 0.00 33.00
Floor 29.Queue 5.93 0.88 0.33 18.44 0.00 33.00
Floor 30.Queue 6.20 0.89 0.27 17.92 0.00 33.00
Floor 31.Queue 6.08 0.83 0.47 16.91 0.00 33.00
Floor 32.Queue 6.84 0.86 0.33 19.08 0.00 32.00
Floor 33.Queue 8.09 0.87 0.57 19.06 0.00 33.00
Floor 34.Queue 8.84 0.84 0.63 18.58 0.00 34.00
Floor 35.Queue 9.29 0.88 1.30 18.39 0.00 33.00
Floor 36.Queue 10.21 0.88 1.87 20.44 0.00 33.00
Floor 37.Queue 10.46 1.04 1.52 23.01 0.00 33.00
Floor 38.Queue 12.48 1.06 1.74 25.38 0.00 34.00
302
Table G12: Floor-Wise Lift Waiting Time (fire on 38th floor)
Waiting Time Average Half
Width
Minimum
Average
Maximum
Average
Minimum
Value
Maximum
Value
Floor 01.Queue 26.71 2.09 12.71 68.84 0.00 248.24
Floor 02.Queue 681.07 73.55 74.60 1751.90 0.80 1928.53
Floor 03.Queue 644.37 80.38 43.41 1927.50 0.40 2107.77
Floor 04.Queue 728.19 79.73 60.16 1727.69 0.83 2161.11
Floor 05.Queue 661.27 78.49 103.28 1558.48 0.07 1916.80
Floor 06.Queue 670.39 76.82 68.28 1948.20 0.79 2106.31
Floor 07.Queue 665.40 77.39 51.19 1717.77 0.54 1868.08
Floor 08.Queue 649.69 66.61 38.90 1808.16 0.56 1951.74
Floor 09.Queue 740.58 84.09 62.00 1789.25 0.51 1873.45
Floor 10.Queue 711.12 83.35 64.61 2006.97 0.12 2081.17
Floor 11.Queue 702.82 68.95 76.10 1584.51 1.82 1710.08
Floor 12.Queue 663.99 70.31 59.52 1819.13 3.40 1877.61
Floor 13.Queue 773.38 71.08 150.82 1703.08 0.78 1910.49
Floor 14.Queue 755.84 68.91 188.86 1704.80 3.11 1830.96
Floor 15.Queue 752.94 82.52 89.94 1966.89 0.08 2100.41
Floor 16.Queue 771.74 81.66 79.17 1790.04 0.23 1872.82
Floor 17.Queue 753.66 77.84 54.90 1730.19 1.79 1836.85
Floor 18.Queue 811.62 73.70 188.93 1850.73 3.43 1940.76
Floor 19.Queue 819.76 72.19 139.51 1971.81 3.02 2032.46
Floor 20.Queue 802.01 83.18 87.89 1922.62 1.55 1979.82
Floor 21.Queue 848.50 81.06 122.31 1988.96 0.02 2118.84
Floor 22.Queue 781.98 76.36 53.34 1850.90 0.53 2001.95
Floor 23.Queue 827.28 79.92 89.37 1872.78 1.62 2092.31
Floor 24.Queue 761.27 76.29 102.35 1647.48 2.21 2033.31
Floor 25.Queue 880.49 88.38 72.03 2051.23 0.19 2133.40
Floor 26.Queue 716.66 73.84 41.26 1528.29 0.03 1749.96
Floor 27.Queue 675.75 74.15 25.08 1579.95 0.12 1972.29
Floor 28.Queue 665.30 78.92 93.09 2121.02 0.08 2190.04
Floor 29.Queue 506.20 63.68 55.68 1688.51 0.14 2031.17
Floor 30.Queue 508.72 71.99 35.71 1375.16 0.08 1752.51
Floor 31.Queue 593.22 75.25 44.83 1486.70 0.02 1768.00
Floor 32.Queue 593.40 78.92 40.18 1657.46 0.15 2037.64
Floor 33.Queue 660.78 77.75 60.75 1706.38 0.01 1885.96
Floor 34.Queue 732.33 79.32 82.06 1757.74 0.01 1980.46
Floor 35.Queue 827.24 75.98 90.75 1794.73 0.04 2051.41
Floor 36.Queue 891.06 73.32 115.46 1688.14 0.11 2059.78
Floor 37.Queue 943.78 90.36 79.92 2001.48 0.00 2406.48
Floor 38.Queue 820.23 73.74 164.80 1613.44 0.79 2152.75
303
Table G13: Floor-Wise Number of Evacuees in Queue (fire on 38th floor)
Number Waiting Average Half
width
Minimum
Average
Maximum
Average
Minimum
Value
Maximum
Value
Floor 01.Queue 0.21 0.02 0.09 0.56 0.00 11.00
Floor 02.Queue 7.85 0.87 0.90 19.62 0.00 32.00
Floor 03.Queue 7.38 0.92 0.53 21.12 0.00 32.00
Floor 04.Queue 8.39 0.94 0.62 21.04 0.00 32.00
Floor 05.Queue 7.66 0.94 1.11 19.59 0.00 32.00
Floor 06.Queue 7.70 0.87 0.72 22.05 0.00 32.00
Floor 07.Queue 7.67 0.92 0.61 20.06 0.00 32.00
Floor 08.Queue 7.44 0.77 0.43 21.94 0.00 33.00
Floor 09.Queue 8.54 0.99 0.77 21.50 0.00 32.00
Floor 10.Queue 8.15 0.95 0.77 21.61 0.00 33.00
Floor 11.Queue 8.04 0.79 0.72 17.57 0.00 33.00
Floor 12.Queue 7.69 0.84 0.71 20.43 0.00 32.00
Floor 13.Queue 8.96 0.86 1.57 19.61 0.00 33.00
Floor 14.Queue 8.66 0.79 2.49 20.10 0.00 32.00
Floor 15.Queue 8.67 0.96 1.15 23.09 0.00 33.00
Floor 16.Queue 8.91 0.97 0.89 22.66 0.00 33.00
Floor 17.Queue 8.68 0.91 0.61 19.73 0.00 32.00
Floor 18.Queue 9.34 0.85 2.08 21.36 0.00 32.00
Floor 19.Queue 9.45 0.86 1.89 23.15 0.00 32.00
Floor 20.Queue 9.26 0.98 1.04 23.67 0.00 33.00
Floor 21.Queue 9.79 0.96 1.33 24.25 0.00 33.00
Floor 22.Queue 9.04 0.92 0.59 20.61 0.00 33.00
Floor 23.Queue 9.57 0.94 1.17 20.85 0.00 33.00
Floor 24.Queue 8.75 0.87 1.30 19.67 0.00 33.00
Floor 25.Queue 10.19 1.05 0.87 23.13 0.00 32.00
Floor 26.Queue 8.31 0.87 0.42 18.08 0.00 33.00
Floor 27.Queue 7.78 0.86 0.30 18.37 0.00 33.00
Floor 28.Queue 7.63 0.87 1.04 20.40 0.00 32.00
Floor 29.Queue 5.85 0.72 0.67 18.49 0.00 32.00
Floor 30.Queue 5.87 0.82 0.47 17.74 0.00 33.00
Floor 31.Queue 6.85 0.88 0.50 17.36 0.00 32.00
Floor 32.Queue 6.84 0.91 0.43 20.20 0.00 33.00
Floor 33.Queue 7.63 0.89 0.67 18.17 0.00 33.00
Floor 34.Queue 8.50 0.93 0.91 20.26 0.00 33.00
Floor 35.Queue 9.56 0.85 0.99 19.55 0.00 33.00
Floor 36.Queue 10.27 0.83 1.46 20.35 0.00 33.00
Floor 37.Queue 10.83 1.00 1.08 22.92 0.00 34.00
Floor 38.Queue 9.51 0.85 1.98 19.49 0.00 33.00
304
Appendix H
Verification of ARENA Model
Table H1: Lift Simulation Model – Lift Evacuation Time (1618.88 seconds)
Table H2: Listing of ELVAC Analysis of 38 Story Building
*************************************************************************** ELVAC VERSION 1.00 WRITTEN BY DANIEL M. ALVORD AND JOHN H. KLOTE CONTRIBUTION OF THE NATIONAL INSTITUTES OF STANDARDS AND TECHNOLOGY (U.S.) NOT SUBJECT TO COPYRIGHT FOR COMPILED VERSION ONLY - PORTIONS (C) COPYRIGHT MICROSOFT CORPORATION, 1988 ALL RIGHTS RESERVED. DOCUMENTATION: KLOTE,J.H., ALVORD,D.M., AND DEAL,S., ANALYSIS OF PEOPLE MOVEMENT DURING ELEVATOR EVACUATION, NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY, (U.S.), NISTIR 4730, 1992. **************************************************************************** Do you want to read about the model (Y or N)? N Enter the title of this run.
Example 38 Story Building
Enter 1 for SI units or 2 for English units: 1 Enter floors that the elevators serve (for example- B G 1-6) 1-38 Typical floor to floor height (m)? 3.0 Height exceptions: Floor range, height(m) (enter END to stop) END Discharge floor? 1 Time to outside after leaving elevator (s)? 0 Trip inefficiency factor (for example .1)? .1 Number of elevator cars in the group? 4 Normal operating velocity of car (m/s)? 3.15 Car acceleration (m/s**2)? 1.0 Elevator Full Load (people)? 16
305
------ MENU OF DOOR TYPES ------ ------------------------------------------------------------------------------------------------------------- Door type Door width Time to Open Door Transfer and Close Inefficiency
-------------------------------------------------------------------------------------------------------------- A Single-Slide 900mm (36in) 6.6 0.10 B Two-Speed 900mm (36in) 5.9 0.10 C Center-Opening 900mm (36in) 4.1 0.08 D Single-Slide 1100mm (42in) 7.0 0.07 E Two-Speed 1100mm (42in) 6.6 0.07 F Center-Opening 1100mm (42in) 4.6 0.05 G Two-Speed 1200mm (48in) 7.7 0.02 H Center-Opening 1200mm (48in) 5.3 0 I Two-Speed 1400mm (54in) 8.8 0.02 J Center-Opening 1400mm (54in) 6.0 0 K Two-Speed 1600mm (60in) 9.9 0.02 L Center-Opening 1600mm (60in) 6.5 0 M Two-Speed, 1600mm (60in) 6.0 0 Center-Opening N OTHER ***Pick one of the door choices A - M. If you wish to specify another type, enter N. H Other transfer inefficiency? 0.0 The start up time for automatically operated elevators is 41.25 seconds. Do you want to enter another value (Y or N)? N Typical Number of People per Floor? 32 People per floor exception: Floor Range, people (enter END to stop) END Percent of people on typical floor using elevator? 100 Percent usage exceptions: Floor Range, Percent (enter END to stop) END
306
Results:
Example 38 Story Building People per floor is 100. Distance between floors is 3.0 m or 10.0 ft. Elevator usage percent is 100.000% Normal car velocity is 3.00 m/s or 590.55 fpm. Car acceleration is 1.20 m/s2 or 3.94 ft/s2. Car full load is 16 people. Full load standing time is 40.26 s. Other transfer inefficiency is 0.0000 Trip inefficiency is 0.100 Door type: H Center-Opening 1200mm (48in) wide Doortime s 5.300 Door inefficiency 0.000
Figure H1 – ELVAC Lift Evacuation Time (1590.5 seconds)
307
Verification of Stair Model from Equation 2.8 (Chapter 2):
Listing of equation and variables used for verification
s
s
tWF
nNt +=1
where
n = 37 floors
N = 16 persons
Fs = 0.5 persons/meter/second
W =1.2 m
ts = 30 seconds
Then
t1 = 1116 seconds
308
Appendix J
Building Characteristics, HRR and Temperature
The enclosure characterisations for fire scenarios are given in Table J1. It is assumed
that the door of apartment is open while the occupants left the unit. The window
opening (0.9 m wide × 0.6 m high) from fire-affected units is considered in the
simulation. Wind pressure is exerted on window openings for a reasonably worst
fire scenario.
Table J1: Enclosure Characterisation of a Hypothetical Building (57m x 20m)
Enclosure Enclosure use Dimensions
L x b x h (m)
Openings
(m)
1 Kitchen/ Living/
Dining room/
Bedroom (fire-
affected SOU)
12m x 8.4m x 2.7m To # 2 – 2.0m x 0.9m – Door
(toward corridor)
2 Corridor 42m x 2.0m x 2.7m To # 1 – 2.0m x 0.9m – Door
To # 3 – 6m x 2.7m – Opening
via smoke lobby
To # 4 – 2.0m x 0.9m – Door
3 Lift lobby 40m x 2.0m x 2.7m To # 2 – 6m x 2.7m – Opening
via smoke lobby
4 Stair smoke lobby 3.0m x 2.0m x 2.7m To # 2 – 2.0m x 0.9m – Door
To # 5 – 2.0m x 0.9m – Door
5 Stair 2.0m x 2.0m x 2.7m To # 4 – 2.0m x 0.9m – Door
6 Opposite SOU 12m x 8.4m x 2.7m To # 2 – 2.0m x 0.9m – Door
(toward corridor)
The area of smoke lobby in double protected lift lobby is considered 1.5 times that of
stair smoke lobby as lifts are concentrated for full evacuation and stairs for half.
Heat Release Rate
The apartment unit of fire origin is approximately 96 m2 in the building. The
ceilings, floors and walls of the apartment unit are assumed to be composed of
concrete. The combustible fuel is based on the wood typically found in drawing
rooms of apartment buildings. The fire source was approximated as a rectangular
object (2.5 m × 2 m × 0.6 m). The fire growth is assumed based on a medium t-
squared curve fire Q= αt2 {where Q is the heat release rate (kW) and α is the fire
growth coefficient (0.01172 kW/s2)} to a constant peak value. The heat release rate
(HRR) is 1000 kW/m2. The combustion yields are given below:
309
SOOT_YIELD = 0.032
NU_O2 = 4.53
NU_CO2 = 4.12
NU_H2O = 3.21
MW_FUEL = 98
EPUMO2 = 11020
The graphical representations of HRR and temperature during Fire Scenario 1 are
given in Figures J1 and J2.
0
200
400
600
800
1000
1200
0 50 100 150 200 250 300 350
Time (second)
HRR (kW)
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 500 1000 1500 2000 2500 3000 3500 4000
Time (second)
HRR (kW)
Figure J1 – HRR during Fire Scenario 1 (HRR for 300 seconds and one hour)
310
0
100
200
300
400
500
600
700
800
900
0 500 1000 1500 2000 2500 3000 3500 4000
Time (second)
Temperature (C)
Figure J2 – Temperature during Fire Scenario 1
311
Appendix K
Occupants’ Movement Causing Door Opening and Closing
Table K1: Door Opening and Closing for Stair and Lift Lobbies and Lift Shaft by the
Entire Population
Protected lift lobby
door opening and
closing by 100 %
floor population
Stair door opening
and closing by 50%
floor population
(1st door)
Stair door opening
and closing by 50%
floor population
(2nd
door)
Lift landing door
opening and
closing
105 – 108
126 – 129
138 – 141
155 – 158
178 – 181
196 – 199
208 – 211
225 – 228
244 – 247
260 – 263
277 – 280
293 – 296
310 – 313
322 – 325
338 – 341
354 – 357
368 – 371
387 – 390
405 – 408
421 – 424
435 – 438
454 – 457
465 – 468
483 – 486
493 – 496
505 – 508
518 – 521
536 – 539
551 – 554
572 – 575
583 – 586
598 – 601
105 – 108
144 –147
183 – 186
215 – 218
235 – 238
266 – 269
296 – 299
336 – 339
375 – 378
418 – 421
452 – 455
486 – 489
499 – 502
520 – 523
545 – 548
574 – 577
620 – 623
108 – 111
147 – 150
186 – 189
218 – 219
238 – 241
269 – 272
299 – 302
339 – 342
378 – 381
421 – 424
455 – 458
489 – 492
502 – 503
523 – 526
548 – 551
577 – 580
623 – 626
For 2nd floor fire
303 – 313
1157 – 1167
1695 – 1705
For 19th floor fire
932 – 942
1634 – 1644
1920 – 1930
For 38th floor fire
390 – 400
1982 – 1992
2105 – 2115
Above table indicates that 100% of the population for lifts and 100% of the
population for two stairs (or 50% for one stair) are using lift lobby or stair lobby
312
doors. Lifts are also serving the floor and lift landing doors are opening and closing
in the lift lobby (see last column). The timings were determined from the stochastic
evacuation model and used in the FDS model for smoke leakages.
Table K2: Door Opening and Closing for Lift and Stair for Partial Population
Protected lift lobby
door opening and
closing by 25 %
floor population
Stair door opening and
closing by 37.5% floor
population
(1st door)
Stair door opening
and closing by 37.5%
floor population
(2nd
door)
105 – 108
178 – 181
244 – 247
310 – 313
368 – 371
435 – 438
493 – 496
551 – 554
105 – 108
144 –147
183 – 186
235 – 238
266 – 269
296 – 299
375 – 378
418 – 421
452 – 455
499 – 502
520 – 523
545 – 548
108 – 111
147 – 150
186 – 189
238 – 241
269 – 272
299 – 302
378 – 381
421 – 424
455 – 458
502 – 503
523 – 526
548 – 551
Above table indicates that lifts are used by 25% of the population while two stairs are
used by 75% of the population (or 37.5% for one stair).
313
Appendix L
Visibility Determination at a Focal Point
Generally, visibility is determined with the help of line of sight method. This
research determined the visibility at a focal point in the lift lobby, lift shaft and stair
shaft. For comparing visibility at a line of sight and at a focal point, an analysis is
given for lift lobby. Figure L1 illustrates the method for determining the visibility via
line of sight. Extinction coefficients were obtained from the averages of three
measurements along the lines of sight (LS).
Figure L1 – Visibility Determination at Three Lines of Sight and Lift Lobby
From the averages of extinction coefficients, it was determined that the time to
exceed tenability limit for visibility arrives at 198 seconds for line of sight 1 (LS1 –
stair1 exit), 230 seconds for LS2 (stair2 exit) and 200 seconds for LS3 (lift exit) (see
Figure L2). Time to exceed tenability limit for visibility arrives at 241 seconds in lift
lobby in Fire Scenario 1. The extinction coefficient is 0.5 m-1.
314
Visibility in corridor and lift lobby
0
5
10
15
20
25
30
35
40
45
0 500 1000 1500 2000 2500Time (second)
Extinction coefficient
(1/m
)
LS_1(TC_1)
LS_1(TC_2)
Occupant_TC
LS_3(TC_2)
LS_3(TC_1)
LS_2(TC_2)
LS_2(TC_1)
Figure L2- Recorded Visibility at Three Lines of Sight
The evacuees have the option of using either stair2 or lifts, after the stair1 becomes
invisible. Considering the walking speed, the occupants are required to travel an
additional 10 m for lifts and 30 m for stair2. Under the deteriorating situation, the
evacuees would be near the lift lobby at the earlier stage then to approach 20 m more
for stair2. Once the evacuees are inside the lift lobby, line of sight method can not be
applied due to small compartment.
315
Appendix M
Species Concentration and Fractional Effective Doses of Smoke, Gases and Heat
Concept Design B (Protected Lift Lobby): Figure M1 gives the results of smoke
extinction coefficient, concentration of gases, temperature and radiant heat flux in
Fire Scenarios 7 to 12. However, slight traces of smoke, gases or minor changes in
temperature and radiant heat flux in the protected lift shaft are not visible in a few
fire scenarios.
Fire Scenario 7 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 7 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
10
20
30
40
50
60
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1
1.5
2
2.5
3
3.5
4
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
316
Fire Scenario 7 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
60
70
80
90
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 7 (Smoke and Gases in Stair)
0
2
4
6
8
10
12
14
16
18
20
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
317
Fire Scenairo 8 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 8 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
10
20
30
40
50
60
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1
1.5
2
2.5
3
3.5
4
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
318
Fire Scenario 8 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
60
70
80
90
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 8 (Smoke and Gases in Stair)
0
2
4
6
8
10
12
14
16
18
20
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
319
Fire Scenario 9 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 9 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
10
20
30
40
50
60
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1
1.5
2
2.5
3
3.5
4
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
320
Fire Scenario 9 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
60
70
80
90
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 9 (Smoke and Gases in Stair)
0
2
4
6
8
10
12
14
16
18
20
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
321
Fire Scenario 10 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 10 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
10
20
30
40
50
60
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1
1.5
2
2.5
3
3.5
4
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
322
Fire Scenario 10 (Smoke and Heat in Lift Shaft)
0
10
20
30
40
50
60
70
80
90
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 10 (Smoke and Gases in Stair)
0
4
8
12
16
20
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
323
Fire Scenario 11 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 11 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
10
20
30
40
50
60
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1
1.5
2
2.5
3
3.5
4
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
324
Fire Scenario 11 (Smoke and Gases in Lift Shaft)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 11 (Smoke and Gases in Stair)
0
2
4
6
8
10
12
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
325
Fire Scenario 12 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440
Time (second)
Extinction coefficient
(1/m
) and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 12 (Temperature and Heat in Lift Lobby)
0
10
20
30
40
50
60
0 360 720 1080 1440
Time (second)
Temperature (C)
1
1.5
2
2.5
3
3.5
4
Radiant heat flux (kW/m
2)
Temperature Radiant heat flux
326
Fire Scenario 12 (Smoke and Gases in Lift Shaft)
0
10
20
30
40
50
60
70
80
90
100
0 360 720 1080 1440
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenairo 12 (Smoke and Gases in Stair)
0
2
4
6
8
10
12
14
16
18
20
0 360 720 1080 1440
Time (second)
Extinction coefficient (1/m
)
and CO (ten ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Figure M1 – Smoke, Gases and Heat in the Lift Lobby, the Lift Shaft and the Stair
Shaft (Fire Scenarios 7 to 12)
327
Concept Design B (Protected Lift Lobby with 25% Population): Figure M2 gives
the results of smoke, gases, temperature and radiant heat flux in Fire Scenarios 13 to
18. Concept Design B is considered for lift and stair partial evacuation. Slight traces
of smoke, gases or minor changes in temperature and radiant heat flux in protected
lift shaft are not visible in a few fire scenarios.
Fire Scenario 13 (Smoke and Gases in Lift Lobby)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 13 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1.5
1.6
1.7
1.8
1.9
2
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
328
Fire Scenario 13 (Smoke and Gases in Lift Shaft)
0
2
4
6
8
10
12
14
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 13 (Smoke and Gases in Stair)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
329
Fire Scenario 14 (Smoke and Gases in Lift Lobby)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 14 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1.5
1.6
1.7
1.8
1.9
2
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
330
Fire Scenario 14 (Smoke and Gases in Lift Shaft)
0
2
4
6
8
10
12
14
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
0
5
10
15
20
25
Radiant heat flux (kW/m
2)
Extinction coefficient CO CO2 O2
Fire Scenario 14 (Smoke and Gases in Stair)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
331
Fire Scenario 15 (Smoke and Gases in Lift Lobby)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 15 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1.5
1.6
1.7
1.8
1.9
2
Radiant heat flux
(kW/m2)
Temperature Radiant heat flux
332
Fire Scenario 15 (Smoke and Gases in Lift Shaft)
0
2
4
6
8
10
12
14
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 15 (Smoke and Gases in Stair)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
333
Fire Scenario 16 (Smoke and Gases in Lift Lobby)
0
10
20
30
40
50
60
70
80
90
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 16 (Temperature and Heat in Lift Lobby)
0
5
10
15
20
25
30
35
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1.5
1.6
1.7
1.8
1.9
2
Radiant heat flux
(kW/m
2)
Temperature Radiant heat flux
334
Fire Scenario 16 (Smoke and Gases in Lift Shaft)
0
2
4
6
8
10
12
14
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 16 (Smoke and Gases in Stair)
0
10
20
30
40
50
60
70
80
90
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
335
Fire Scenario 17 (Smoke and Gases in Lift Lobby)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 17 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1.5
1.6
1.7
1.8
1.9
2
Radiant heat flux
(kW/m
2)
Temperature Radiant heat flux
336
Fire Scenario 17 (Smoke and Gases in Lift Shaft)
0
2
4
6
8
10
12
14
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m)
and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2(%
)
Extinction coefficient CO CO2 O2
Fire Scenario 17 (Smoke and Gases in Stair)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
337
Fire Scenario 18 (Smoke and Gases in Lift Lobby)
0
20
40
60
80
100
0 360 720 1080 1440 1800 2160
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 18 (Temperature and Radiant Heat Flux in Lift
Lobby)
0
5
10
15
20
25
30
35
40
0 360 720 1080 1440 1800 2160
Time (second)
Temperature (C)
1.5
1.6
1.7
1.8
1.9
2
Radiant heat flux
(kW/m
2)
Temperature Radiant heat flux
338
Fire Scenario 18 (Smoke and Gases in Lift Shaft)
0
2
4
6
8
10
12
14
16
0 360 720 1080 1440 1800 2160
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 18 (Smoke and Gases in Stair)
0
10
20
30
40
50
60
70
80
90
100
0 360 720 1080 1440 1800 2160
Time (second)
Extinction coefficient (1/m
)
and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Figure M2 – Smoke, Gases and Heat in the Lift Lobby, the Lift Shaft and the Stair
Shaft (Fire Scenarios 13 to 18)
339
Concept Design C (Double Protected Lift Lobby): Figure M3 gives the results of
smoke, gases, temperature and radiant heat flux in Fire Scenarios 19 to 24. Slight
traces of smoke, gases or minor changes in temperature and radiant heat flux in
protected lift shaft are not visible in a few fire scenarios.
Fire Scenario 19 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient (1/m
)
and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 19 (Temperature and Radiant Heat Flux in Lift
Lobby)
19.7
19.8
19.9
20
20.1
20.2
20.3
20.4
20.5
20.6
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Temperature (C)
1.67
1.672
1.674
1.676
1.678
1.68
Radiant heat flux
(kW/m
2)
Temperature Radiant heat flux
(Temperature and radiant heat flux have not increased significantly in double protected lift shaft)
340
Fire Scenario 19 (Smoke and Gases in Lift Shaft)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
(Smoke and gases have not increased significantly in double protected lift shaft)
Fire Scenario 20 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
341
Fire Scenario 21 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 22 (Smoke and Gases in Lift Lobby)
0
10
20
30
40
50
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
342
Fire Scenario 23 (Smoke and Gases in Lift Lobby)
0
5
10
15
20
25
30
0 360 720 1080 1440 1800 2160 2520 2880 3240 3600
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Fire Scenario 24 (Smoke and Gases in Lift Lobby)
0
20
40
60
80
100
120
0 360 720 1080
Time (second)
Extinction coefficient
(1/m
) and CO (ppm)
0
5
10
15
20
25
CO2 (%) and O2 (%)
Extinction coefficient CO CO2 O2
Figure M3 – Smoke, Gases and Heat in the Lift Lobby and the Lift Shaft (Fire
Scenarios 19 to 24)
343
FED – Concept Design B (Protected Lift Lobby): Figure M4 gives the
representations for FED smoke, asphyxiant gases and heat in Fire Scenarios 7 to 12.
Traces of asphyxiant gases and temperature less than 60°C are not shown in a few
fire scenarios.
Fire Scenario 7 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 7 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
(Traces of CO are invisible)
344
Fire Scenario 7 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
Fire Scenario 8 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
345
Fire Scenario 8 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
(Traces of CO are invisible)
Fire Scenario 8 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
346
Fire Scenario 9 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 9 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED
asphyxiant
Smoke Asphyxiant
(Traces of CO are invisible)
347
Fire Scenario 9 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
Fire Scenario 10 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (second)
FED smoke and FED asphyxiant
Smoke Asphyxiant
348
Fire Scenario 10 (FED in Lift shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 10 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
349
Fire Scenario 11 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED
asphyxiant
Smoke Asphyxiant
Fire Scenario 11 (FED in Lift Shaft)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
350
Fire Scenario 11 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
(Fire on 38th level – Smoke accumulated in the stair top and has not diluted)
Fire Scenario 12 (FED in Lift Lobby)
0
1
2
3
4
5
0 5 10 15 20 25 30
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
351
Fire Scenario 12 (FED in Lift Shaft)
0
1
2
3
4
5
0 5 10 15 20 25 30
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 12 (FED in Stair)
0
1
2
3
4
5
0 5 10 15 20 25 30
Time (minute)
FED smoke
Smoke
Figure M4 – FED in Fire Scenarios 6 to 12
352
FED – Concept Design B (Protected Lift Lobby with Partial Evacuation): Figure
M5 gives the FED for smoke, gases and heat in Fire Scenarios 13 to 18. Concept
Design B is considered for lift and stair partial evacuation.
Fire Scenario 13 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 13 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
353
Fire Scenario 13 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
Fire Scenario 14 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED
asphyxiant
Smoke Asphyxiant
354
Fire Scenario 14 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
Fire Scenario 14 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
355
Fire Scenario 15 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 15 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
356
Fire Scenario 15 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
Fire Scenario 16 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
357
Fire Scenario 16 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
Fire Scenario 16 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
(Traces of CO are invisible)
358
Fire Scenario 17 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 17 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
359
Fire Scenario 17 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 18 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
360
Fire Scenario 18 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 18 (FED in Stair)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
Figure M5 – FED in Fire Scenarios 13 to 18
361
FED – Concept Design C (Double Protected Lift Lobby): Figure M6 gives the
FED for smoke, gases and heat in Fire Scenarios 19 to 24.
Fire Scenario 19 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED
asphyxiant
Smoke Asphyxiant
Fire Scenario 19 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
362
Fire Scenario 20 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 20 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
363
Fire Scenario 21 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 21 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
364
Fire Scenario 22 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED asphyxiant
Smoke Asphyxiant
Fire Scenario 22 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
365
Fire Scenario 23 (FED in Lift Lobby)
0
1
2
3
4
5
0 10 20 30 40
Time (minute)
FED smoke and FED
asphyxiant
Smoke Asphyxiant
Fire Scenario 23 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30 40
Time (minute)
FED smoke
Smoke
366
Fire Scenario 24 (FED in Lift Lobby)
0
1
2
3
4
5
0 5 10 15 20 25
Time (minute)
FED smoke and FED
asphyxiant
Smoke Asphyxiant
Fire Scenario 24 (FED in Lift Shaft)
0
0.1
0.2
0.3
0.4
0.5
0 5 10 15 20 25
Time (minute)
FED smoke
Smoke
Figure M6 – FED in Fire Scenarios 19 to 24
367
Appendix N
Calculation of Fractional Effective Doses of Smoke, Gases and Heat
The fractional effective doses (FEDs) are calculated for Fire Scenario 1. The FEDs
are related to smoke, asphyxiant toxic gases and radiant and convective heats. The
safe and incapacitation doses are shown in bold. The safe criterion is taken as one-
tenth of incapacitation.
TIME (minutes) 0 1 2 3 4 5 6 7 8 9 10
Extinction coefficient (1/m) 0 0 0 0 0.23 0.97 2.20 3.90 6.37 9.06 11.99
CO (ppm) 0 0 0 0 10.42 44.78 104.88 189.88 316.39 447.33 588.22 CO2 (%) 0 0 0 0 0.09 0.39 0.91 1.65 2.75 3.89 5.11 O2 (%) 20.72 20.72 20.72 20.72 20.60 20.22 19.55 18.61 17.20 15.74 14.18
Temperature (oC) 20.00 20.00 20.00 20.00 24.28 33.61 42.39 48.97 55.23 53.54 51.43
Heat Flux (kW/m2)
Lift lo
bby
1.67 1.67 1.67 1.68 1.68 1.75 1.89 2.08 2.29 2.34 2.35
Extinction coefficient (1/m) 0 0 0 0 0.08 0.28 0.52 1.15 2.00 2.17 2.34
CO (ppm) 0 0 0 0 3.61 12.52 23.36 52.43 91.61 98.85 105.71 CO2 (%) 0 0 0 0 0.03 0.11 0.20 0.46 0.80 0.86 0.92 O2 (%) 20.72 20.72 20.72 20.72 20.68 20.58 20.46 20.13 19.69 19.61 19.54
Temperature (oC) 20.00 20.00 20.00 20.00 21.43 23.91 25.11 28.25 30.31 27.73 25.93
Heat Flux (kW/m2)
Lift sh
aft
1.67 1.67 1.67 1.67 1.68 1.68 1.68 1.69 1.71 1.71 1.72
FED smoke 0 0 0 0 0.46 1.93 4.40 7.80 12.75 18.11 23.97
FID CO 0 0 0 0 0 0 0 0.01 0.01 0.02 0.02
FID CO2 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.73 2.18 2.78
FID O2 0 0 0 0 0 0 0 0 0 0 0
FED asphyxiant 0 0 0 0 0 0 0 0.01 0.02 0.03 0.06
ΣΣΣΣFED asphyxiant 0 0 0 0 0 0 0.01 0.01 0.03 0.06 0.12
FED heat 0 0 0 0 0 0 0 0 0 0 0
ΣΣΣΣFED heat
Lift lo
bby
0 0 0 0 0 0 0 0 0 0 0
FED smoke 0 0 0 0 0.16 0.56 1.04 2.30 3.99 4.35 4.67
FID CO 0 0 0 0 0 0 0 0 0 0 0
FID CO2 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
FID O2 0 0 0 0 0 0 0 0 0 0 0
FED asphyxiant 0 0 0 0 0 0 0 0 0 0 0
ΣΣΣΣFED asphyxiant 0 0 0 0 0 0 0 0 0.01 0.01 0.01
FED heat 0 0 0 0 0 0 0 0 0 0 0
ΣΣΣΣFED heat
Lift sh
aft
0 0 0 0 0 0 0 0 0 0 0
369
TIME (minutes) 11 12 13 14 15 16 17 18 19 20
Extinction coefficient (1/m) 13.72 15.21 15.17 16.40 20.28 23.82 27.43 33.28 36.49 39.08
CO (ppm) 686.43 770.39 758.96 939.33 1151.45 1292.32 1387.30 1590.02 1707.85 1814.67 CO2 (%) 5.96 6.69 6.59 8.16 10.00 11.23 12.05 13.81 14.84 15.76 O2 (%) 13.08 12.15 12.28 10.27 7.91 6.34 5.29 3.03 1.72 0.53
Temperature (oC) 57.88 62.00 57.87 105.83 102.45 85.81 61.41 42.86 36.47 33.98
Heat Flux (kW/m2)
Lift lo
bby
2.54 2.67 2.59 4.30 4.55 3.77 2.80 2.32 2.16 2.12
Extinction coefficient (1/m) 1.81 3.37 4.18 4.76 6.45 5.81 4.87 5.26 9.51 8.22
CO (ppm) 81.19 153.47 190.60 222.04 314.39 276.91 224.50 251.25 545.83 494.88 CO2 (%) 0.71 1.33 1.66 1.93 2.73 2.41 1.95 2.18 4.74 4.30 O2 (%) 19.81 19.00 18.59 18.23 17.20 17.62 18.21 17.91 14.61 15.18
Temperature (oC) 24.12 27.74 28.32 35.25 49.30 41.13 31.35 42.18 107.90 125.83
Heat Flux (kW/m2)
Lift sh
aft
1.72 1.73 1.76 1.82 2.00 1.87 1.92 2.30 3.55 7.94
FED smoke 27.43 30.41 30.33 32.79 40.56 47.63 54.86 66.56 72.97 78.16
FID CO 0.02 0.03 0.03 0.03 0.04 0.05 0.05 0.06 0.06 0.07
FID CO2 3.30 3.81 3.74 5.11 7.39 9.44 11.14 15.84 19.44 23.40
FID O2 0.02 0.03 0.03 0.09 0.33 0.76 1.35 4.57 9.28 17.62
FED asphyxiant 0.10 0.14 0.13 0.26 0.63 1.20 1.91 5.48 10.48 19.16
ΣΣΣΣFED asphyxiant 0.22 0.36 0.49 0.75 1.38 2.58 4.49 9.96 20.44 39.60
FED heat 0.04 0.07 0.06 0.24 0.23 0.15 0.07 0.05 0.04 0.04
ΣΣΣΣFED heat
Lift lo
bby
0.04 0.11 0.18 0.42 0.65 0.80 0.87 0.92 0.95 0.99
FED smoke 3.61 6.75 8.37 9.51 12.90 11.62 9.74 10.52 19.03 16.44
FID CO 0 0 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.01
FID CO2 1.00 1.00 1.00 1.00 1.73 1.62 1.48 1.55 2.58 2.36
FID O2 0 0 0 0 0 0 0 0 0 0
FED asphyxiant 0 0 0.01 0.01 0.02 0.01 0.01 0.01 0.04 0.03
ΣΣΣΣFED asphyxiant 0.01 0.02 0.02 0.03 0.04 0.06 0.07 0.08 0.11 0.15
FED heat 0 0 0 0 0 0 0 0 0.23 0.47
ΣΣΣΣFED heat
Lift sh
aft
0 0 0 0 0 0 0 0 0.23 0.70
370
TIME (minutes) 21 22 23 24 25 26 27 28 29 30
Extinction coefficient (1/m) 39.92 39.79 39.64 39.41 39.17 38.84 38.81 39.03 39.05 38.34
CO (ppm) 1858.65 1853.66 1847.71 1840.73 1833.29 1826.36 1822.89 1817.62 1810.51 1798.25 CO2 (%) 16.15 16.10 16.05 15.99 15.93 15.87 15.84 15.79 15.73 15.62 O2 (%) 0 0 0 0 0 0 0 0 0 0
Temperature (oC) 34.80 34.96 35.17 35.80 36.43 37.90 37.51 34.90 33.52 37.09
Heat Flux (kW/m2)
Lift lo
bby
2.17 2.17 2.21 2.20 2.19 2.14 2.13 2.02 1.99 2.08
Extinction coefficient (1/m) 10.29 11.35 10.88 13.11 15.79 18.46 21.03 28.06 35.13 39.60
CO (ppm) 698.44 857.93 813.19 861.59 1052.40 1076.88 1099.92 1515.43 1685.19 1801.34 CO2 (%) 6.07 7.45 7.06 7.48 9.14 9.36 9.56 13.17 14.64 15.65 O2 (%) 12.90 11.11 11.61 11.07 8.93 8.66 8.40 3.75 1.85 0.00
Temperature (oC) 177.10 228.44 227.52 162.43 170.44 114.50 74.29 86.21 46.11 34.17
Heat Flux (kW/m2)
Lift sh
aft
9.68 18.87 17.93 12.82 10.26 5.39 4.05 3.75 2.51 2.01
FED smoke 79.85 79.59 79.28 78.82 78.34 77.68 77.63 78.06 78.10 76.68
FID CO 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07
FID CO2 25.26 25.04 24.79 24.49 24.17 23.88 23.74 23.52 23.24 22.75
FID O2 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48
FED asphyxiant 25.18 25.16 25.14 25.11 25.08 25.06 25.04 25.03 25.00 24.96
ΣΣΣΣFED asphyxiant 64.78 89.94 115.08 140.19 165.27 190.33 215.37 240.39 265.39 290.35
FED heat 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.03 0.04
ΣΣΣΣFED heat
Lift lo
bby
1.03 1.07 1.11 1.15 1.19 1.23 1.26 1.30 1.33 1.37
FED smoke 20.57 22.70 21.77 26.21 31.58 36.92 42.06 56.12 70.26 79.20
FID CO 0.02 0.02 0.02 0.02 0.03 0.03 0.03 0.04 0.05 0.05
FID CO2 3.37 4.44 4.11 4.47 6.22 6.49 6.76 13.92 18.69 22.87
FID O2 0.02 0.06 0.04 0.06 0.19 0.22 0.25 3.10 8.65 23.48
FED asphyxiant 0.09 0.16 0.14 0.17 0.37 0.41 0.46 3.69 9.52 24.62
ΣΣΣΣFED asphyxiant 0.23 0.40 0.53 0.70 1.07 1.48 1.94 5.62 15.14 39.76
FED heat 1.14 2.72 2.65 1.03 1.05 0.32 0.13 0.15 0.05 0.03
ΣΣΣΣFED heat
Lift sh
aft
1.84 4.56 7.20 8.23 9.28 9.60 9.72 9.87 9.92 9.96
371
TIME (minutes) 31 32 33 34 35 36 37 38 39 40
Extinction coefficient (1/m) 38.31 38.18 38.05 37.86 37.71 37.75 37.69 38.66 39.66 40.25
CO (ppm) 1797.17 1794.63 1791.25 1788.54 1785.84 1780.32 1774.39 1789.59 1801.42 1806.53 CO2 (%) 15.61 15.59 15.56 15.54 15.51 15.47 15.41 15.55 15.65 15.69 O2 (%) 0 0 0 0 0 0 0 0 0 0
Temperature (oC) 37.19 37.74 38.21 39.33 40.07 38.83 38.29 33.11 27.36 23.75
Heat Flux (kW/m2)
Lift lo
bby
2.08 2.09 2.11 2.14 2.16 2.12 2.10 2.00 1.84 1.75
Extinction coefficient (1/m) 38.67 38.55 37.23 38.16 37.64 37.59 38.64 39.79 40.32 40.69
CO (ppm) 1775.32 1764.11 1757.81 1762.48 1750.65 1748.82 1741.10 1757.44 1771.22 1773.84 CO2 (%) 15.42 15.33 15.27 15.31 15.21 15.19 15.13 15.27 15.39 15.41 O2 (%) 0 0 0 0 0 0 0 0 0 0
Temperature (oC) 39.43 39.47 49.99 42.51 45.88 46.28 36.80 29.20 26.28 23.76
Heat Flux (kW/m2)
Lift sh
aft
2.11 2.17 2.37 2.29 2.41 2.48 2.12 1.88 1.82 1.76
FED smoke 76.61 76.37 76.11 75.72 75.43 75.49 75.37 77.32 79.32 80.51
FID CO 0.07 0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.07 0.07
FID CO2 22.70 22.60 22.47 22.37 22.26 22.05 21.82 22.41 22.87 23.08
FID O2 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48
FED asphyxiant 24.95 24.94 24.93 24.92 24.92 24.90 24.88 24.93 24.97 24.99
ΣΣΣΣFED asphyxiant 315.31 340.25 365.18 390.11 415.02 439.92 464.80 489.73 514.69 539.68
FED heat 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.03 0.03 0.03
ΣΣΣΣFED heat
Lift lo
bby
1.41 1.45 1.49 1.53 1.57 1.60 1.64 1.68 1.71 1.73
FED smoke 77.34 77.10 74.47 76.32 75.28 75.17 77.28 79.57 80.64 81.38
FID CO 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05
FID CO2 21.86 21.44 21.20 21.37 20.94 20.87 20.60 21.19 21.70 21.80
FID O2 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48 23.48
FED asphyxiant 24.55 24.52 24.51 24.52 24.49 24.49 24.47 24.51 24.54 24.55
ΣΣΣΣFED asphyxiant 64.31 88.83 113.34 137.86 162.35 186.83 211.30 235.80 260.34 284.89
FED heat 0.04 0.04 0.05 0.04 0.05 0.05 0.04 0.03 0.03 0.03
ΣΣΣΣFED heat
Lift sh
aft
10.00 10.04 10.09 10.13 10.18 10.23 10.27 10.30 10.33 10.36