|Author||David Charters and Ganapathy Ramachandran|
Quantitative Risk Assessment
in Fire Safety
Quantitative Risk Assessment
in Fire Safety
and David Charters
First published 2011
by Spon Press
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
Simultaneously published in the USA and Canada by Spon Press
270 Madison Avenue, New York, NY 10016, USA
Spon Press is an imprint of the Taylor & Francis Group, an informa business
© 2011 Ganapathy Ramachandran and David Charters
The right of Ganapathy Ramachandran and David Charters to be
identified as authors of this work has been asserted by them in accordance
with sections 77 and 78 of the Copyright, Designs and Patents Act 1988.
Typeset in Goudy by
HWA Text and Data Management, London
Printed and bound in Great Britain by
CPI Anthony Rowe, Chippenham, Wiltshire
All rights reserved. No part of this book may be reprinted or reproduced or
utilised in any form or by any electronic, mechanical, or other means, now
known or hereafter invented, including photocopying and recording, or in
any information storage or retrieval system, without permission in writing
from the publishers.
This publication presents material of a broad scope and applicability.
Despite stringent efforts by all concerned in the publishing process, some
typographical or editorial errors may occur, and readers are encouraged
to bring these to our attention where they represent errors of substance.
The publisher and author disclaim any liability, in whole or in part, arising
from information contained in this publication. The reader is urged to
consult with an appropriate licensed professional prior to taking any
action or making any interpretation that is within the realm of a licensed
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Ramachandran, G. (Ganapathy)
Quantitative risk assessment in fire safety / Ganapathy Ramachandran &
David A. Charters.
Includes bibliographical references and index.
1. Fire risk assessment. 2. Quantitative research. I. Charters,
David A. (David Anderson), 1949– II. Title.
ISBN13: 978-0-419-20790-0 (hbk)
ISBN13: 978-0-203-93769-3 (ebk)
List of figures
List of tables
About the authors
2 Qualitative and semi-quantitative risk assessment techniques
3 Quantitative risk assessment techniques
4 Acceptance criteria
6 Design fire size
7 Fire spread beyond room of origin
8 Performance and reliability of detection, alarm and suppression 268
9 Performance and reliability of human response and evacuation
10 Performance and effectiveness of fire service intervention
11 Whole project analysis
13 Combining data from various sources – Bayesian techniques
Breakdown of different fire risk assessment methods
Schematic of full quantitative fire risk assessment process
An F–n curve
An example of combustible materials
An example of a car fire
Summary report for an unstructured fire risk assessment
Record sheet from a structured qualitative fire risk assessment
An example of a matrix method flow chart
Venn diagram of union set
Venn diagram of intersection set
An example of a fault tree
General form of an event tree
Event tree of the early stages of fire event development
Simplified event tree for bus garage fires
General form of a fault tree
Fault tree for failure to detect a fire within five minutes of ignition
Average time (min) and area damaged (m2)
Damage and building size – textile industry, UK
Fire growth within room of origin
Discovery time and fatality rate
Density function curve of fire loss
Textile industry, United Kingdom – probability distribution of area
Pareto distribution of area damage – retail premises
The survivor probability distribution of fire loss for each class
in the textile industry
Probability tree for textile industry
Room layout and corresponding graph
Probabilistic network of fire spread of Room 1 to C2
Equivalent fire spread network with five-minute unrated doors
Typical consequence probability distribution
List of figures vii
3.23 Correlation of the number of non-fatal injuries per fire and the
area damaged by fire, heat and smoke
3.24 Ticket hall after the fire at King’s Cross underground station
3.25 Fire testing of a shutter
3.26 Example of a full-scale fire experiment
3.27 Typical schematic of building fire control volume model
3.28 Schematic of mass flow between control volumes
3.29 Schematics of a plume in natural ventilation
3.30 Typical schematic of building fire heat transfer processes
3.31 Typical results for a three-layer control volume model compared
with experimental data
3.32 Example of computational fluid dynamics
3.33 Example of discrete evacuation analysis
4.1 Life safety acceptance criteria shown on an F–n Curve.
4.2 Experience per 106 person years of fires with N or more fatalities
4.3 Targets for acceptability for societal risk from process industries
4.4 Multiple fatality fires in buildings
4.5 Proposed target for acceptibility from multiple fatality fires in all
buildings other than dwellings for different populations
5.1 Probability tree for rooms and offices
5.2 Sprinkler alternative probability tree
7.1 Effect of beam deflection on a fire-resisting wall
9.1 Occupants’ response – room of fire origin
9.2 Occupants’ response – room other than room of fire origin
10.1 Effectiveness of early intervention by fire brigade – example
based on UK fire statistics, non-sprinklered buildings
10.2 Effectiveness of early intervention by fire brigade – example
based on UK fire statistics, sprinklered buildings
11.1 Typical quantified fire risk assessment process
11.2 Typical output from SAFiRE
12.1 Damage and building size
12.2 Damage and compartment size
Typical matrix method record sheet
Hazard /accident frequency
Accident trigger probability
Accident frequency categories
Accident severity (life safety)
Cost of asset damage or destroyed
Cost of performance penalties or operational loss
Life safety risk rating
Business risk rating
Overall station risk rating
Example of part of a risk ranking spreadsheet (most likely fire event) 57
Relative S values for active systems
Values of components
Sample of a summary sheet
Relationship between random processes and probability modelling 72
Truth table for an OR gate
Truth table for an AND gate
Frequency of ignition per year
Frequency of ignition normalised per building
Frequency of ignition normalised by floor area
General data on the reliability of fire protection systems
Data used to quantify the probability of failure to detect a fire
within five minutes of ignition
3.9 Annual chance of a fire outbreak for various occupancies
3.10 Probable damage in a fire – parameters of Equation (3.41)
3.11 Industrial buildings – annual damage
3.12 Retail premises – assembly areas – frequency distribution of area
List of tables ix
3.13 Average loss per fire at 1966 prices
3.14 Pathways for fire spread equivalent network assuming
five-minute unrated corridor doors
3.15 Pathways for the fire spread equivalent network assuming
self–closing 20-minute rated corridor doors
3.16 Passive fire protection measures in hospitals
3.17 Probabilities of success and failure (standard normal distribution)
4.1 Typical types of acceptance criteria
4.2 Number of deaths per building year and per occupant year
4.3 Societal risk criteria for structural collapse due to fire
4.4 Discounted cash flow for bus garage sprinkler system
5.1 Spinning and doubling industry – places of origin of fires and
sources of ignition
5.2 Fire growth parameters
5.3 Expected values of the fire growth rate θ for different scenarios
– paper, printing and publishing industries, UK
5.4 Maximum values of the fire growth rate θ for different scenarios
– paper, printing and publishing industries, UK
5.5 Growth rate and doubling time
5.6 Probability of flashover
5.7 Fire extent of spread – textile industry, UK
5.8 Location of casualties – single and multiple occupancy dwellings
5.9 Fatal casualties in dwellings by whereabouts of casualties and
cause of death
6.1 Horizontal fire-damaged area in retail premises, 1978
6.2 Area of direct burning and number of sprinkler heads opening for
different fractiles, sprinklered retail premises
6.3 Pareto parameters
6.4 Design fire size
6.5 Pareto distribution of area damage – rooms without and with
6.6 Design fire size
7.1 Probability of fire spreading beyond room of origin
7.2 Material ignited first – probability of fire spreading beyond room
of origin (chemical and allied industries)
8.1 Experiment 1: Summary of results and overall rankings
8.2 Experiment 2: Summary of results and overall rankings
8.3 Effectiveness of sprinklers – non-industrial buildings
9.1 Design evacuation time and fatality rate for single and multiple
10.1 Direct loss in relation to attendance time – non-dwellings, UK
10.2 Percentage of small fires and fires extinguished by sprinklers
and by fire brigade
10.3 Control time and arrival time: regression coefficients
11.1 Risk–cost benefit analysis
About the authors
Ganapathy Ramachandran, PhD, DSc, FIFireE, MSFPE, held senior scientific
appointments for 23 years (1965–1988) at the Fire Research Station,
Borehamwood, Herts, UK. Since retiring in November 1988, he has been
practising as a private consultant mainly in research problems in the application
of statistical, probabilistic and economic techniques to fire risk evaluation, fire
protection engineering and fire insurance. He has published several papers on
these topics including chapters in the four editions of the SFPE Handbook of Fire
Protection Engineering. He is the author of Economics of Fire Protection (Spon 1998)
and co-author of Evaluation of Fire Safety (Wiley 2004).
Ramachandran was a Visiting Professor in Glasgow Caledonian University,
University of Hertfordshire and University of Manchester. For the past 12 years
he is a Visiting Professor in the University of Leeds. For three and a half years (up
until July 2010), he was Technical Director, Fire and Risk Engineering at Faber
David Charters, BSc, PhD, CEng, FIFireE, MIMechE, MSFPE is Director of
Fire Engineering at BRE Global (incorporating the Fire Research Station, FRS)
where he has been responsible for projects such as research into the use of lifts
and escalators for emergency evacuation and quantitative fire risk assessments
for nuclear and non-nuclear fire engineering projects. He has over 20 years of
fire safety experience. Prior to joining BRE Global, he was a Director of Arup
Fire responsible for the fire engineering of projects including; Channel Tunnel
Rail Link (CTRL) tunnels, City of Manchester Stadium and Attatürk Airport.
Previously he was Chief Engineer and Manager; Fire and Safety at NHS Estates,
Department of Health where he was responsible for policy advice to ministers
and the FIRECODE suite of guidance. Prior to this he was a Senior Fire Risk
Assessment Engineer with the UK Atomic Energy Authority.
In 2005, David became International President of the Institution of Fire
Engineers. He was appointed Visiting Professor in Fire Risk Analysis at the
University of Ulster in 2003. In 1994, David was awarded a PhD in Fire Growth
and Smoke Movement in Tunnels from the University of Leeds. He is co-author of
the chapter on Building Fire Risk Analysis in the SFPE Handbook on Fire Protection
Engineering. David was also responsible for quantitative fire risk assessments of
London Underground stations following the King’s Cross fire and Tartan Alpha
following the Piper Alpha disaster.
As viable alternative for prescriptive rules, particularly for large and complex
buildings, performance-based fire safety codes and building design methods
are being developed and applied in many countries. These rules, codes and
methods are mainly based on a qualitative assessment of fire risk supported by
experimental data, case studies, deterministic (scientific) models and professional
engineering practice. Statistical data provided by real fires are rarely analysed and
the results (evidence) produced by such an analysis are seldom included in the
risk assessment. Quantitative methods of risk assessment discussed in this book,
on the other hand, explicitly consider statistical data on real fires, in addition to
experimental data, and take account of uncertainties governing the occurrence of
a fire, spread of fire, damage caused in a fire, reliability of passive and active fire
protection systems and evacuation of building occupants.
Fire safety regulations, codes and standards do provide some unquantified
levels of safety particularly for the occupants of a building, but these levels may or
may not be adequate for some large, tall and complex buildings. Also, these levels
may not be acceptable to property owners who have to consider also property
damage and consequential losses such as business interruption and loss of profits.
Given that there are no quantitative criteria for fire risk in buildings, it is not
clear whether the safety levels provided by prescriptive rules are acceptable to
the society at large. Criteria for determining acceptable safety levels for property
owners and the society are discussed in Chapter 4.
Buildings can be designed according to acceptable levels for life safety and
property protection by applying quantitative methods of fire risk assessment.
These methods are discussed in detail in Chapter 3. Methods applicable to
some particular problems in fire safety engineering are discussed in other
chapters. These problems include initiation of fires (Chapter 5), design fire size
(Chapter 6), flashover and spread of a fire beyond room of fire origin (Chapter 7)
and performance and reliability of detection, alarm and suppression systems
(Chapter 8). Building designs should sufficiently consider response and evacuation
capability of occupants as discussed in Chapter 9.
Prescriptive rules specified in fire regulations, codes and standards should be
further verified, validated and improved, if necessary, in the light of quantitative
risk assessment. Otherwise, fire safety engineers will not be able to recommend
enhanced fire protection necessary for large tall and complex buildings. This is
because owners of such buildings are generally reluctant to spend more money
than the cost required for complying with prescriptive rules and performance-based
codes. The owners may not appreciate at present that enhanced fire protection may
be more cost-effective than protection provided by prescriptive rules and codes.
Enhanced fire protection for large, tall and complex buildings, can be
economically attractive if, in addition to cost of fire protection, monetary values
of the cost per life saved property/business loss are also considered in a cost–
benefit analysis. From among alternative fire protection strategies identified by a
performance-based code, a building owner may select a strategy which is the most
cost-effective in terms of costs and benefits of fire protection measures considered
and interactions and synergies between these measures (Chapter 12). The owner
would also consider the costs and benefits due to insurance/self-insurance options.
A detailed framework for carrying out a cost-benefit analysis as described above
has been discussed in Ramachandran’s book The Economics of Fire Protection
(1998). Quantitative risk assessment is an integral part of this framework.
As discussed above, the four major stakeholders or decision makers in the fire
safety field are property owners, architects and designers, consultant firms engaged
in fire safety engineering and government departments and organisations involved
in the development and enforcement of fire safety regulations, codes and standards.
Public fire and rescue services constitute the fifth major stakeholder. These bodies
have to provide adequate fire cover to properties in their geographical areas by
providing a sufficient number of strategically located fire stations, with enough
firefighters and other resources. This problem should necessarily consider the fire
risk in the properties in each area and the effectiveness of fire-protection measures
and fire brigade performance in reducing the risk – see Chapter 10. Taking into
account the interactions and synergies (Chapter 12) with fire protection measures,
a fire and rescue service can identify an economically optimum fire cover strategy
for any area.
The fire insurance industry is the fifth major stakeholder that has to estimate
appropriate premiums that can be charged for different types of properties. If
an insurance firm underestimates the premiums to be collected, it might face
bankruptcy in a market-driven economic environment involving keen competition
with other insurance firms. In the national interest the firm should promote fire
safety by offering sufficient rebates in premiums for fire protection systems and
self-insurance deductibles. Traditionally, most of the fire insurance underwriters
adopt semi-quantitative points schemes discussed in Chapter 2, Section 2.2.2.
The underwriters should validate and improve their premium calculation
methods by applying statistical models described in Chapter 3, section 3.3 and in
Chapter 11 of Ramachandran’s book The Economics of Fire Protection (1998). The
statistical models would provide more accurate estimates of the ‘risk premium’ and
the ‘safety loading’ to be added to this premium. An insurer firm can add another
loading towards expenses and profits to estimate the total premium to be charged.
The authors hope this book will provide most of the methods and tools for
quantitative risk assessment needed by stakeholders in the fire safety field.
Ganapathy Ramachandran (‘Ram’) would like to thank Liz Tattersall for word
processing his Chapters 5, 6, 7, 8, 9, 10 and 12, and sections in Chapters 2, 3,
4 and 13. Liz coped admirably well with all the statistical and mathematical
functions and formulae in these chapters and sections. Ram’s wife, Radha, gave
him considerable assistance in checking the typescript of the manuscript and
printed page proofs for spotting mistakes to be corrected.
David Charters would like to thank Dr Roth Phylaktou for the permission
to incorporate content from the MSc Module on Fire Risk Assessment and
Management at the University of Leeds. He would also like to thank Dominic
Vallely and James Holland (formerly and currently respectively) of Network
Rail, Paul Scott, Fermi Ltd and Matthew Salisbury, MSA for their support in the
development of the book. David would like to thank BRE Global and all the other
people who gave permission for material in this book. He would also like to thank
his sons, Jack and Theo for their encouragement and support.
The authors acknowledge the following permissions to reproduce material in this
Figures 1.2 and 11.1 are reproduced from D. A. Charters, ‘A review of fire
risk assessment methods’, Interflam 04, reproduced with permission of Interscience
Figures 1.3, 4.1, 2.5 and Tables 2.1, 2.2, 2.3, 2.4, 2.5, 2.6 and 2.7 are reproduced
with permission of Agetro Ltd
Figures 2.1, 2.2, 3.24, 3.25, 3.32 and 3.33 are reproduced with permission of
BRE Global Ltd.
Table A.2.1. is reproduced from E. C. Wessels (1980) ‘Rating techniques for
risk assessment’, Fire International, 67, 80–89. with permission from Keyways
Tables A.2.2. and A.2.3. are reproduced from P. Stollard (1984), ‘The
development of a points scheme to assess fire safety in hospitals’, Fire Safety
Journal, 7, 2, 145–153 with permission of Elsevier.
Figures 3.1, 3.2 and 3.3 are reproduced from D. A Charters, Fire Risk
Assessment and Management MSc module reproduced with permission of the
Energy and Resources Research Institute, University of Leeds.
Figures 3.5, 3.6, 3.8, 3.9, Tables 3.7 and 4.2 are reproduced from D. A Charters
and G. Ramachandran, committee draft of BSI PD 7974 ‘Application of fire
safety engineering principles to the design of buildings’, Part 7 ‘Probabilistic risk
assessment’. Permission to reproduce extracts from PD 7974 ‘Application of fire
safety engineering principles to the design of buildings’ is granted by BSI. British
Standards can be obtained in PDF or hard copy formats from the BSI online shop:
www.bsigroup.com/Shop or by contacting BSI Customer Services for hardcopies
only: Tel: +44 (0)20 8996 9001, Email: [email protected]
Figure 3.7 and Table 4.4 are reproduced from D. A. Charters ‘Fire safety
assessment in bus transportation’, Fire Safety in Transport, reproduced with
permission of the Institution of Mechanical Engineers.
Figures 3.10 to 3.19, 5.2, 12.1 and 12.2 are repoduced from D.J .Rasbash, G.
Ramachandran, B. Kandola, J. M. Watts and M. Law (2004) Evaluation of Fire
Safety with permission of John Wiley and Sons Ltd (Figures 3.10, 3.11, 12.1 and
12.2 were previously published in G. Ramachandran (1998) The Economics of Fire
Proctection published by Spon)
Figures 3.20 and 3.21 and Tables 3.14 and 3.15 are reproduced from W. T. C
Ling and R. B. Williamson (1985) ‘The modeling of fire spread through probabilistic
networks’, Fire Safety Journal, 9, 287–300 with permission of Elsevier, and from
D.J. Rasbash, G. Ramachandran, B. Kandola, J. M. Watts and M. Law (2004)
Evaluation of Fire Safety with permission of John Wiley and Sons Ltd.
Figure 3.28, 3.29, 3.30 and 3.31 are reproduced from D.A. Charters, ‘Control
volume modelling of tunnel fires’ in A. Beard and R. Carvel (eds) The Handbook
of Tunnel Fire Safety, reproduced with permission of Thomas Telford Publishing.
Figures 4.2, 4.3, 4.4 and 4.5 are reproduced from D.J. Rasbash (1984), ‘Criteria
for acceptability for use with quantitative approaches to fire safety’, Fire Safety
Journal, 8, 141–158 with permission of Elsevier.
Figure 7.1 is reproduced from BS5950, Part 8, 1990, Section 4 with permission
of British Standards Institution.
Table 11.1 is reproduced from D. A. Charters, M. Salisbury and S. Wu, ‘The
development of risk-informed performance based codes’, 5th International
Conference on Performance Based Codes and Design Methods, reproduced with
permission of The Society of Fire Protection Engineers.
Figure 11.2 is reproduced from D. A. Charters, ‘Quantitative fire risk assessment
in the design of major multi-occupancy buildings’, Interflam 01, reproduced with
permission of Interscience Communications.
Every effort has been made to seek permission to reproduce copyright material
before the book went to press. If any proper acknowledgement has not been made,
we would invite copyright holders to inform us of the oversight.
We all take risks all the time, whether it is crossing the road, driving to work or
watching television. The risk may vary from being knocked down, to being in a
car accident or suffering ill health due to lack of exercise. The same can be said
of fire safety in buildings. As long as we occupy buildings where there is a chance
that ignition sources and combustible materials may be present together, there will
be a risk of death and injury due to fire in addition to property damage. We need
not be fatalistic, however, this simply identifies the need to manage the risk. It also
indicates that, although we should work towards reducing risk, the ultimate goal
of zero risk is not currently a realistic expectation.
As Benjamin Franklin once said, ‘But in this world nothing is certain, but death
and taxes’. It follows that whilst we live there is a risk of death and the only way
of not dying is not to live in the first place. This may be s elf-evident, but it is very
important when we start to consider specific risks that we consider them in the
context of other risks.
There is a practical benefit to looking at risks in context. Society may decide
that it would like to dedicate more resources to addressing one risk than another.
For example, for healthcare, it may typically cost about £20,000 to save a life,
whereas for fire safety in buildings, it may cost more than say £1million to save a
life (Charters 1996). Therefore, society (and/or its representatives) could decide
to put more resources into healthcare than into fire safety in buildings. Equally,
society may be more concerned about the suffering of people killed and injured
by multi-fatality fires than it is about the provision of every possible healthcare
intervention to all patients, irrespective of need or prognosis.
For fire safety in buildings, the annual fire statistics indicate that there is a finite
level of fire risk in buildings (Office of the Deputy Prime Minister 2005). This may
also indicate that if a building complies with the appropriate fire safety standards,
then its level of fire risk is broadly tolerable (or possibly acceptable). It could also
be said that applying the fire standards to a non-standard building could result in
intolerable levels of fire risk. However, no criteria for fire risk in buildings have
been set in the UK (British Standards Institute PD 7974 Part 7 2003).
For the fire safety engineering of a non-standard building, this means that the
level of risk should be designed to be the same or lower than that for an equivalent
standard building (Office of the Deputy Prime Minister 2005; British Standards
2 Quantitative Risk Assessment in Fire Safety
Institute BS 7974 Code of Practice 2001). Since there are no quantitative risk
criteria, this means comparison of a non-standard building with a compliant
building. Similarly, since regulations are not generally framed in terms of risk,
this results in an assessment of physical hazards and the balancing of a qualitative
arguments about risk. So, in conclusion, we could say that, with respect to fire
safety in buildings, we are all taking a risk but rarely, if ever, is it calculated.
The risk of an undesirable event can be defined as the combination of:
the severity of its outcome (its consequences); and
how often it might happen (its frequency) (British Standards Institute PD
7974 Part 7 2003).
If we simply consider the potential consequences of fires, then we may not
be adequately addressing fires with lower consequences but whose risk is much
higher due to their frequency. It is easy to imagine very severe fire events in
buildings that have no fire precautions. We can identify ignition sources and
combustible material, estimate how quickly a fire would grow and how quickly
untenable conditions would develop. We can then consider the potential number
of occupants, their likely behaviour and how quickly they might escape from the
building. It is less easy to imagine very severe events in buildings with many fire
precautions, but they can, and do, occur. Severe fire events like these seem to
catch us by surprise and are characterised by a series of failures of the many fire
precautions. The lower frequency of these events leads to the sense of surprise.
The fact that a series of failures had to occur before the severe event occurred
indicates that we had ‘defence in depth’. Defence in depth is where many systems
are present, but only a few need to work to achieve a safe outcome. Finally, this
indicates that all fire precautions have a level of unreliability and occasionally
many of the fire precautions may fail at the same time, leading to the severe event.
Fortunately, these very severe events are relatively rare, but this rarity means that
it is very difficult to assess their frequency directly. For example, if we have a
thousand buildings of a certain type and we want to address fires that may occur
once every million building years, then on average this fire could be expected only
to happen once every thousand years. This is a long period over which to collect
and analyse data and make a decision about the adequacy of the fire precautions.
What makes it even harder is that this ‘one in a million building years’ event
could happen tomorrow and the day after. This makes the direct estimation of
high consequence/low frequency events very uncertain. Therefore, we use events
that happen more often to estimate the frequency of rarer events. The frequency
of ignition is the usual starting point for fire safety in buildings. To estimate the
frequency of the severe fire events we need to understand how ignition may, or may
not, lead to them. This involves the study of the reliability of fire safety systems.
For example, probabilities can be attributed to the following series of events:
Does the fire grow?
If yes, is the fire detected early?
If yes, is the fire extinguished using first aid fire fighting?
Is the fire suppressed and/or vented by automatic systems?
Does the fire spread beyond the compartment of fire origin?
In this way we can assess the impact of fire retardant or non-combustible
materials, fire detection, extinguishers, sprinklers, vents, compartmentation etc
on reducing the frequency of severe fires if, say, all these fire precautions fail.
We can combine the frequencies and consequences of these severe events to
estimate their levels of risk. The levels of risk can be expressed as a frequency
distribution (or ‘Fn’ curve) of risks, with lower risk events near the origin and
higher risk events to the upper right-hand side. See Section 1.4 on acceptance
criteria for risk and the role of Fn curves. Some events may have trivial or no
consequences and so can be discounted from the analysis. Other events are high
risk and should be addressed.
Because there are no quantitative criteria for the fire risk in buildings in the UK,
it is not possible to say in absolute terms when the risk is low enough. However, if
we know where a compliant standard building lies on the distribution, it is possible
to do it by comparison. So we can say that we are safe enough, when the risk is
lower, than a compliant standard building.
1.1 Fire engineering
Fire engineering can be defined as (BS 7974 2001):
the application of scientific and engineering principles to the protection of
people, property and the environment from fire.
Fire engineering can address one or more objectives. These objectives can
property loss prevention/business continuity; and
Fire engineering is generally more demanding, technically and in terms of
resources, than the application of simple prescriptive fire safety guidance used
to support building regulations (ADB 2007). Therefore, fire engineering is
generally used where simple fire safety guidance may not adequately address
the fire scenarios or issues of concern. Simple fire safety guidance may not
adequately address the fire scenarios or issues of concern when the building
is large, complex or unique, or when application of the simple prescription
conflicts with the function of the building (usually rendering the fire precaution
highly unreliable) or when application of the simple rules is not the most cost-
effective approach. Other factors that might indicate where fire engineering is
typically used include:
4 Quantitative Risk Assessment in Fire Safety
when there is an atrium;
when there are multiple purpose groups in one building;
where a highly innovative design is used to facilitate the function of the
when there are unique or challenging fire hazards.
Examples where fire engineering is typically used include the larger assembly
buildings, hotels, hospitals, industrial and commercial premises, transport
interchanges and tunnels, landmark, heritage and headquarters buildings, ships
and offshore installations.
1.2 Deterministic approaches
Deterministic approaches use quantitative analysis of physical processes, such
as smoke movement and evacuation, to aid decision-making during the design
process (BS 7974 2001). To understand the deterministic approaches, a more
detailed consideration of the design process may be of benefit.
Design can be characterised as an essentially creative, largely intuitive, often
divergent process where many design parameters are manipulated in three and
sometimes four dimensions, to meet multiple design objectives. Design objectives
may include function, efficiency, cost, buildability, safety, durability, reliability,
aesthetics etc. Design parameters may include the location of the site, the size
and interconnection of different spaces, means of access and egress, the number
of levels, the type of construction etc. For fire safety design, the main design
objective is life safety (occupants, fire fighters and others in the vicinity), but other
objectives such as property protection/business continuity and the environment
may also be included. In many respects, the use of simple prescriptive guidance
suppresses the design aspect of fire safety and can encourage the perception that
there is only one solution and ‘this is it’. However, the nature of many modern
buildings means that simple fire safety rules are increasingly difficult and costly to
apply. In these circumstances alternative fire safety design solutions are required
to ensure that an appropriate level of safety is achieved. This is where analysis
is used to assess the level of safety and indicate whether the fire safety design
objective(s) have been met. Analysis can be characterised as an essentially logical,
structured, rigorous process which takes certain input parameters, undertakes a
series of operations/calculations and produces one or more output parameters. For
fire safety analysis, the input parameters could include the material contents and
their burning characteristics, size of a space, the number, width and distribution
of its exits, the number of occupants, a walking speed and a flow rate through the
exit. Deterministic analysis can be defined as the use of point values for the above
variables and a purely physical model. Through the use of deterministic egress
calculation methods a single value of the output parameter of time for occupants
to move through an exit can be calculated (see Section 3.6 on consequence
analysis). This single value is regarded as an exact value, ignoring the uncertainties
governing the input and output parameters.
Decision-making can be characterised as an essentially convergent process
of identifying an issue that needs to be addressed, gathering information on it
(including that from analysis) and reaching a conclusion based on the evidence.
For fire safety, the issue could be the number and width of exits required for egress
from a large and complex space. Based on knowledge of the number and width of
exits for smaller/simpler spaces, the basis of the input data, analytical model and
results of the analysis, a decision on the appropriate number of exits for a large and
complex space can be made (Charters 2000).
Quantified fire risk assessment may best be used where deterministic fire safety
engineering may not adequately address the fire scenarios of concern. This tends
to occur when the (life safety, loss prevention and/or environmental protection)
consequences of a fire may be very high/intolerable.
1.3 Probabilistic approaches
Fire risk assessment allows the performance and reliability of fire precautions to be
explicitly taken into account in the fire safety engineering of a building (PD 7974
Part 7 2003). It can be used to:
select fire scenarios for deterministic analysis; and/or
quantify levels of fire risk (to life, property, business and/or the environment).
In the selection of fire scenarios, an event tree for a broad class of occupancies
and first order data estimates of frequencies and consequences may be used to
identify and define an appropriate fire scenario for deterministic fire engineering
analysis. It is important that the scenario selected provides a reasonably severe
challenge to the fire safety design, yet is reasonably credible in terms of its
frequency. Typically, this may include the failure of fire prevention, reaction to
fire of materials, natural fire breaks, first aid fire fighting and non-fire resisting
construction. Additional scenarios can be analysed to assess the dependence on a
particular fire safety system such as sprinklers, smoke control or compartmentation.
Fire risk assessment can also be useful in quantifying the levels of risk for an
individual building design:
in demonstrating equivalent life safety to a code-compliant building or
satisfaction of fire risk criteria;
for cost-benefit analysis of property protection/business continuity; and/or
in assessing the environmental impact of large fires.
Therefore, circumstances where fire risk assessment could be useful include
Life safety is affected by buildings:
• containing a very large number of people who are unfamiliar with the
Author David Charters and Ganapathy Ramachandran Isbn 9780419207900 File size 10MB Year 2011 Pages 384 Language English File format PDF Category Architecture Book Description: FacebookTwitterGoogle+TumblrDiggMySpaceShare Fire safety regulations in many countries require Fire Risk Assessment to be carried out for buildings such as workplaces and houses in multiple occupation. This duty is imposed on a “Responsible Person” and also on any other persons having control of buildings in compliance with the requirements specified in the regulations. Although regulations only require a qualitative assessment of fire risk, a quantitative assessment is an essential first step for performing cost-benefit analysis of alternative fire strategies to comply with the regulations and selecting the most cost-effective strategy. To facilitate this assessment, various qualitative, semi-quantitative and quantitative techniques of fire risk assessment, already developed, are critically reviewed in this book and some improvements are suggested. This book is intended to be an expanded version of Part 7: Probabilistic risk assessment, 2003, a Published Document (PD) to British Standard BS 7974: 2001 on the Application of Fire Safety Engineering Principles to the Design of Buildings. Ganapathy Ramachandran and David Charters were co-authors of PD 7974 Part 7. Quantitative Risk Assessment in Fire Safety is essential reading for consultants, academics, fire safety engineers, fire officers, building control officers and students in fire safety engineering. It also provides useful tools for fire protection economists and risk management professionals, including those involved in fire insurance underwriting. Download (10MB) Building Dynamics: Exploring Architecture Of Change Universal Principles of Design: 100 Ways to Enhance Usability, Influence Perception, Increase Appeal, Make Better Design Chartered Institution of Building Services, Transportation Systems in Buildings Concrete Buildings Scheme Design Manual Building Control with Passive Dampers Load more posts