Quantitative Risk Assessment in Fire Safety by David Charters and Ganapathy Ramachandran


5150Od0kwL._SY291_BO1204203200_QL40_.jpg Author David Charters and Ganapathy Ramachandran
Isbn 9780419207900
File size 10MB
Year 2011
Pages 384
Language English
File format PDF
Category architecture



 

Quantitative Risk Assessment in Fire Safety Quantitative Risk Assessment in Fire Safety Ganapathy Ramachandran 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 professional practice. 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.     p. cm.   Includes bibliographical references and index.   1. Fire risk assessment. 2. Quantitative research. I. Charters,   David A. (David Anderson), 1949– II. Title.   TH9176.R36 2009   363.37’7—dc22 2009022649 ISBN13: 978-0-419-20790-0 (hbk) ISBN13: 978-0-203-93769-3 (ebk) Contents List of figures List of tables About the authors Preface Acknowledgements 1 Introduction vi viii x xi xiii 1 2 Qualitative and ­semi-­quantitative risk assessment techniques 34 3 Quantitative risk assessment techniques 70 4 Acceptance criteria 175 5 Initiation 196 6 Design fire size 239 7 Fire spread beyond room of origin 255 8 Performance and reliability of detection, alarm and suppression 268 9 Performance and reliability of human response and ­evacuation 295 10 Performance and effectiveness of fire service i­ntervention 310 11 Whole project ­analysis 327 12 Interactions 341 13 Combining data from various sources – Bayesian ­techniques 352 Index 359 Figures 1.1 1.2 1.3 2.1 2.2 2.3 2.4 2.5 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15 3.16 3.17 3.18 3.19 3.20 3.21 3.22 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 method An example of a matrix method flow chart Venn diagram 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 damaged 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 13 17 20 36 37 39 41 44 75 76 77 79 89 90 92 95 96 104 107 108 108 112 113 114 118 123 126 127 128 138 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 139 139 140 141 142 143 147 147 148 151 152 180 184 185 186 187 218 229 265 296 297 323 324 330 337 343 344 Tables 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14 2.15 2.16 A2.1 A2.2 A2.3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 Typical matrix method record sheet 45 Hazard /accident frequency 46 Accident trigger probability 46 Accident frequency categories 47 Accident severity 47 Risk classification 48 Risk tolerability 48 Hazard frequency 54 Accident severity (life safety) 54 Accident probability 54 Cost of asset damage or destroyed 54 Cost of performance penalties or operational loss 55 Life safety risk rating 55 Business risk rating 56 Overall station risk rating 56 Example of part of a risk ranking spreadsheet (most likely fire event) 57 Relative S values for active systems 65 Values of components 66 Sample of a summary sheet 67 Relationship between random processes and probability modelling 72 Truth table for an OR gate 78 Truth table for an AND gate 78 Frequency of ignition per year 87 Frequency of ignition normalised per building 88 Frequency of ignition normalised by floor area 88 General data on the reliability of fire protection systems 93 Data used to quantify the probability of failure to detect a fire within five minutes of ignition 97 3.9 Annual chance of a fire outbreak for various occupancies 99 3.10 Probable damage in a fire – parameters of Equation (3.41) 100 3.11 Industrial buildings – annual damage 102 3.12 Retail premises – assembly areas – frequency distribution of area damage 115 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 sprinklers 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 occupancy dwellings 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 117 129 129 130 167 177 178 192 194 198 204 207 210 212 221 227 231 232 240 241 243 244 246 247 256 258 283 284 290 307 318 320 322 339 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 Maunsell/AECOM. 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. Preface 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 xii  Preface 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. Acknowledgements 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 book. 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 Communications. 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 Publishing Ltd. 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. xiv  Acknowledgements 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. 1 Introduction 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? Introduction  3 • • • 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 include: • • • life safety; property loss prevention/business continuity; and environmental protection. 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 building; or 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. Introduction  5 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 where: 1 Life safety is affected by buildings: • containing a very large number of people who are unfamiliar with the building;

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

Leave a Reply

Your email address will not be published. Required fields are marked *