=> ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY NSW CHAPTER Practice Note for Design Fires Version 1.0, dated 22.03.12 Prepared by Society of Fire Safety NSW Chapter Engineers Australia ‘THE SOCIETY OF FIRE SAFETY, ENGINEERS AUSTRALIA AND THEIR MEMBERS ACCEPT NO LIABILITY FROM ANY DAMAGES, LOSS OR INJURY RESULTING FROM THE USE OF THIS DOCUMENT Sociely of Fire Safety NSW Chapter PO Box 2466, Werrbeo, VIC, 3030 Tek + 619.9965 8666 + Fox 14) 2°9865 8415, Email inlo@[Link] + Websie: nw [Link]ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE Fi ‘OR DESIGN FIRES. Contents 1 INTRODUCTION LIMITATIONS OF THIS PRACTICE NOTE DESIGN FIRE SCENARIO 3 Introduction 3.2 Selection of Design Fire Scenarios 3.2.1 Simplified Analysis, 3.2.2 Complex Projects ~ Event Tree Analysis 3.2.3 Risk Ranking 3.3. Redundancy. 3.4 Sensitivity 3.5 Fire Brigade Intervention 4 DESIGN FIRE DERIVATION. 4 Introduction 4.2 Design Fire Derivation Strategy 4.2.1 Identify Design Occupancy 4.2.2 \deniily Fire Sofely Design Issues. 4.2.3 Identify Relevant Subsystems. 4.2.4 Determine Characteristics of Design Fire 5 CHARACTERISTICS OF DESIGN FIRES 5. General. 5.2 Location of Design Fire 5.3 Type of Fire. 5.4 Type and Amount of Fuel 5.5 Radiative Fraction. 5.6 Soot and Species Yields and Heat of Combustion. 5.7 Heat Release Rate 5.7.1 Incipient Stoge 5.7.2 Growth Stage 5.7.3. Fully Developed Stage ~ The Maximum Heat Release Rate of Unintervened Design Fires 5.7.4 Impact of Fire Safely Systems [Link] The Maximum Heat Release Rale of Inlervened Design Fires [Link] Impact of a Smoke Control System 5.7.5 Impact of Building Element Failure. 5.7.6 Duration of Burning Stage 5.7.7 Decay Stoge 6 EXAMPLE APPLICATION ON A LOW RISE SHOPPING CENTRE. 7 REFERENCES. APPENDIKA Event Trees Version 1.0, dated 22051 SREP EGGSSSS TI U USS eeu aaaan 20 20 22 22 22 22 23 31 34ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES List of Tables Table 5-1: Species Yields ond Heat of Combustion Table 5-2: Fire Growth Rates for Typical Occupancies. Table 4-1: Low Rise Shopping Centre Example List of Figures Figure 4-1: Design Fire Derivation Strategy. Figure 5-1: Ideolised Heat Release Rate Curve Figure 5-2: Determination of Heat Release Rate of Design Fires Figure 6-1: Event Tree in Major Retail Tenancy Figure 6-2: Event Tree in Moll Figure 6-3: Event Tree in Carpark, Version 1.0, dated 2203.12 18 25 u 16 28 29 30ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page4 of 35 1 Introduction Since the inception of the performance based building code in Australia, fire safety engineering as a discipline has grown significantly. Along with the advancement in fire safety seience, fire safety ‘engineering practice is also evolving, The process, the methodology and the data can be regarded as, the three pillars of fire safety engineering, as noted in the International Fire Engineering Guidelines (IFEG)'”. A rational approach to fire safety engineering design relies on clarity in the process, as well as scientific rigor in the analysis and verifiable data, The process is an overarching framework within which engineering designs are carried out, and the design methodology and data are implemented. The development of the process is an essential part of advancing fire engineering science and it often stems from practitioners’ experience and the need for consistent quality assurance in engineering practice. ‘On 16 June 2009, the Society of Fire Safety New South Wales chapter held a seminar on the topic of | Design Fire. The issue of consistency in engineering design practice was discussed. The attendees felt that there was no specific guidance to practitioners on how to derive design fires, and that there were discrepancies found in fire engineering reports prepared by different consultants, or in some instances, by different offices of the same consultant, in terms of the design fire adopted for projects. As an ‘outcome of the seminar, a consensus was reached that there is a need of clear guidance to practitioners ‘on how to derive design fires to address the problem of discrepancies mentioned above. ‘According to the IFEG!", in order to specify the design fires that are to be used in a fire engineering ‘evaluation, three steps should be undertaken: 1. Determine potential fire scenarios. 2. From these possibilities, select the design fire scenarios to be used for developing the design fires. 3. Forcach of these design fire scenarios, specify a design fire. With respect to Step 3 above, the limitations of the tools available to fire safety practitioners to date do not allow prediction of the initiation and development of the design fire from first principles or even. through complex Computational Fluid Dynamics (CFD) models, but require the fire safety practitioners to prescribe the design fire. This Practice Note was drafted by a group of fire safety engineers on behalf of the Society of Fire Safety. It serves as a supplementary guide to the IFEG' for fire safety engineers and aims to assist the fire safety engineers to derive design fires and design fire scenarios in a consistent manner. However; ‘the Society of Fire Safety does not claim that the proposed method in this practice note is the ‘only method for derivation of design fire scenarios and design fires. This practice note is divided into seven sections. Section 1 provides the background of producing the note. Section 2 states the limitations of this note, the status of this note in regulatory regimes, and the responsibility of the users. Section 3 discusses the relationship between design fire scenarios and design fires, and the strategy to derive design fire scenarios. Section 4 describes the strategy to derive design fires. Section 5 provides information on the types and characteristics of design fires, and items to consider for derivation. Section 6 includes an example of how the practice note can be used to derive design fires, using a low rise shopping centre for illustration. Section 7 lists the literature referenced for producing this practice note. Appendix A describes the mathematical relationships ‘within event trees, Version 1.0, dated 22.03.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 5 of 35 2 Limitations of This Practice Note This practice note has been prepared as a guide to practitioners on derivation of design lire scenarios and design fires, for the purpose of enhancing consistency in the fire engineering industry. The proposed methodology should not be claimed to be the only means of derivation, This practice note should be used in conjunction with the International Fire Engineering Guidelines. Whilst the design fire scenario derivation methodology has employed a risk based approach, this practice note was predominantly prepared for the derivation of design fires to be used in fire ‘modellings or calculations in a deterministic analysis. Where a semi-probabilistic or probabilistic approach is to be employed, relevant guidance from the fire engineering literature should be consulted. ‘This practice note has been prepared under the regime of the National Construction Code of Australia and may not be 100% applicable under building regulatory regimes in other countries. ‘The data presented in this practice note was based on the information in fire engineering literature available to the working group at the time of preparation. It is the onus of the practitioners to make sure that the data are accurate, up to date, and applicable to specifi projects with which the practitioners are involved, This pract note does not address acts of terrorism, explosions or multiple occurrences of arson, Version 1.0, dated 22.03.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 6 of 35 3 Design Fire Scenario 3.1. Introduction Fire safety engineering analysis is by and large a mixture of deterministic and probabilistic analyses. ‘These types of approaches are undertaken in engineering assessment processes at different levels. In analysing a design fite scenario, the most common approach is deterministic, and most commonly used computational tools are of this type. However, the identification of the design fire scenarios for the engineering analysis relies on a probabilistic approach, either explicitly or implicitly. Ithas long been recognised that fire safety engineering design parameters are associated with uncertainties due to the complex nature of reality”. The number of possible fire scenarios in reality may approach infinity and it would be extremely difficult to base a fire safety engineering design on the analysis of all possible fire scenarios!) Instead, a limited number of typical, or so called “worst credible”, scenarios may be used to carry out the analysis"! ‘The term ‘fire scenario’ is defined as a qualitative description of the course of a fire (event) with time. ‘A design fire scenario is a specific fire scenario of which the sequence of events will be quantified and a fire safety engineering analysis will be conducted" A design fire is @ quantitative description of assumed fire characteristics within the design fire scenario, Design fires are part of design fire scenarios and are inseparable from the design fire scenarios, The determination of design fires is an important step in fire engineering analyses. Although design fire scenarios and design fires are intrinsically related, a differentiation is made in the ‘current document, This chapter focuses on the selection of design fire scenarios. Detailed discussions ‘on the determination of design fires are given in Chapter 4, Design fire scenarios and design fires are also closely related to building occupancy and the fire safety design issues, or the proposed Alternative Solutions. There is no design fire scenario or design fire that, will fit all building occupancies or can be used to assess all Alternative Solutions, For the purpose of discussion in this note, the building occupancy and the Alternative Solutions that are dealt with in a fire engineering project are referred to as design occupancy and fire safety design issues, 3.2 Selection of Design Fire Scenarios ‘The selection of design fire scenarios is part of the Fire Engineering Brief (FEB) process as outlined in the IFEG. The design fire scenarios should be agreed by all stakeholders of the project. A systematic approach to the identification of significant fire scenarios and the selection of design fire scenarios for quantitative analysis in fire safety design is documented in ATS $387.2", ISO/TR- 13387! and the SFPE Handbook", where the selections of design fire scenarios and design fires are ‘combined. In the current document the selections of design fire scenarios and design fires are discussed separately ‘The selection of design fire scenarios depends on the building classification, project issues and the relevant fire safety subsystems installed in the building. For example, the sprinkler system installed in « high rise apartment building means that its design fire scenario will be different from that for a low rise building of the same classification without sprinklers. Generally, multiple design fire scenarios need to be analysed as a requirement for sensitivity studies”, Based on the SFPE Handbook’ and ISO/TR-13387", a procedure for the derivation of desi scenarios is outlined below: 1, Identify the design occupancy and the fire safety design issues; Version 1.0, dated 22.03.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 7 0f 35, 2, Identify all subsystems related to the fire safety design issues; 3. Identify all possible statuses of each and every subsystem relevant to the fire engineering assessment; 4, Quantify the associated probabilities; 5. Use an event tree to establish the possible scenarios; 6. Calculate the probability of each scenario; 7. Estimate the potential consequences of all the scenarios using engineering or expert judgement; 8. Use risk ranking and / or probabilities to select the design fire scenarios for analysis. Whilst the quantitative evaluation of risk relies on the outcome of the fire scenario analysis, it may be possible to combine the event tree analysis and expert judgement in risk ranking to determine the fire scenarios for engineering assessment, Whether such a process should be undertaken entirely, or partially, for a specific project is to be determined by the project stakeholders. It may depend on the seale of the project and the nature of the design solutions, In case insufficient data of experience or expertise is available for expert judgement, the selection of fire scenarios may be based on probability of an event, The base case design fire scenario would be a fire scenario in which all proposed fire safety measures in the trial fire safety design are assumed to operate and achieve their design objectives. Other design fire scenarios would be fire scenarios where one or a combination of the proposed fire safety measures are assumed to fail or cannot achieve their design objectives; these scenarios are likely to include low probability/high consequences fires. 3.2.1 Simplified Analysis For simple projects it may not be necessary to go through all the above steps. When steps I and 2 have been undertaken, existing guidelines such as CIBSE Guide E'’ can be utilised to determine design fires. This approach may be appropriate: + When only minor non-compliances with the Deemed-to-Satisfy Provisions are addressed. + For comparative assessments where layout of building and possible consequences are similar to a Deemed-to-Satisty compliant design. More complex projects may also be suitable for a simplified analysis; however this would depend on the issues being assessed and would require agreement from stakeholders. 3.2.2 Complex Projects — Event Tree Analysis scenarios outlined in the existing literature, such as ATS 5387.2", ISO/TR-13387"" and the SFPE Handbook", consists of a number of steps. The core of the design fire scenario selection process is the event tree analysis, It is through this analysis that, probabilistic description and evaluation of events is carried out, ‘The approach to the selection of design In the event te, the initiating event, namely the stat of afte, isthe cause ofthe subsequent events, Which are represented by the nodes that branch off the previous event. Each branch in the event tree represents a scenario, Each event in the event tre will have an associated probability of occurrence ‘and consequence. These probabilities will help the risk ranking process through which the design fire scenario can be selected!) Figure 3-1 is an example of an event tree. A more detailed description of ‘event tres can be found in Appendix A. ‘The probabilities for each event can be determined based on published statistical data or can be determined by using fault tree analysis Version 1.0, dated 22.03.12ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 8 of 35 Note that the Fire Initiation and Development and Control subsystem (Sub-system A") occupies a special place in the event tree in that its status, or nodes, represent the initiating events. Each node represents an initiating design fire which can also be regarded as the descriptions of the status of this particular subsystem. Like any nodes in the event tree, initiating design fires have associated probabilities of occurrence. In that sense, each design fire defines an event in the fire initiation subsystem, More comprehensive discussions of design fires are given in Chapter 4 of this document. Aw Suh-event Snhcevent — Snb-event Serena Ps Pa Po Ps Ps B Po A Pa Ps Po Pa B Pa Pa B Pa R Pa Pra Ps Figure 3 Example of an Event Tree! "ssub-event inthe diagram mears “subsequent eventENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 9 of 35 3.2.3 Risk Ranking isk ranking is an important step in the selection of design fire scenarios. Risk is defined as the ‘expected loss, or the product of the probability of a scenario and the consequence of that scenario! For a particular scenario i 9 R= PC, 3-1) where R; P, and C; are the risk, probability of occurrence and consequence of scenario i respectively. It is seen that risk ranking requires not only the knowledge of probabilities associated with various possible scenarios, or the events that constitute scenarios, but also the knowledge of consequences of the scenarios, which is not usually a priori before the quantitative analysis is carried out. Engineering, judgement may be employed in determining the possible consequences", If insufficient information or experience can be referred to, then risk ranking may be reduced to probability ranking and the scenario selection may be based on probability alone. Fire incident statistics with reasonable qualities can be referred to as a source of information for identification of the initial set of possible ignition scenarios. More specifically, quality fire incident statistics can be used to determine (a) the most likely type(8) of ignition scenario; and (b) the most likely ignition scenario(s) with significant consequence, ‘The design fire scenarios can be determined on the basis of risk ranking. The scenarios with significant risks in the ranking or with high probabilities in the ranking should be considered for analysis, 3.3 Redundancy Redundancy analysis examines the redundant measures of a trial design that essentially fulfils the same function, For example a smoke exhaust system may be activated manually, by detect sprinklers. Its not expected that each redundant component will deliver the exact same performance!” Consideration should be made for redundancy in fire safety designs, particularly when the ‘consequences of some of the scenarios may not be quantifiable for the risk ranking before the ‘engincering assessment is carried out. Examination of redundancy can be carried out by rendering one ‘or a combination of key fire safety systems to be ineffective in order to test the robustness of the design, It should however be noted that the safety margin can be relaxed in comparison to that used in the assessment of the tral fire safety design'™” Consideration of the following should be taken when determining if'a redundancy analysis is required: # Are there single points of failure? + Is the consequence in case of failure comparable to a Deemed-to-Satisfy compliant design? ‘Is the reliability of the system comparable to a Deemed-to-Satisfy compliant design’? Ifthe risk is considered higher than a comparable Deemed-to-Satisty compliant design or unacceptable, it can be lowered by reducing the probability of the scenario to occur (for instance ‘through enhancing the reliability of key fire protection systems) or by modifying the fire protection systems to reduce the consequence. The IFEG!" states that “Care and judgement should be used to avoid unnecessarily analysing events with a very low probability of occurrence, but where the scenario may have a very high adverse consequence, due consideration should be given if not for the primary analysis at least in the sensitivity studies’. It has been recognised that in the design of major ‘commercial complexes it is better to attempt to reduce the probability of occurrence of catastrophic ‘events than try to design for these events!" Version 1.0, dated 22.03.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 10 of 35 3.4 Sensitivity Sensitivity analysis measures the impact on the results of the fire engineering analyses by changing ‘one or more key input variables. This is especially important if there is some doubt about their ‘quantification or evidence that there is a heavy reliance on a particular fire safety measure. It allows the identification of those parameters that are most important to the outcome!” 3.5 Fire Brigade Intervention ‘A performance based Alternative Solution should account for, and facilitate, fire brigade intervention ‘where it is deemed relevant to do so. In accounting for fire brigade intervention, fire fighters should be given a reasonable time to undertake search and rescue activities before conditions in the building or structure(s) become untenable or unsafe for fire fighters, Consideration for fire brigade intervention and conditions against the tenability limit criteria for fre fighters needs to be evaluated at the anticipated arrival time, Given the nature and use of buildings, the model duration may also need to be extended to consider the time fire service personnel inside a building for search and rescue operations during which tenable conditions for fire fighters and structural stability may need to be maintained. ‘A performance based Altemative Solution should not use fire brigade intervention in isolation as a ‘means or justification of the issues being addressed. Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES lors 4 Design Fire Derivation 4.1 Introduction ‘This section illustrates the design fire derivation procedure. Whether a desiga fire is to be considered ina fire engineering assessment is dictated by the design fire scenario derivation procedure. As illustrated in the previous chapter, a design fire constitutes the status of the Fire Initiation and. Development and Control sub-system and is one of the determining factors of design fire scenarios. A design fire is characterised within the following sub-sections, 4.2. Design Fire Derivation Strategy Figure 7-1 presents a recommended process for the derivation of design fires. Further illustrations of the process are presented in the subsequent sections. Identify design occupancy Identify design issues I Identify relevant subsystems I Identify design fire scenarios via event tree or simplified method I Determine characteristics of design fire (see Section 5) (ona) Figure 7-1: Design Fire Derivation Strategy 4.2.1 Identify Design Occupancy Important items to consider here include: + Building classification based on BCA classification + Occupaney type * Occupant characteristies + Relevant hazards Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 12 of 35 4.2.2 Identify Fire Safety Design Issues Important items to consider here include: * Involvement of all relevant stakeholders + Identification of applicable fire safety issues (Le, issue relating to egress will result in a different fire scenario to an issue relating the structural capacity) 4.2.3. Identify Relevant Subsystems Fire safety subsystems that could have an impact on the growth and development of fires should be identified, These subsystems typically include fire suppression or extinguishing systems, smoke control systems, and (failure of) building elements. Potential impacts of these subsystems on design fires are discussed in Section 5.7.4 4.2.4 Determine Characteristics of Design See Section 5, Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES 13 0f 35 5 Characteristics of Design Fires 5.1 General Design fires are usually characterised in terms of the following parameters: + Location (e.g, at comer or middle of room, height of fire, near exit, in unoccupied room, in concealed space, or external), + Type of fire or buming regime (e.g,, smouldering, flaming or fully developed; fuel or ventilation controlled), + Type and amount of fuel (material, configuration, fuel load ... ete), + Radiative fraction, + Soot and species (e.g., CO, CO:, HCN, HCI ete) yields, + Heat of combustion, + Heat release rate (HRR) history, + Interaction with other fire safety measures. ‘Other parameters such as HR per unit area or volume, fuel ignition temperature, emissivity and fuel ‘material chemical composition may be required as inputs to sophisticated fire models. An understanding of the building hazards will assist in determining relevant parameters, Al of the parameters listed above are interrelated. Some parameters are usually specified as functions of time), The specifications of design fire parameters depend on the fire safety design issues at ‘hand, building layout and the level of sophistication of fire simulation models used. Some of the design fire parameters are discussed in the following sections. For the purpose of the present document, the burning behaviour of design fires can be categorised into, ‘wo types; intervened and unintervened. An intervened design fire will interact with other fire safety subsystems and its HRR will change according to the change in the status of other subsystems (see Section 5.7.4 and Figure 7-3). Unintervened design fires in this practice note refer to design fires that have not been subjected to any ‘manual fire suppression (by occupants or fire brigade) or automatic fire suppression or change in ‘mechanical ventilation conditions, Its considered that in most cases an arson fire will be covered by the following process. Should the ‘occupancy, contents or other building specifies result in a deliberately lit fire outside the design fire characteristics presented below, and this scenario is considered a likely event, a more detailed analysis may be required. 5.2 Loca ‘The fire location is related to design occupancy and design issues. Variation of fire location may also be required as part of the sensitivity study in an engineering assessment, ‘The selection of fire location ccan be aided by event tree analysis whereby the locations at which fires have either highest likelihood ‘of occurrence or most adverse consequences may be selected. The selection list is likely to include, but ‘may not be limited tol", n of Design Fire ‘© Challenging fire ~ credible worst case scenario that will challenge the fire protection features ofa building Blocked exit fire located near primary exit Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES. 4 of 35 + Unoccupied room / concealed space + Smouldering fie ~in proximity to sleeping occupants + Potential spread to neighbouring properties + Fire source external to building Moreover, precise fire source location in the room of fire origin may be required by some fire models to determine entrainment rate and detector or sprinkler activation time. 5.3. Type of Fire Fire burning behaviour has been traditionally classified into smouldering, flaming and fully developed. With the sophistication of modem fire models, the need to distinguish flaming and fully developed fires in the design fire specification process is diminishing. However, smouldering fire is a special ‘mode of burning and may need to be considered for certain design occupancies and fire safety design issues, 5.4 Type and Amount of Fuel The first step to establish a desig fire isto identify potential fuel and its form (gaseous, liquid and/or solid) in the design occupancy. The configuration of the fuel items is also an important parameter that ‘may influence the buming behaviour. 5.5 Ra tive Fraction Radiative fraction of the heat released from fires depends on the type of fuel, combustion efficiency, soot concentration and the diameter of burning surface area. It may vary between 0.1 and 0.35!" The lower bound represents clean fuel fires similar to alcohol fires whilst the upper bound represents, smoky hydrocarbon fires where sooty smoke is released, Without further details on the characteristics ‘of the fuel load, this fraction is typical assumed to be 0.3, 5.6 Soot and Species Yields and Heat of Combustion ‘The values of toxie species yields (e.g., CO and CO;) and smoke/soot yield for a wide range of ‘commonly encountered combustible materials are presented in the SEPE Handbook!" Soot yield is a critical input to fire models for predicting visibility through smoke. However, there is little validation ‘of such predictions of smoke visibility for real fires in large compartments. Therefore visibility predictions for real-scale fires using fire models should be treated with caution, using engineering judgement, In particular, when these soot yield "data" are used in CFD simulations, the recent research findings”) as summarised below should be taken into account "The study showed that FDS significantly over-predicted smoke concentrations compared to the experiments, Exploratory findings indicate that soot deposits to the ceiling above the plume may be as high as 37 percent of the soot produced, Current versions of FDS do not account for this substantial soot loss. Another important finding of this work was a large diserepancy between reported small-scale soot ‘yields and yields from larger scale fires. Commonly used soot yields in the literature for a range of ‘fuel materials have been obtained for small fires conducted in bench-scale calorimetry apparatus. This work kas shown that measured soot yields for 100 KW fires range from 2 to 5 times smaller than the reported values. This finding is consistent with other data in the literature; however, there are ‘minimal published studies that address this issue." Version 1.0 dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Is 0f 35, ‘The values for yields of CO and CO; and smoke/soot for a few selected commonly encountered. combustible materials are presented in Table 7-1. These values are taken from the SFPE Handbook''® and represent data measured in the bench-scale ASTM E2088 fire propagation apparatus at FM Global, under well ventilated conditions with the sample material exposed to a ‘constant external radiant heat flux. The conditions such as ventilation and exposure of materials to radiant flux in real-scale fires can vary substantially from those in the bench-scale measurement, and will affect the soot and species yield values. ‘The heat of combustion, AH, isthe energy released as heat when a compound undergoes complete ‘combustion with oxygen under standard conditions. The values for a sclected few commonly ‘encountered combustible materials are presented in Table 7-1. These values are taken from the SFPE Handbook", the Fire Protection Handbook"? and the Fire Engineering Design Guide™, The user of this practice note should always refer to the latest edition of the literature for the most current data, ‘The heat of combustion is one of the factors required to quantify the pre-and post-flashover fire. With ‘other factors, it ean be used to determine the total energy in the fuel (MJ), the ventilation controlled heat release rate, whether flashover occurs or not, the equivalent time of fire exposure and the duration of the buming period. In most cases it is considered reasonable to use average values to cover the likely contents. However; in situations where the contents are not well mixed, the values for the worst-case contents should be used. In view of the above discussion, caution and good engineering judgement should be exercised in prescribing species yield values as inputs to fire models for prediction of parameters such as visibility through smoke, ‘Table 7-1: Species Yields and Heat of Combustion Material CO, yield | CO yield | Soot Vield | AMT, cr) ee cr ung) Gasoline 2.85 oon 0.038 437-468 (Heptane/ Octane) Kerosene 283 0012 0.042 2-468 Wood /eelislose [13 0.004 | 0015 16-225 PMMA 212 oor 0.022 249) Polyethylene [2.76 0.024 0.06 3-477 Polystyrene foam [233 0.06 ois 021 [397-422 Polypropylene [2.79 0.024 0.059 343-462 PVC 0.46 0.068 0078-0172 | 164-268 Polycarbonate | 15 0.054) 02 297-315 Polyester 16 007 0.09 28 Polyurethane 1s 0.01- 0.04 0.013- 0.23 2-28.64 foam Nylon 2.06 0.038 0.075 2317 Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 16 of 35 5.7 Heat Release Rate ‘One of the most important parameters to define a design fire is the HRR which is often specified as a function of time”, Figure 7-2 illustrates an idealised HRR history for a compartment fire, from ignition through to complete burnout. crowen Fly developed Decay Hoat Release Rate rately Supgowedy Time Figure 7-2: Idealised Heat Release Rate Curve ‘Note that fire brigade intervention has been ignored in the above curve. The resultant HRR may. ‘change at any stage of the fire curve due to fire brigade intervention, ‘The process of determining the HRR can be described by Figure 7-3 Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES 1 0f3s Estimate incipient duration Determine fire growth rate TS the fire intervened BY suppression subsystems? Yes Determine the time and effect of suppression Yes Is the fire fuel controlled? Determine the total fuel burning surface area amount of fuel = Determine the activation ‘time and air supply rate Ts the fire affected by mechanical ventilation? Yes Determine the time and extent of failure Determine the maximum HRR_fe——H4H4HH I £ Determine the duration of the burning stage and decay stage Figure 7-3: Determination of Heat Release Rate of Design Fires Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 18 of 35 5.7.1 Incipient Stage ‘The incipient stage is difficult to quantify. It depends on the ignition source, type of combustibles and ventilation conditions, Whether to include the incipient stage in the specification of HRR for a design fire is dictated by the design issue and the design occupancy. For example, in a large space where smoke control system capacity is a design issue, inclusion of the incipient stage is not critical. However, in a hospital where early detection is crucial, a design fire with incipient phase may be ‘considered in one of the design fire scenarios. 5.7.2 Growth Stage Iflarge scale experimental data of HRR versus time is available and is deemed to be appropriate for the scenario being analysed, it can be used to describe the fire growth stage of the design fire. The HRRs of some furniture and appliances have been documented in the SEPE Handbook”! When specific information about the combustible items is not available, the HRR of the fire growth phase is typically modelled using a t-squared growth rate, which depends on the type of combustibles and their configuration, e.g., horizontally distributed or stacked, The t-squared fire growth can be ‘thought of in terms of a burning object with a constant HRR per unit arca in which the fire is spreading ina circular pattern with a constant radial flame speed”. ‘The fire growth rates of various commonly encountered configurations of combustible materials are ‘well documented in literature including but not limited to: © BS EN 1991-1-2:2002"; ‘© PD 7974-[Link] Application of fire safety engineering principles to the design of buildings. Part 1: Initiation and development of fire within the enclosure of origin (Sub-system 1)"""; © BS 9999:2008™, © CIBSE Guide E: Fire Engineering! and © NEPA 920 ‘Typical growth rates for t-squared growth fires, as summarised from the above decumentation are listed in Table 7-2. This table provides a general guide only and docs not rule out other possible ‘growth rates. The fire safety engineer should examine the fuel properties, ventilation conditions and ‘even the possible ignition source to determine what is the appropriate growth rate. ‘Table 7-2: Fire Growth Rates for Typical Occupancies Building Occupancies Fire t squared Growth Rate (a) Residential Medium Hotel (Room) Medium Office Medium Retail Fast Kitehen Fast Picture gallery Slo Display area Slow to medium Storage and warehousing Ultrafast Teaching laboratories Fast Workshop Medium to Fast Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOC BTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES. Page 19 of 35 Building Occupancies Fire tsquared Growth Rate (@) Hospital (Room) Medium Library Fast Classroom of a school Medium ‘Gymnasium Medium “Assembly hall seating Medium to Fast Whether the condition inside the burning enclosure attains flashover depends on the enclosure ‘geometry, HRR and ventilation condition'""!. Modern sophisticated software packages are now capable ‘of predicting if, and when, flashover will be attained and even the consequence of flashover in terms of variation in the HRR due to change of ventilation conditions with reasonable accuracy. Fire growth rate may be simulated using sophisticated fire models with adequate sub-models for pyrolysis, flame and phase combustion. Caution, however, needs to be taken as fire growth models are not yet mature and validated to a stage where they can reliably predict the HRR of real fires. For most applications itis therefore recommended that the fire growth rate is determined by the user via the input variables, Estimation of the HRR of a fuel controlled fire using Computational Fluid Dynamics (CFD) models such as Fire Dynamies Simulator (FDS) must be considered wisely because the model output is sensitive to minor changes in the input properties of bounding surfaces”, 5.7.3 Fully Developed Stage — The Maximum Heat Release Rate of Unintervened Design Fires Unintervened design fires in this practice note refer to design fires that have not been subjected to any ‘manual fire suppression (by occupants or fire brigade) or automatic fire suppression (by fixed fire suppression systems) ora change in mechanical ventilation ‘Without intervention, fires may spread and continue to grow, depending on the availability of fuel and oxygen. At the early growth stage a fre is likely to be fuel controlled, whereas at the fully developed sage a fire is likely to be ventilation controlled if there is sufficient fuel available for combustion, In the fully developed stage, the HRR may be assumed to attain a constant (maximum) value due to either fuel controlled or ventilation controlled burning, For unintervened ventilation controlled fires, there are empirical formulas in fire engineering literature for estimating the maximum HRR that can bbe supported by the ventilation available!") ‘The peak HRR of unintervened ventilation controlled design fires is recommended to be the maximum HR estimated from these empirical formulas, For unintervened fuel controlled fires, the peak HRR of the fire would depend on the maximum possible burning surface arca and the HRR per unit surface area", Fuel composition and configuration (local ventilation) also influences the peak HRR of fuel controlled design fires. Fire science and engineering literature should be consulted on the burning characteristics of individual fuel items, such as furniture pieces, cars, stored product pallets, etc. Alternatively bench-scale to full-scale fire tests can be carried out to capture the HRR and species yield data of items that have not been documented in the literature. Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 20 of 35 ‘The fire safety practitioners are recommended to consider the following when deriving unintervened fuel controlled design fires: + Fire load density and distribution of fire load within the enclosure © Characteristics of fuel packages (such as flame spread rate, HRR, percentage of heat loss through radiation, critical heat flux at ignition etc.) within the enclosure © Arrangement of fuel packages (c.g. are packages stacked on top of cach other and spatial distribution of fuel packages) + Effect of radiant heat flux feedback from hot smoke layer ‘The Building Code of Australia (BCA) Deemed-to-Satisfy Provisions require the smoke exhaust rates for smoke hazard management purposes to be determined based on a peak heat release rate of S MW, 10 MW or 15 MW depending on the classification of the building”! These may not be appropriate for every situation, If these fire sizes are to he used, justification of why they are considered appropriate for the design should be provided and agreement sought from stakeholders. 5.7.4 Impact of Fire Safety Systems [Link] The Maximum Heat Release Rate of Intervened Design Fires Fire suppression systems, either manual or automatic, could alter the HIRR of a design fire, The maximum HIRR can be determined by the time and effect of the system activation, Manual intervention by occupants involves the use of devices such as fire hose reels and portable fire extinguishers. Manual intervention by the brigades may also occur ata later stage of the fire by fire hydrants, foam systems etc. When developing design fires on a deterministic basis, manual intervention is recommended to be ignored. If a probabilistic approach is applied, manual suppression ‘can be considered and modelled by reducing the probability of larger fires occurring. ‘The most common automatic intervention system within buildings is a sprinkler system. Other automatic intervention systems include gas suppression systems, water mist systems and fire fighting. foam systems ete. ‘The consequence of the intervention by suppression subsystems may be categorised as listed below and shown in Figure 7-2: 1. Controlled by a fire suppression system, ie, the HRR remaining constant following sprinkler activation; or 2. Suppressed by a fire suppression system, i.e, the HRR decreasing following sprinkler activation; 3. Controlled or suppressed by fire brigade. ‘A sprinkler controlled fire is often defined by a constant HRR following the activation of sprinklers. ‘The peak HRR is typically determined from the fire growth rate and time of the sprinkler activation using computer software and the following parameters"! © Rate of fire growth (c.g. fast or ultra-fast. + Radial distance between the fire and the sprinkler(s). To provide a worst credible estimate for sprinkler activation time it is recommended thatthe fire is located in the middle between sprinkler heads, ‘© Sprinkler head activation temperature and the RTI value. + Number of heads required for sprinkler control. It is recommended that this is based upon statistical data for the type of occupancy. Ifa deterministic approach is undertaken, Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 21 of 35 with appropriate safety factors applied, it is considered reasonable to use the number of heads controlling 90 % of the fires. This will determine if the radial distance to the frst, second or third row of sprinkler is appropriate to use for sprinkler activation calculations, © Height of base of fire to ceiling. For occupancies with the majority of the fuel load located above the floor level it may be appropriate to raise the base of the fire. If the configuration of fucl load is not known, the fire should be located at floor level. + Location of fire ~ e.g. whether the fire is located in the middle of the room or located lose to a corner / wall. Potential forthe fire to be shielded should also be considered, ‘© Any heat transfer loss from sprinkler bulb (o the sprinkler pipework, ie, C-factor. © Ambient temperature within area protected by the sprinkler system, ‘An automatic sprinkler system that is correctly designed is intended to control or suppress a fire. In many instances, the fire size will be reduced or suppressed following sprinkler activation. If specific data is not available or shielded fires are likely, itis recommended to assume that a sprinkler system ‘controls the fire and that the fire remains at the fire size reached at the time of sprinkler activation for the purpose of developing a design fire. Otherwise, based on the reliability of modem sprinkler systems it is considered reasonable to assume suppression or a reduction in fire size. This is particularly relevant for systems such as storage (previously called ESFR) sprinkler systems that have been designed to suppress the fire, In some circumstances, such as spaces with very high ceiling heights (e.g tall atrium voids) the sprinkler system may not control the fire size, The maximum fire size should then be determined based upon fuel and ventilation conditions, ‘There are several computer programs and empirical models available to predict sprinkler activation time and fire size upon sprinkler activation. The sprinkler activation time is the time when the sprinkler bulb reaches the activation temperature, Convective heat transfer from the gases in the ceiling jet to the sprinkler bulb is the primary heat transfer mechanism. However, for a small enclosure ‘where the ceiling jet is immersed within a hot upper smoke layer, additional heat transfer from the hot upper layer to the sprinkler bulb also occurs. There are also heat transfer losses from the sprinkler bulb to the sprinkler pipework’, Radiant heat transfer during the early stages of a fire, when sprinkler activation is expected, is generally neglected ‘The most appropriate model to use will depend on the building being assessed. An important factor is the geometry of the area being protected by sprinklers, For an enclosed room the upper hot layer will significantly impact on the time for sprinkler activation. For enclosed rooms the use of equations developed for unconfined ceilings may therefore significantly over-estimate the time for sprinkler operation" The location of the sprinkler below the ceiling should also be considered. Some models assume that the sprinkler head is located at the highest temperature and velocity of the ceiling jet while others considers the actual location of the sprinkler head below the ceiling. Sprinkler activation models are generally not developed for flush mounted or recessed sprinkler heads! It should also be noted that experiments indicate the sprinkler activation time is hard to predict when. the sprinkler head activation temperature is close to the ceiling jet temperature™” Consideration of ceiling obstructions such as beams should also be taken into account, as they may delay sprinkler activation, Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIBTY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES 22 0f35 [Link] Impact of a Smoke Control System ‘The impact of a smoke control system on the HRR must be considered during the design process because additional air provided by a smoke control system could potentially increase the rate of air ‘entrainment into a fire plume. Research has shown that the burning rate may be increased by a factor of two or three, which could have an impact on the HRR™!. In tunnels, the HRR of a heavy goods vehicle could increase in value by a factor of 4 to 10 depending on the ventilation flow ratel™! Examples of smoke control systems that could have an impact on the HRR include, but not limited to: 1, Activation of longitudinal jet fans in tunnels to prevent back-layering; 2. High velocity make-up air near the fire origin, If Storage Sprinklers are installed in the building to provide fire suppression in high challenge fire risks, special consideration must be given to the smoke control system, if installed. This sprinkler system has special design requirements and limitations in respect to the building structure and the system application may potentially have an impact on the smoke control system, ic. delaying the activation time of smoke control system, Similarly, activation of the smoke exhaust system may negatively impact on the sprinkler system. 5.7.5 Impact of Building Element Failure Building clement failure may alter the HRR of a fire, A typical building clement failure is window lass breaking, The consequence of this is the creation of additional ventilation which increases the RR of ventilation controlled fires. In other cases, such as failure of a fire door, additional fuel may be involved in the fire development. Depending on the sophistication of the models used for fire simulation, if separate models have to be used to predict window breaking time and extent, then the specification of the change of ventilation boundary condition relies on an iterative approach. A trial run of a fire model may be needed to ‘evaluate time and extent of failure. Then the impact on the HRR can be estimated and the design fire ‘modified for the subsequent fire simulations. 5.7.6 Duration of Burning Stage The fully developed fire is most important when considering property protection and structural stability. During this stage the fire is usually ventilation controlled. The duration depends on the total ‘amount of fuel and the available ventilation. Time temperature curves and empirical calculations are available to determine the duration of the burning stage. 5.1.7 Decay Stage The requirement to assess the decay phase is dependent on the design issue. The fire duration may need io be determined to cover the required safe egress time, structural stability ofthe building or time for fire brigade intervention, with an adequate and appropriate margin, Decay occurs as the fuel becomes consumed, and the HRR declines. Normally this is assumed to being. ‘when 70-80 % of the fuel has been consumed’, The fire may change from ventilation controlled to fuel controlled during this period. So far, rates of decay of temperature can only be established empirically or by making conservative or highly idealized assumptions”! Version 1.0, dated 2203.12ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 23 of 35 6 Example Application on a Low Rise Shopping Centre ‘The building isa single storey retail mall with a basement level carpark. The building is a large isolated building greater than 3,500 m? of Type C construction with Deemed-to-Satisfy compliant perimeter access. The building does not exceed 18,000 min floor area nor 108,000 m* in volume. ‘The retail area contains a number of speciality shops < 1,000 m? in area with one major tenaney > 2,000m: in area, The shops are connected by a mall with a number of kiosks within the mall. See Figure 6.1. The basement carpark contains more than 40 vehicles and is connected to the retail level by a travelator separated at basement level by drencher protected glass. ‘The building is fully sprinkler protected on both levels. The major tenancy and the mall are provided with automatic smoke exhaust systems, activated by smoke detectors installed in accordance with ASINZS 1668.1. Basement level is provided with mechanical carpark exhaust ventilation only. BCA Decmed-to-Satisfy non-compliances to be addressed arc: + Extended travel to point of choice, nearest exit and between altemative exits in the retail mall, ‘major tenancy and carpark. © Engineered smoke exhaust rate and no smoke baffle in the major tenancy and the mall © Glazed lobby between retail / earpark — ie. lack of fire separation between classifications. For this stage of the Practice Note, the event trees are provided as a guide to the process only ‘Therefore, the quantification of probabilities has not been included. Version 1.0, dated 2203.12aaquag Suyddoyg asry MoT] & Jo node] IpEMUDNDS :1°9 24N5Ly ee PNP HOME, (m1 000¢ ~) wera s0fe 14 sory asnoy Jo eg wom ‘Surpeo7y SEO PEE ‘SRI NOISTC WOd SLON AOTLOVWd ALHAVS RIL AO ALAIOOS VIIVALSAV SWAINIDNAENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY development, spread and control), D (lire detection, warning and suppression), E (occupant evacuation and control) and F (Fire Services intervention) development, spread and control), D (fire detection, warning and suppression), E (occupant evacuation and, control) and F (Fire Services intervention) PRACTICE NOTE FOR DESIGN FIRES Page 25 of 35 Table 7-3: Low Rise Shopping Centre Example Egress ‘Smoke exhaust | Compartmentation Step 1 —identify | Class 6 retail and Class 7a carpark. design occupancy Step 2—identily | Extended travel 0 | Engineered smoke] Glazed lobby design issues point of choice, | exhaust rate and no_| between retail / nearest exitand | smoke baffle in | carpark between alternative | major tenancy and exits in the retail | mall small, major tenancy and carpark Step 3— identify | A (fire initiation, | A (fire initiation, | A (ire initiation, relevant development and | development and _| development and subsystems control), B (smoke | control), B (smoke | control), (ire spread, impact and control) and D (fire detection, warning and suppression) Step 4— identity design fire Event tree, see Figure 7-4 and Figure 7-5 Event tr Figure 7 see Event tree, see Figure 7- Step 5 — determine characteristies of design fire Step Sa - location Major — centre of ‘tenaney and corner of tenancy Mall —in a speciality shop and middle of mall Carpark — centre of carpark Major — centre of tenaney, and comer of tenancy (blocking exit) Mall — ina speciality shop and mall food court (main exit blocked) Carpark — adjacent to travelator Step Sb— type of fire Flaming ¢ growth until sprinkler activation Redundancy (sprinkler failure) — growth to peak, HRR Flaming ¢ growth until sprinkler activation Redundancy (sprinkler failure) — growth to peak, HRR Flaming t growth until sprinkler activation Redundancy (sprinkler failure) ~ growth (o peak HRR Version 1.0 dated 2.03.12ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY radiative fraction PRACTICE NOTE FOR DESIGN FIRES Page 26 of 35 Egress ‘Smoke exhaust | Compartmentation Step Sb—type & | Retail - mix of Mix of plastic and | Car fire up to amount of fuel | plastic and cellulosic materials | 200 Mim? cellulosic materials | up to 1300 MJim* up to 1300 Mim? Carpark — car fire up to 200 Mim? Step Se - 03 03 03 Step Sd— soot & species yields & heat of combustion Retail — average value based on mix of plastic and cellulosic materials; CO; -1.8 e/g CO-0.04 gig Soot - 0.06 gig AH, ~ 30 Mike, Car—average value based on mix of plastic materials and gasoline; CO; -2.0 g/g CO-0.04 gig Soot - 0.08 g AH, ~34 Mikg. “Average value based on mix of plastic and cellulosic materials; COs-18 gig 0 - 0.04 pie. Soot - 0.06 gig H, = 30 Muikg “Average value based on mix of plastic materials and gasoline; CO; =2.0 gig CO-0.04 gig Soot - 0.08 gig AH, -34 Miike Verso 1.0 dated 2.03.12ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 27 of 35 Egress ‘Smoke exhaust | Compartmentation Step Se-heat | Retail — ‘no incipient ‘+ no incipient release rate « no incipient growth growth growth, fast & growth + medium ¢ growth « fast & growth extinguished by | « controlled by extinguished by | sprinkler sprinkler sprinkler activation activation to one activation ‘no impact from the | ‘no impact from the smoke exhaust + no building failure HR decay based con sprinkler activation ‘redundancy ~ fast growth to peak HRR based on available fuel Campark — no ineipient growth ‘* medium t growth * controlled by sprinkler activation to one car « no building failure © HR constant until occupant ogress and fire brigade arrival redundancy — fast # growth to peak HRR based on available fuel smoke exhaust no building failure + HR decay based ‘on sprinkler activation ‘redundancy — fast # growth to peak HRR based on available fuel + no building failure © ARR constant until occupant egress and fire brigade arrival ‘* redundancy ~ fast growth to peak RR based on available fuel Version 1.0 dated 2.03.12vee PNP HONEA, Aoueuay, yey9y soley Ut 994, UAT 2p-L 2undIy, oo ey ——— Avan tna rE“ amy Jonni \ EET i found sous pense Saas spon 5630 aC oA ‘SHRI NOISTC WOd SLON AOTLOVWd ALAAVS RIL AO ALAIOOS VITVALSNV SWAINIONAee PNP HONEA, We wy a2ay, uaa 25-1 aunty, 5630 6c oA SHRI NOISTC WOd SLON AOTLOVWd ALAAVS RIL AO ALAIOOS WITVALSNV SWAINIDNAoro nueue eo [Link]> se Gauci) ‘baa womans even '0 {won saedan uy 994, 9A 19-1 94n cane arom sor| poems soiiae| / amoral! = a ——_/ aepresuna or \ ee] £30 05 oa SRI NOIS9C WOd SLON AOTLOVWd ALAAVS @RIla AO ALAIOOS WIIVALSNV SWAINIONALENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRI 10, 13 Page 31 of 35 References Australian Building Codes Board, International Fire Engineering Guidelines, Canberra, Australia, 2005, Frantzich, H., Magnusson, S. E., Holmquish, B., Ryden, J., Derivation of Partial Safety Factors for Fire Safety Evaluation Using the Reliability Index Beta Method. In: Hasemi Y (ed). Proceedings of the Fifth International Symposium on Fire Safety Science, Melboume: International Association for Fire Safety Science, 1997: 667-678. He, Y., Horason, M., Taylor, P., Ramsay, C. Stochastic Modelling for Risk Assessment. In: Evans DD (ed). Fire Safety Science — Proceedings of the Seventh (7th) International Symposium. Worcester, MA, USA: Intemational Association for Fire Safety Science, 2003: 333-344, ATS 5387.2, Australian Technical Specification, Guidelines - Fire safety engineering, Part 2: Design fire scenarios and design fires. Standards Australia, 2006, Hadjisophocleous, G.V., Mehaffey, J. R., Fire Scenarios. In: DiNenno P.I (ed), SFPE Handbook of Fire Protection Engineering 4” ed, Society of Fire Protection Engineers, USA, 2008: 5.186-5,205. ISO/TR-13387, Fire Safety Engineering. International Standards Organisation, 1999. Hurley, MJ. and Rosenbaum, E.R, Performance-Based Design, In: DiNenno P.J. (ed), SFPE Handbook of Fire Protection Engineering 4” ed. Society of Fire Protection Engineers, USA, 2008: 3.440-3.455. CIBSE, Fire Engineering CIBSE Guide E, The Chartered Institution of Building Services Engineers London, 2010. Joglar, F., Reliability, Availability, and Maintainability. In: DiNenno, PJ, (cd), SFPE Handbook of Fire Protection Engineering 4th ed. Society of Fire Protection Engineers, USA, 2008: 5.25-5.68 Watts, J MJ., Fire Risk Indexing. In: DiNenno PJ. (ed), SFPE Handbook of Fire Protection Engineering 4" ed. Society of Fire Protection Engineers, USA, 2008: 5.168-5.185, Johnson, P., Gildersleeve, C., Boverman, D., Design Fires "How do We Know We Have Them Right?," Charting the Course ~ Proceedings of Society of Fire Safety 10% International Conference, 2008. Bennetts, LD., Thomas, LR., Poh, K.W., Design of Sprinklered Shopping Contre Buildings for Fire Safety, Onesteel - Market Mills, Newcastle Australia, November 2000, BSI, PD 7974-6:2004, The Application of Fire Safety Engineering Principles 10 Fire Safety Design of Buildings ~ Human Factors: Life Safety Strategies — Occupant Evacuation, Behaviour and Conditions, July 2004, Verso 1.0 dated 2.03.12ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRI 14, 15 16. 17, 18 24 25, 26, 27. 28, Page 32 of 35 NFPA 5000, Building Construction and Safety Code, NFPA 2009. Drysdale, D.D., An Introduction to Fire Dynamics, 2” Edition, Chichester, UK., 1999, John Wiley and Sons. ‘Tewarson, A., Generation of Heat and Chemical Compounds in Fires. In: DiNenno P.J. (ed), SEPE Handbook of Fire Protection Engineering 4" ed, Society of Fire Protection Engineers, USA, 2008: 3.109-3.194, Gottuk, D., Mealy. C., Floyd. J., Smoke Transport and FDS Validation, Fire Safety Science - Proceedings of the Ninth International Symposium, International Association for Fite Safety Science, 2008, pp. 9-129 to 9-140. Appendix C, Fuel Properties and Combustion Data, In: DiNenno P.J. (ed), SEPE Handbook of Fire Protection Engineering 4" ed. Society of Fire Protection Engineers, USA, 2008: A34- A43, Babrauskas, V., Tables and Charts, In: Cote A.E. (ed), Fire Protection Handbook 20” ed. National Fire Protection Association, USA, 2008: Table 6.17.1 = 6.17.3. Fire Engineering Design Guide 2008, 3" Edition, M. Spearpoint (ed), New Zealand Cente for Advanced Engineering, Christchurch, New Zealand. Babrauskas, V., Peacock, R. D., Heat Release Rate: The Single Most Important Variable in Fire Hazard. Fite Safety Journal No 18, 1992: 255-272. Babrauskas, V., Heat Release Rates. In Dinenno, P.J. (ed), SFPE Handbook of Fire Protection Enginecring 4" ed. Society of Fire Protection Engineers, USA, 2008: 3.1 -3.59. BS EN 1991-1-2:2002, Eurocode 1: Actions on Structures - Part 1.2 General Actions on Structures Exposed to Fire (BS1 2002). BS 9999-2008, Code of practice for fire safety in the design, management and use of buildings (BSI 2008). NFPA 92, Standard for Smoke Control Systems, National Fire Protection Association, 2012, MeDermott, R, etl, Fire Dynamics Simulator (Version 5) Teclnical Reference Guide Volume 2: Verification, NIST Special Publication 1018-5, National Institute of Standards and Technology, 2010 ‘Australian Building Codes Board, BCA 2010, Building Cade of Australia, Class 2 to Class 9 Buildings, Volume 1, Australian Building Codes Board, Canberra, Australia, 2010. Custer, RLP., Meacham B.J., and Schifiliti R-P., Design of Detection ‘Systems. In: DiNenno P.J. (ed), SEPE Handbook of Fire Protection Engineering 4" ed, Society of Fire Protection Engineers, USA, 2008: 4.1-4.44. Verso 1.0 dated 2.03.12ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRI 31 33, Page 33 of 35 Fleming, R.P., Automatic Sprinkler System Calculations. In; DiNenno P.J. (ed), SEPE Handbook of Fire Protection Engineering 4” ed. Society of Fire Protection Engineers, USA, 2008: 4.72-4.88. Madrzykowski, D., Evaluation of Sprinkler Activation Prediction Methods, ASIAFLAM’95, International Conference on Fire Science and Engineering, 1" Proceedings, 1995: 211-218, Wade, C., Spearpoint, M.J, Bitten, A., Tsai, K., Assessing the Sprinkler Activation Predictive Capability of the BRANZFIRE Fire Model. Fire ‘Technology, Vol 43, No 3, September, 2007: 175-193, Maudrzykowski, D., The Effect of Recessed Sprinkler Installation on Sprinkler Activation Time and Prediction, Master Thesis, University of Maryland, 1993 He, Y., Effect of Pressurisation and Smoke Management Systems on Fire Growth and Smoke Movement in a Multi-Storey Building. International Journal on Engineering Performance-Based Fire Codes, Vol. 1, No. 3, 1999: 148-155, Beard, A., Carvel, R., 2005: “The handbook of tunnel fire safe Telford Publishings, 1 Heron Quay, London E14 4JD, UK, 2008. ‘homas Staffansson, L., “Selecting Design Fires”, Report 7032, Department of Fire Safety Engineering and Systems Safety, Lund University, Sweden, 2010, Verso 1.0 dated 2.03.12ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRI Page 34 0f 35 Appendix A Event Trees In fire safety engineering analysis, the events in the event tree can be associated with the fire safety subsystems as described in the IFEG. The status of a subsystem defines an event, ot a node, in the event tree (see Figute 2-1), and each event has the associated probability of attainment, py (1, 2, .-. my and j-1, 2, .., N) where iand | are indices of a node, m is the number of possible status of the jth subsystem and Vis the total number of subsystems involved. The probability of a scenario Py (1, 2, ... ‘M) can be calculated by: (any where M is the total number of scenarios and co) well, In the event tree, as depicted in Figure 3-1, the subsequent events represent the outcomes of the bi-valued subsystems, or the subsystems that attain either the status of operation or failure, with associated probabilities. The probability of operation is also referred to as reliability. If the reliability of a subsystem is independent of the status of essor or precursor, some of the probability values at the same level of the tevent tree may be equal. For example, it is possible that: (A3) Py = Px and prs Furthermore: Pot Pa = Prt Pas = ay) Generally, Eq (A-3) may not hold and the status of subsystem may depend on the status of its predecessor and so may the probabilities. For example, ifthe sprinkler subsystem has successfully suppressed a fire, the probability of the fire spreading to adjacent compartments diminishes. It is possible for a subsystem to have more than two outeomes. For example, the interaction between a fire and a sprinkler subsystem may be suppression, control or failure, Each of these outcomes will define a scenario and have the associated design fire and probability Verso 1.0 dated 2.03.12ENGINEERS AUSTRALIA SOCIETY OF FIRE SAFETY PRACTICE NOTE FOR DESIGN FIRES Page 35 of 35 Generally, the sum of the probabilities of all possible status of a subsystem should be unity: Lev=1 (as) (a6) Equations (A-5) and (A-6) can be used to verify the event tree analysis. Verso 1.0 dated 2.03.12