Fire Safety Engineering
Sometimes building codes are not applicable or perhaps do not deal with a particular building configuration and fire scenario and it is here that a Fire Safety Engineer may be able to assist in the design of a building which otherwise would not be code compliant.
Fire safety engineering is a method of design which satisfies the functional requirements of building regulations without following the prescriptive codes. A Fire Safety Engineer can overcome problems associated with fire risk, for example, separation distances between buildings to prevent fire spread, open plan conditions, number of staircases, travel distance, compartment size and level of fire resistance. Sometimes smoke control calculations are made to show that there is adequate time for escape before conditions become untenable. Fire Safety Engineering solutions are increasingly accepted by fire regulators.
Some of the most challenging projects in London are conversions in residential and commercial buildings which may not satisfy the prescriptive guidance in the regulations and codes and therefore require a fire safety engineered approach using compatible passive and active fire defence measures.
In all this work it is essential to gain a complete understanding of the 3-dimensional geometry and compartmentation of the building so that all possible paths of fire effluent can be considered when deciding on the tenability of humans who may be anywhere in or near the building. Only then can an effective fire safety strategy be developed that satisfies the functional fire regulations and other requirements.
Prescriptive codes – the move away from
The effects of fire in or near buildings are mitigated by fire precautions required by national regulations and codes of practice. Over the past four decades building regulations have moved away from comprehensive prescriptive regulations to functional regulations supported by detailed technical guidance. It is generally accepted that existing prescriptive guidance, such as Approved Document B (AD B), is not suited to large, less compartmented and more complex buildings and is not suited to buildings which are refurbished when many of the features required simultaneously by prescription do not exist, for instance in single staircase buildings with only one obvious escape route.
Fire Safety Engineering – the principles
While the present combination of regulations and guidance has served well, and will continue to be appropriate for simple buildings, there remains a pressing need for wider acceptance of Fire Safety Engineering (Fire Safety Engineering) – see definition below. This should include performance-based principles, acceptance criteria, and calculation tools which permit the use of time-based calculations to address the important relationship between the time required for escape (which is determined by people behaviour, internal planning of the building and fire defence systems) and the time available for escape (which is determined by the human tenability of the fire effluent for each person at his or her location in the building). It seems generally true that performance based guidance is weakest in means of escape and the challenge here is to gain further information on the time it takes for people to get away from danger in, for instance, shops, public assembly and institutional buildings where large numbers of people, including disabled people, may be present. To accomplish this goal a methodology based on sound engineering principles, which employs calculation tools and expert judgement, is needed. This methodology is called fire safety engineering
Fire Safety Engineering – a definition
Fire Safety Engineering may be defined as the application of engineering principles, rules and expert judgement, based on a knowledge of human behaviour and a scientific understanding of the phenomena of fire and its effects, to:
- save life, protect property and preserve the environment and heritage
- quantify the hazards and risk of fire and their effects
- evaluate analytically the optimum protective and preventative measures necessary to limit within prescribed levels the consequences of fire.
Fire Safety Engineering – the challenge
The task in Fire Safety Engineering is daunting and exciting due to the many imponderables in fire. Will the fire be accidental or malicious? In which room will the fire occur? What will be the item first ignited? Will doors be open or closed? What will be the size of design fire? Will the suppression system operate and will it control the fire? Where will the occupants be located? How soon will occupants be aware of the fire and how rational will be their escape behaviour? These kinds of questions cannot be quantified in deterministic design but can be accounted for in probabilistic design if appropriate statistics are available. It is the availability of such statistics which is currently limiting probabilistic design quite apart from the need for all members of the design and enforcement team to be appropriately educated in quantitative risk assessment. In deterministic design, calculations are made assuming a limited range of plausible worst-case fire scenarios, and often accompanied by a large measure of engineering judgement and experience. An example of the deterministic approach and methodology can be seen in the first paper listed in My Fire Safety Publications under Fire Safety Engineering – click here to see
Fire Safety Engineering – how is it used?
If you should ask ‘What kind of guidelines, standards, computer models are used as Fire Safety Engineering tools?’, my answer would be:
“UK building regulations are functional, not prescriptive. For England and Wales the official guidance on how you satisfy the regulations is in a government document, Approved Document B ‘Fire Safety’ (AD B), which states that the prescriptive guidance in the AD does not have to be used and it allows a fire safety engineering approach to be used. It is similar in Ireland and Scotland. Technical guidance on fire safety engineering is given in a set of BSI documents in the series PD 7974 under the umbrella of BS 7974: 2001 ‘Application of fire safety engineering principles to the design of buildings’. This series replaces BSI DD 240 with which I was very much involved. The recently published BS 9999 sits between Approved Document B and the PD 7974 series in terms of technical complexity.
Computer models are used for calculating smoke fill times, sometimes using computational fluid dynamics (CFD) for very large and complex building geometries. However, in all my audits of the work of fire safety engineers, 2-zone plume equations have been employed for smoke fill type problems using a spreadsheet such as Excel for number-crunching and making parametric sensitivity studies. Computer-aided numerical models are sometimes used for people-evacuation modelling of complex large spaces. Computer models are not often used for calculations of structural fire resistance except when it is possible to show that steelwork may be used with less fire protection or none at all. Simple computer models are used for predicting thermal radiation hazards to nearby buildings, but not often. In my experience most consulting fire engineers are not making probabilistic analyses for ordinary buildings – such analyses are confined to nuclear power stations, petrochemical plant and sometimes in hospitals. Interestingly, in the UK we are free to use whatever technical guidance is available and appropriate which may come from ISO, CEN, or BSI or even a professional body within the UK (such as CIBSE on smoke management) or elsewhere (such as American NFPA standards), but this variation and freedom then makes it difficult for the checking and approving authorities.
Zone models and their use and limitations in Fire Safety Engineering.
A fire zone model is a calculation method for predicting the fire effects within an enclosure. The calculations are based on the conservation of mass and energy applied separately to control volumes that subdivide an enclosure into one or more zones. At any instant in time, the properties of each zone are assumed to be uniform throughout the layer: The fire is treated as a source of mass and energy and is a user-prescribed input to the calculation. A fire zone model is most commonly a numerical fire model in the form of a computer program but calculations can be done using spreadsheet applications or even by hand.
Most commonly used zone models comprise two zones in the form of a hot upper layer and a cool lower layer. This provides sufficient resolution for many simple pre-flashover fire simulations. The author has presented two example calculations using zone models and these are given in www.cookeonfire.com>About Me>My Calculations. One zone models have also been developed e.g for fully developed postflashover room fires where the assumption of a well-mixed uniform zone may be reasonable. Alternatively, the fundamental equations for mass and energy conservation can also be extended to more than two zones to provide greater resolution over the height of a compartment.
The design fire is the quantitative description of the fire characteristics to be used by the zone model. The most important input variable influencing the course of the fire and in particular the gas temperatures reached is the rate of the heat release as a function of time. If a zone model is used to assess factors that are dependent on the smoke density such as visibility through smoke or the response of optical smoke detectors, appropriate selection of the smoke/soot yield input parameter is also very important.
Since in most zone models, the design fire characteristics are input variables to be provided by the user, it is critical that for design applications in particular, the project stakeholders and regulators agree beforehand on the design fire to be used including rate of heat release and species production rates. Guidance on selection of design fire scenarios and design fires is given in ISO/TS 16733 and in BS 7974 series.
Zone models usually require enclosures to be represented as rectangular volumes with uniform cross-sectional area over their height. Non rectangular spaces can be modelled as an equivalent rectangular volume such that the enclosure area and height are kept the same (and hence volume is the same). The user can vary the length and width of the enclosure as needed. It is important that the height is conserved because the plume entrainment and smoke production calculations are strongly influenced by the vertical distance between the fire and the smoke layer interface. Some zone models also allow for variations in cross section area over the height (e.g. for a sloping ceiling) in the calculation of the upper layer volume and layer height.
Applications of a zone model include: predicting the smoke-filling time for a compartment of a given size and for a fire with known time-dependent characteristics; evaluating the life safety tenability of a fire environment for comparison with design criteria; reconstructing a past fire event to support or refute theories about the development of a fire;
determine the likely fire size at the time of sprinkler operation (where a detector/sprinkler submodel is included); determining the response time of a detector or sprinkler (where a detector/sprinkler submodel is included); determining the smoke extract capacity for naturally or mechanically ventilated spaces;
determining the impact on important equipment to ensure its continued functionality.
The simplified physics included in zone models mean they are less computationally demanding and are relatively quick to run compared to state-of-the-art models (e.g. CFD computational fluid dynamics) that attempt to describe the physics using the best available methods. Zone model simulations may therefore have greater limitations in predicting the fire environment than more detailed models for a known scenario. There are also advantages in being able to run a larger number of simulations for the same computing resource. This allows the sensitivity of the results to the various input parameters to be investigated in greater depth. This is particularly useful for design applications where the exact fire parameters are not known, yet may have a large influence on the predicted outputs. The accuracy achieved by a zone model can be sufficient for a specific application from the engineering or authorities viewpoint. CFD models may not be more accurate for an application, given the uncertainties involved. It is often necessary to provide validatory data resulting from a fully instrumented fire test.
Finite element method of temperature analysis for structures.
Behaviour of structures in fire is also amenable to computation, typically by use of the finite element method requiring the use of computers to solve hundreds of simultaneous equations. Load-bearing capacity decreases with temperature while the deformation generally increases with temperature. Both of these quantities require knowledge of thermal and mechanical properties which vary with temperature. Computation of temperatures reached in structures exposed to fire is possible using a number of commercially available softwares.
A powerful program for calculating temperatures in sections is TASEF and its use has been simplified with related software TASEFplus enabling it to be used by fire consultants, researchers and students. The incentive to use such programs is that accurate calculations yield, in general, considerably lower temperatures than simple standard predictions. Therefore the costs of fire protection of e.g. steel structures can be reduced.
TASEF handles plane or axi-symmetric cross-sections. TASEF uses the finite element method for thermal analysis in two dimensions. The meshing of the cross-section is shown by a graphical interface. Cut outs and voids can be incorporated in the section geometry.
Certain standard fires are built into TASEFplus including ISO 834, EN 1363-1 and ASTM E-119. There is a facility to adopt Eurocode parametric fires. Completely customised fires can also be specified. The user specifies the times at which the temperature output is desired. The program can handle several fires for various boundaries of the cross-section. These include typical fires defined by time temperature curves, incident heat flux by radiation combined with convection heat transfer from adjacent gases, boundaries with prescribed temperatures that may follow a specified fire curve, as well as ambient non-fire boundaries. A special feature is the possibility to model heat transfer by radiation and convection in voids.
Thermal material properties like conductivity and specific volumetric enthalpy (density and specific heat capacity) can vary with temperature and latent heat due to water evaporation can be modelled. Properties of steel and concrete based on Eurocodes 2-4 are built into the input generator TASEFplus. Completely customised material properties can also be defined.
TASEFplus has the facility to view colour contours for calculated temperature distributions. In addition, a completely annotated text file output is generated. TASEFplus including TASEF is supplied with a detailed User Manual and Example Manuals in which the user is guided step-by-step through specific problems, see www.tasefplus.com
Fire Safety Engineering – my involvement
I have been closely associated with fire safety engineering from its beginnings in the UK and was a member of the small group contracted by government (DTI) to prepare the first British Standard DD240 on the subject. I was also fortunate to get Personal Promotion in BRE to focus on Fire Safety Engineering and have published several papers on Fire Safety Engineering- click here to see. I have been an active member of BSI committee FSH24 Fire Safety Engineering and also remain very active in the International Standards Organisation TC 92 SC4 on Fire Safety Engineering.
Some of my PowerPoint presentations on Fire Safety Engineering (in PDF format)
- Upgrading the fire behaviour of heritage buildings – click here to see
- Design against chip-pan fires – click here to see
- Behaviour of concrete in fire – click here to see
- Sandwich panels in fire (stability theory) – click here to see
- Sandwich panels in fire (surprises) – click here to see
- Some projects – click here to see
- Resistance of tall buildings to impact and fire – click here to see
- Ad hoc fire test for Eurobond – click here to see
- Fire phenomena and defects observed – click here to see